AI 日报

Nvidia, Fujitsu Unveil AI, Robotics Partnership

The partners said their expanded collaboration will help support the 'AI industrial revolution.'

OpenAI and AMD Strike Major AI Partnership

The multibillion-dollar deal between OpenAI with AMD for compute power could make AMD a key competitor to AI chip giant Nvidia.

OpenAI unveils AgentKit that lets developers drag and drop to build AI agents

OpenAI launched an agent builder that the company hopes will eliminate fragmented tools and make it easier for enterprises to utilize OpenAI’s system to create agents.

AgentKit, announced during OpenAI’s DevDay in San Francisco, enables developers and enterprises to build agents and add chat capabilities in one place, potentially competing with platforms like Zapier.

By offering a more streamlined way to create agents, OpenAI advances further into becoming a full-stack application provider.

“Until now, building agents meant juggling fragmented tools—complex orchestration with no versioning, custom connectors, manual eval pipelines, prompt tuning, and weeks of frontend work before launch,” the company said in a blog post.

AgentKit includes:

• Agent Builder, which is a visual canvas where devs can see what they’ve created and versioning multi-agent workflows

• Connector Registry is a central area for admins to manage connections across OpenAI products. A Global Admin console will be a prerequisite to using this feature.

• ChatKit enables users to integrate chat-based agents into their user interfaces.

Eventually, OpenAI said it will build a standalone Workflows API and add agent deployment tabs to ChatGPT.

OpenAI also expanded evaluation for agents, adding capabilities such as datasets with automated graders and annotations, trace grading that runs end-to-end assessments of workflows, automated prompt optimization, and support for third-party agent measurement tools.

Developers can access some features of AgentKit, but OpenAI is gradually rolling out additional features, such as Agent Builder. Currently, Agent Builder is available in beta, while ChatKit and new evaluation capabilities are generally available. Connector Registry “is beginning its beta rollout to some API and ChatGPT Enterprise and Edu users.

OpenAI said pricing for AgentKit tools will be included in the standard API model pricing.

To clarify, many agents are built using OpenAI’s models; however, enterprises often access GPT-5 through other platforms to create their own agents. However, AgentKit brings enterprises more into its ecosystem, ensuring they don’t need to tap other platforms as often.

Demonstrated during DevDay, the company stated that Agent Builder is ideal for rapid iteration. It also provides developers with visibility into how the agents are working.

During the demo, an OpenAI developer made an agent that reads the DevDay agenda and suggests panels to watch. It took her just under eight minutes.

Other model providers saw the importance of offering developer toolkits to build agents to entice enterprises to use more of their tools. Google came out with its Agent Development Kit in April, expanding multi-agent system building “in under 100 lines of code.” Microsoft, which runs the popular agent framework AutoGen, announced it is bringing agent creation to one place with its new Agent Framework.

OpenAI customers, such as the fintech company Ramp, stated in a blog post that its teams were able to build a procurement agent in a few hours instead of months.

“Agent Builder transformed what once took months of complex orchestration, custom code, and manual optimizations into just a couple of hours. The visual canvas keeps product, legal, and engineering on the same page, slashing iteration cycles by 70% and getting an agent live in two sprints rather than two quarters,” Ramp said.

AgentKit’s Connector Registry would also enable enterprises to manage and maintain data across workspaces, consolidating data sources into a single panel that spans both ChatGPT and the API. It will have pre-built connectors to Dropbox, Google Drive, SharePoint and Microsoft Teams. It also supports third-party MCP servers.

Another capability of Agent Builder is Guardrails, an open-source safety layer that protects against the leakage of personally identifiable information (PII), jailbreaks, and unintended or malicious behavior.

Since most agentic interactions involve chat, it makes sense to simplify the process for developers to set up chat interfaces and connect them with the agents they’ve just built.

“Deploying chat UIs for agents can be surprisingly complex—handling streaming responses, managing threads, showing the model thinking and designing engaging in-chat experiences,” OpenAI said.

The company said ChatKit makes it simple to embed chat agents on platforms and embed these into apps or websites.

However, some OpenAI competitors have begun thinking beyond the chatbot and want to offer agentic interactions that feel more seamless. Google’s asynchronous coding agent, Jules, has introduced a new feature that enables users to interact with the agent through the command-line interface, eliminating the need to open a chat window.

The response to AgentKit has mainly been positive, with some developers noting that while it simplifies agent building, it doesn’t mean that everyone can now build agents.

Several developers view Agent Kit not as a Zapier killer, but rather as a tool that complements the pipeline.

Zapier debuted a no-code tool for building AI agents and bots, called Zapier Central, in 2024.

OpenAI announces Apps SDK allowing ChatGPT to launch and run third party apps like Zillow, Canva, Spotify

OpenAI's annual conference for third-party developers, DevDay, kicked off with a bang today as co-founder and CEO Sam Altman announced a new "Apps SDK" that makes it "possible to build apps inside of ChatGPT," including paid apps, which companies can charge users for using OpenAI's recently unveiled Agentic Commerce Protocol (ACP).

In other words, instead of launching apps one-by-one on your phone, computer, or on the web — now you can do all that without ever leaving ChatGPT.

This feature allows the user to log-into their accounts on those external apps and bring all their information back into ChatGPT, and use the apps very similarly to how they already do outside of the chatbot, but now with the ability to ask ChatGPT to perform certain actions, analyze content, or go beyond what each app could offer on its own.

You can direct Canva to make you slides based on a text description, ask Zillow for home listings in a certain area fitting certain requirements, or ask Coursera about a specific lesson's content while dit plays on video, all from within ChatGPT — with many other apps also already offering their own connections (see below).

"This will enable a new generation of apps that are interactive, adaptive and personalized, that you can chat with," Altman said.

While the Apps SDK is available today in preview, OpenAI said it would not begin accepting new apps within ChatGPT or allow them to charge users until "later this year."

ChatGPT in-line app access is already rolling out to ChatGPT Free, Plus, Go and Pro users — outside of the European Union only for now — with Business, Enterprise, and Education tiers expected to receive access to the apps later this year.

Built atop common MCP standard

Built on the open source standard Model Context Protocol (MCP) introduced by rival Anthropic nearly a year ago, the Apps SDK gives third-party developers working independently or on behalf of enterprises large and small to connect selected data, "trigger actions, and render a fully interactive UI [user interface]" Altman explained during his introductory keynote speech.

The Apps SDK includes a "talking to apps" feature that allows ChatGPT and the underlying GPT-5 or other "o-series" models piloting it underneath to obtain updated context from the third-party app or service, so the model "always knows about exactly what you're user is interacting with," according to another presenter and OpenAI engineer, Alexi Christakis.

Developers can build apps that:

• appear inline in chat as lightweight cards or carousels

• expand to fullscreen for immersive tasks like maps, menus, or slides

• use picture-in-picture for live sessions such as video, games, or quizzes

Each mode is designed to preserve ChatGPT’s minimal, conversational flow while adding interactivity and brand presence.

Early integrations with Coursera, Canva, Zillow and more...

Christakis showed off early integrations of external apps built atop the Apps SDK, including ones from e-learning company Coursera, cloud design software company Canva, and real estate listings and agent connections search engine, Zillow.

Altman also announced Apps SDK integrations with additional partners not demoed officially during the keynote including: Booking.com, Expedia, Figma and Spotify and in documentation, said more upcoming partners are on deck: AllTrails, Peloton, OpenTable, Target, theFork, and Uber, representing lifestyle, commerce, and productivity categories.

The Coursera demo included an example of how the user onboards to the external app, including a new login screen for the app (Coursera) that appears within the ChatGPT chat interface, activated simply by a text prompt from the user asking: "Coursera can you teach me something about machine learning"?

Once logged in, the app launched within the chat interface, "in line" and can render anything from the web, including interactive elements like video.

Christakis explained and showed the Apps SDK also supports "picture-in-picture" and "fullscreen" views, allowing the user to choose how to interact with it.

When playing a Coursera video that appeared, he showed that it automatically pinned the video to the top of the screen so the user could keep watching it even as they continued to have a back-and-forth dialog in text with ChatGPT in the typical input/output prompts and responses below.

Users can then ask ChatGPT about content appearing in the video without specifying exactly what was said, as the Agents SDK pipes the information on the backend, server-side, from the connected app to the underlying ChatGPT AI model. So "can you explain more about what they're saying right now" will automatically surface the relevant portion of the video and provide that to the underlying AI model for it to analyze and respond to through text.

In another example, Christakis opened an older, existing ChatGPT conversation he'd had about his siblings' dog walking business and resumed the conversation by asking another third-party app, Canva, to generate a poster using one of ChatGPT's recommended business names, "Walk This Wag," along with specific guidance about font choice ("sans serif") and overall coloration and style ("bright and colorful.")

Instead of the user manually having to go and add all those specific elements to a Canva template, ChatGPT went and issued the commands and performed the actions on behalf of the user in the background.

After a few minutes, ChatGPT responded with several poster designs generated directly within the Canva app, but displayed them all in the user's ChatGPT chat session where they could see, review, enlarge and provide feedback or ask for adjustments on all of them.

Christakis then asked for ChatGPT to turn one of the slides into an entire slide deck so the founders of the dog walking business could present it to investors, which did it in the background over several minutes while he presented a final integrated app, Zillow.

He started a new chat session and asked a simple question: "based on our conversations, what would be a good city to expand the dog walking business."

Using ChatGPT's optional memory feature, it referenced the dog walk conversation and suggested Pittsburgh, which Christakis used as a chance to type in "Zillow" and "show me some homes for sale there," which called up an interactive map from Zillow with homes for sale and prices listed and hover-over animations, all in-line within ChatGPT.

Clicking a specific home also opened a fullscreen view with "most of the Zillow experience," entirely without leaving ChatGPT, including the ability to request home tours and contact agents and filtering by bedrooms and other qualities like outdoor space. ChatGPT pulls up the requested filtered Zillow search as well as provides a text-based response in-line explaining what it did and why.

The user can then ask follow-up questions about the specific property — such as "how close is it to a dog park?" — or compare it to other properties, all within ChatGPT.

It can also use apps in conjunction with its Search function, searching the web to compare the app information (in this case, Zillow) with other sources.

Safety, privacy, and developer standards

OpenAI emphasized that apps must comply with strict privacy, safety, and content standards to be listed in the ChatGPT directory. Apps must:

• serve a clear and valuable purpose

• be predictable and reliable in behavior

• be safe for general audiences, including teens aged 13–17

• respect user privacy and limit data collection to only what’s necessary

Every app must also include a clear, published privacy policy, obtain user consent before connecting, and identify any actions that modify external data (e.g., posting, sending, uploading).

Apps violating OpenAI’s usage policies, crashing frequently, or misrepresenting their capabilities may be removed at any time. Developers must submit from verified accounts, provide customer support contacts, and maintain their apps for stability and compliance.

OpenAI also published developer design guidelines, outlining how apps should look, sound, and behave. They must follow ChatGPT’s visual system — including consistent color palettes, typography, spacing, and iconography — and maintain accessibility standards such as alt text and readable contrast ratios.

Partners can show brand logos and accent colors but not alter ChatGPT’s core interface or use promotional language. Apps should remain “conversational, intelligent, simple, responsive, and accessible,” according to the documentation.

A new conversational app ecosystem

By opening ChatGPT to third-party apps and payments, OpenAI is taking a major step toward transforming ChatGPT from a chatbot into a full-fledged AI operating system — one that combines conversational intelligence, rich interfaces, and embedded commerce.

For developers, that means direct access to over 800 million ChatGPT users, who can discover apps “at the right time” through natural conversation — whether planning trips, learning, or shopping.

For users, it means a new generation of apps you can chat with — where a single interface helps you book a flight, design a slide deck, or learn a new skill without ever leaving ChatGPT.

As OpenAI put it: “This is just the start of apps in ChatGPT, bringing new utility to users and new opportunities for developers.”

There remain a few big questions, namely: 1. what happens to all the data from those third-party apps as they interface with ChatGPT and its users...does OpenAI get access to it and can it train upon it? 2. What happens to OpenAI's once much-hyped GPT Store, which had been in the past promoted as a way for third-party creators and developers to create custom, task-specific versions of ChatGPT and make money on them through a usage-based revenue share model?

We've asked the company about both issues and will update when we hear back.

Eli Lilly to Invest ₹9,000 Crore in Hyderabad, Boost Telangana’s Pharma Leadership

The investment is expected to create thousands of high-value jobs and attract further global partnerships in the sector.

The post Eli Lilly to Invest ₹9,000 Crore in Hyderabad, Boost Telangana’s Pharma Leadership appeared first on Analytics India Magazine.

CoreWeave to Acquire Monolith, Offer Full Stack Research & Design

With this acquisition, CoreWeave will expand into automotive, aerospace, manufacturing and other industries.  

The post CoreWeave to Acquire Monolith, Offer Full Stack Research & Design appeared first on Analytics India Magazine.

‘Almost All New Code Written at OpenAI Today is From Codex Users’

Sam Altman stated that nearly every pull request at OpenAI undergoes a Codex review. 

The post ‘Almost All New Code Written at OpenAI Today is From Codex Users’ appeared first on Analytics India Magazine.

Deloitte’s A$440,000 ‘Human Intelligence Problem’

Australian Sen. Deborah O’Neill at a senate hearing blasted that “Deloitte has a human intelligence problem.”

The post Deloitte’s A$440,000 ‘Human Intelligence Problem’  appeared first on Analytics India Magazine.

Your Native Language Might Just Be Your Next Tech Skill

“Startups that can build culturally and linguistically nuanced AI will have a strong global advantage”

The post Your Native Language Might Just Be Your Next Tech Skill appeared first on Analytics India Magazine.

ChatGPT Enters the Super App Era With Apps SDK

The company also announced a new no-code, drag-and-drop AI agent builder. 

The post ChatGPT Enters the Super App Era With Apps SDK appeared first on Analytics India Magazine.

MLDS 2026 Set to Redefine the Future of Agentic AI for Developers

Bangalore | 19–20 February 2026 | Nimhans Convention Center

The post MLDS 2026 Set to Redefine the Future of Agentic AI for Developers appeared first on Analytics India Magazine.

Leaked Preview Shows OpenAI’s Agent Builder, a Potential n8n Rival

The tool is expected to be launched at OpenAI’s DevDay event on October 6. 

The post Leaked Preview Shows OpenAI’s Agent Builder, a Potential n8n Rival appeared first on Analytics India Magazine.

Discovery of cells that keep immune responses in check wins medicine Nobel Prize

Their work revealed a peripheral mechanism that keeps immune system from causing harm.

Ted Cruz picks a fight with Wikipedia, accusing platform of left-wing bias

Cruz sends letter demanding answers from Wikimedia Foundation.

OpenAI wants to make ChatGPT into a universal app frontend

Spotify, Canva, Zillow among today's launch partners, more coming later this year.

The neurons that let us see what isn’t there

A standard optical illusion triggers specific neurons in the visual system of mice.

Trump’s EPA sued for axing $7 billion solar energy program

"Solar for All" was designed to drop energy bills by $350 million, lawsuit says.

From the telegraph to AI, our communications systems have always had hidden environmental costs

The telegraph was hailed for its revolutionary ability to span distance. Now AI is being hailed as a great leap forward. But both came with environmental costs.

Robin Williams’ daughter Zelda hits out at AI-generated videos of her dead father: ‘stop doing this to him’

Film-maker tells the public to stop sending her videos, saying: ‘You’re not making art, you’re making disgusting, over-processed hotdogs out of the lives of human beings’

Zelda Williams, the daughter of the late actor and comedian Robin Williams, has spoken out against AI-generated content featuring her father.

“Please, just stop sending me AI videos of Dad,” Zelda wrote in an Instagram story on Monday. “Stop believing I wanna see it or that I’ll understand, I don’t and I won’t. If you’re just trying to troll me, I’ve seen way worse, I’ll restrict and move on. But please, if you’ve got any decency, just stop doing this to him and to me, to everyone even, full stop. It’s dumb, it’s a waste of time and energy, and believe me, it’s NOT what he’d want.

Continue reading...

What happens when AI doesn’t do what it’s supposed to? | Fiona Katauskas

Somebody always ends up paying

• See more of Fiona Katauskas’s cartoons here

Continue reading...

Consultants Forced to Pay Money Back After Getting Caught Using AI for Expensive “Report”

"Deloitte has a human intelligence problem."

The post Consultants Forced to Pay Money Back After Getting Caught Using AI for Expensive “Report” appeared first on Futurism.

Stalker Already Using OpenAI’s Sora 2 to Harass Victim

"It is scary to think what AI is doing to feed my stalker's delusions."

The post Stalker Already Using OpenAI’s Sora 2 to Harass Victim appeared first on Futurism.

Taylor Swift Fans Furious as She’s Caught Using Sloppy AI in Video for New Album

"She’s far too rich to be this f***ing cheap."

The post Taylor Swift Fans Furious as She’s Caught Using Sloppy AI in Video for New Album appeared first on Futurism.

Jony Ive Says He Wants His OpenAI Devices to ‘Make Us Happy’

“I don’t think we have an easy relationship with our technology at the moment,” the former Apple designer said at OpenAI's developer conference in San Francisco on Monday.

OpenAI Wants ChatGPT to Be Your Future Operating System

At OpenAI’s Developer Day, CEO Sam Altman showed off apps that run entirely inside the chat window—a new effort to turn ChatGPT into a platform.

OpenAI's Blockbuster AMD Deal Is a Bet on Near-Limitless Demand for AI

OpenAI’s latest move in the race to build massive data centers in the US shows it believes demand for AI will keep surging—even as skeptics warn of a bubble.

WIRED Roundup: The New Fake World of OpenAI’s Social Video App

On this episode of Uncanny Valley, we break down some of the week's best stories, covering everything from Peter Thiel's obsession with the Antichrist to the launch of OpenAI’s new Sora 2 video app.

Jaguar Land Rover to Restart Output in UK After Cyberattack

Jaguar Land Rover Automotive Plc will resume production at some of its UK sites after weeks-long outages following a cyberattack that snarled its global manufacturing network.

Hedge Funds and Crypto Lure Gulf Heirs Set to Inherit $1 Trillion

In 2020, two young brothers at a 135-year-old Gulf merchant family approached one of its money-managers with an idea: bet on Bitcoin.

Telefonica Seeks Leaner, Bolder Structure Under Chief’s Review

An executive shake-up, potential job cuts and a pledge to do more deals all offer glimpses into the Telefonica SA chairman’s ongoing strategic review, set to be presented next month with the ambition of kick starting growth at the Spanish carrier.

LG Electronics India Starts Taking Bids for $1.3 Billion IPO

LG Electronics Inc.’s Indian unit started taking orders for its $1.3 billion initial public offering on Tuesday, joining Tata Capital Ltd. to launch deals in what could be a record month for new listings in the country.

Elon Musk Names Ex-Morgan Stanley Banker CFO of xAI, FT Reports

Elon Musk named a former Morgan Stanley executive the chief financial officer of xAI, the Financial Times reported, appointing an ally to help oversee one of the largest US artificial intelligence startups.

OpenAI Lets ChatGPT Users Connect With Spotify, Zillow In App

OpenAI is making it easier for ChatGPT users to connect with third-party apps within the chatbot to carry out tasks, the company’s latest bid to turn its flagship product into a key gateway for digital services.

Qualtrics Agrees to Buy Press Ganey Forsta in $6.75 Billion Deal

Qualtrics International Inc. agreed to buy health-care survey firm Press Ganey Forsta in a deal valued at $6.75 billion, in the latest private equity-driven acquisition of an enterprise software business.

Supreme Court Denies Google’s Bid to Pause App Store Changes

Alphabet Inc.’s Google lost a US Supreme Court bid to pause major changes to its Google Play app store in an antitrust case filed by Fortnite-maker Epic Games Inc.

AMD's Chip Deal With OpenAI Triggers Explosive Rally | Bloomberg Tech 10/6/2025

AMD CEO Lisa Su and OpenAI President Greg Brockman join Bloomberg Tech to weigh in on the two companies' deal to roll out AI infrastructure, a pact the chipmaker said could generate tens of billions of dollars in new revenue. Plus, White House AI and Crypto Czar David Sacks defends the Trump administration’s approach to China in the context of the global AI race, and Meta CMO Alex Schultz has a new book on how AI is changing the advertising landscape. (Source: Bloomberg)

OpenAI’s Golden Touch Spreads as Stocks Soar

OpenAI has already proved it has a golden touch when it comes to partnering with other tech firms to deploy its artificial intelligence products. Its power to move shares, though, is rapidly spreading to companies it even briefly discusses teaming up with.

AppLovin Probed by SEC Over Its Data-Collection Practices

The Securities and Exchange Commission has been probing the data-collection practices of the mobile advertising tech company AppLovin Corp., according to people familiar with the matter. The company’s shares slid.

Entergy in Talks With Hyperscalers on Big Louisiana Data Centers

Multiple technology companies are seeking to build large data centers in Northern Louisiana after Meta Platforms Inc. announced plans to construct its largest AI-focused facility in the region, according to Entergy Louisiana Chief Executive Officer Phillip May.

David Ellison has a vision for his media empire — whether you like it or not

David Ellison has big plans for Paramount Skydance, backed by his father's billions. Employees and other media insiders are nervous but curious.

Amazon, Meta, Microsoft, and Google are gambling $320 billion on AI infrastructure. The payoff isn't there yet

Big Tech is set to spend $320 billion on capex, mainly for AI infrastructure like data centers. It's a construction boom unlike any in living memory.

Planning a vacation was already a nightmare. Then came AI.

Plane tickets, Airbnbs, rental cars: AI has taken over the travel industry.

Americans are spending like mad. It may not be enough to stop a recession.

Consumers are still spending like mad. But increasing unemployment and a decline in business investment are worrying recession signals.

Athletes on TikTok are freaking out about the Strava-Garmin lawsuit, which hit just before big fall races

Athletes on TikTok are not happy about Strava and Garmin, the "mum and dad" of fitness, clashing in court.

Taylor Swift explains why she won't perform at the Super Bowl: 'I'm just too locked in'

Swift said she did not get an official offer from Jay-Z's team to perform at the Super Bowl, but she also had another reason why she wouldn't do it.

George Clooney says his kids have a 'much better life' living in France than in LA

"But now, for them, it's like — they're not on their iPads, you know?" George Clooney said.

Rivian CEO: Tesla is going all in on cameras, but self-driving cars still need LiDAR

"It's a really great sensor that can do things that cameras can't," Rivian CEO RJ Scaringe said of LiDAR.

Jony Ive says he's juggling up to 20 ideas for OpenAI gadgets

OpenAI acquired Jony Ive's startup earlier this year, sparking speculation about future AI devices the former Apple designer is working on.

Play Store downloads could soon get cheaper after the Supreme Court denies Google's bid to delay antitrust changes

The Supreme Court denied Google's bid to temporarily block parts of a lower court ruling in its legal fight with Epic Games, the maker of Fortnite.

I am an AI scientist at Amazon Web Services. It's not too late to get a Ph.D. in AI.

Girik Malik, 30, says a Ph.D. in AI will become more valuable, not less, as the technology becomes more prevalent.

We retired early and teach a course on financial independence. These are 5 mistakes we keep seeing people make.

Katie and Alan Donegan suggest setting up your index funds and automating your investments — then relax, enjoy life, and stay on track for early retirement.

Bali made sense for business. Moving back to the US was right for our family — for now.

Carey Bentley said the Asia-US time difference made running the company from Bali difficult and prompted them to relocate again.

I lost 60 pounds in my 50s, and I've never felt stronger. Here's how I did it and kept the weight off.

I didn't want my weight holding me back so I transformed my diet and exercise routine after 50. I went from a size 16 to a size 6.

Boeing's 777X was supposed to lead its comeback, but it has been delayed — again. This is why that's such a big deal.

The 777X faces certification in a new era of FAA scrutiny, where Boeing has far less freedom to sign off on its own work.

OpenAI dishes more details on its internal AI software tools, including a sales assistant and customer service agent

The startup revealed more details on its AI-powered internal software tools, including a sales assistant and a customer service agent.

OpenAI just launched its own version of an app store, taking aim at Apple and Google

OpenAI is bringing apps inside ChatGPT and announced a kit for developers to build AI-native apps, CEO Sam Altman announced at its DevDay event.

Sam Altman is spreading OpenAI's money far and wide to keep the AI race alive

OpenAI CEO Sam Altman is thinking long-term about his company's access to the AI chips that are so vital in the AI race.

This $855 million acquisition could be a boon for Trump's lofty Golden Dome project

Firefly Aerospace acquires defense analytics firm SciTec for $855 million for an edge in defense tech and Trump's Golden Dome project.

The coolest building in every US state

From futuristic-looking skyscrapers to kitsch roadside attractions, the architecture across all 50 states is as diverse as its population and history.

Jony Ive Says He Wants His OpenAI Devices to ‘Make Us Happy’

“I don’t think we have an easy relationship with our technology at the moment,” the former Apple designer said at OpenAI's developer conference in San Francisco on Monday.

OpenAI Wants ChatGPT to Be Your Future Operating System

At OpenAI’s Developer Day, CEO Sam Altman showed off apps that run entirely inside the chat window—a new effort to turn ChatGPT into a platform.

OpenAI's Blockbuster AMD Deal Is a Bet on Near-Limitless Demand for AI

OpenAI’s latest move in the race to build massive data centers in the US shows it believes demand for AI will keep surging—even as skeptics warn of a bubble.

WIRED Roundup: The New Fake World of OpenAI’s Social Video App

On this episode of Uncanny Valley, we break down some of the week's best stories, covering everything from Peter Thiel's obsession with the Antichrist to the launch of OpenAI’s new Sora 2 video app.

Startups And The Shutdown: When Your Primary Customer Folds Overnight

For thousands of early-stage companies working on AI, biotech, climate tech and more, the U.S. government shutdown is an existential crisis. It doesn’t matter how brilliant their tech is or how slick their pitch deck looks. What matters is whether they’ve built a company resilient enough to survive the silence until Washington switches back on, writes regular columnist Aron Solomon.

Cerebras Systems Pulls Plug On Its IPO Days After Big Fundraise

Just days after raising $1.1 billion in funding, AI processor developer Cerebras Systems has filed to withdraw its hotly anticipated initial public offering.

Decomposing Attention To Find Context-Sensitive Neurons

arXiv:2510.03315v1 Announce Type: new Abstract: We study transformer language models, analyzing attention heads whose attention patterns are spread out, and whose attention scores depend weakly on content. We argue that the softmax denominators of these heads are stable when the underlying token distribution is fixed. By sampling softmax denominators from a "calibration text", we can combine together the outputs of multiple such stable heads in the first layer of GPT2-Small, approximating their combined output by a linear summary of the surrounding text. This approximation enables a procedure where from the weights alone - and a single calibration text - we can uncover hundreds of first layer neurons that respond to high-level contextual properties of the surrounding text, including neurons that didn't activate on the calibration text.

Graph-S3: Enhancing Agentic textual Graph Retrieval with Synthetic Stepwise Supervision

arXiv:2510.03323v1 Announce Type: new Abstract: A significant portion of real-world data is inherently represented as textual graphs, and integrating these graphs into large language models (LLMs) is promising to enable complex graph-based question answering. However, a key challenge in LLM-based textual graph QA systems lies in graph retrieval, i.e., how to retrieve relevant content from large graphs that is sufficiently informative while remaining compact for the LLM context. Existing retrievers suffer from poor performance since they either rely on shallow embedding similarity or employ interactive retrieving policies that demand excessive data labeling and training cost. To address these issues, we present Graph-$S^3$, an agentic textual graph reasoning framework that employs an LLM-based retriever trained with synthetic stepwise supervision. Instead of rewarding the agent based on the final answers, which may lead to sparse and unstable training signals, we propose to closely evaluate each step of the retriever based on offline-extracted golden subgraphs. Our main techniques include a data synthesis pipeline to extract the golden subgraphs for reward generation and a two-stage training scheme to learn the interactive graph exploration policy based on the synthesized rewards. Based on extensive experiments on three common datasets in comparison with seven strong baselines, our approach achieves an average improvement of 8.1\% in accuracy and 9.7\% in F$_1$ score. The advantage is even higher in more complicated multi-hop reasoning tasks. Our code will be open-sourced.

Implicit Values Embedded in How Humans and LLMs Complete Subjective Everyday Tasks

arXiv:2510.03384v1 Announce Type: new Abstract: Large language models (LLMs) can underpin AI assistants that help users with everyday tasks, such as by making recommendations or performing basic computation. Despite AI assistants' promise, little is known about the implicit values these assistants display while completing subjective everyday tasks. Humans may consider values like environmentalism, charity, and diversity. To what extent do LLMs exhibit these values in completing everyday tasks? How do they compare with humans? We answer these questions by auditing how six popular LLMs complete 30 everyday tasks, comparing LLMs to each other and to 100 human crowdworkers from the US. We find LLMs often do not align with humans, nor with other LLMs, in the implicit values exhibited.

Morpheme Induction for Emergent Language

arXiv:2510.03439v1 Announce Type: new Abstract: We introduce CSAR, an algorithm for inducing morphemes from emergent language corpora of parallel utterances and meanings. It is a greedy algorithm that (1) weights morphemes based on mutual information between forms and meanings, (2) selects the highest-weighted pair, (3) removes it from the corpus, and (4) repeats the process to induce further morphemes (i.e., Count, Select, Ablate, Repeat). The effectiveness of CSAR is first validated on procedurally generated datasets and compared against baselines for related tasks. Second, we validate CSAR's performance on human language data to show that the algorithm makes reasonable predictions in adjacent domains. Finally, we analyze a handful of emergent languages, quantifying linguistic characteristics like degree of synonymy and polysemy.

Omni-Embed-Nemotron: A Unified Multimodal Retrieval Model for Text, Image, Audio, and Video

arXiv:2510.03458v1 Announce Type: new Abstract: We present Omni-Embed-Nemotron, a unified multimodal retrieval embedding model developed to handle the increasing complexity of real-world information needs. While Retrieval-Augmented Generation (RAG) has significantly advanced language models by incorporating external knowledge, existing text-based retrievers rely on clean, structured input and struggle with the visually and semantically rich content found in real-world documents such as PDFs, slides, or videos. Recent work such as ColPali has shown that preserving document layout using image-based representations can improve retrieval quality. Building on this, and inspired by the capabilities of recent multimodal models such as Qwen2.5-Omni, we extend retrieval beyond text and images to also support audio and video modalities. Omni-Embed-Nemotron enables both cross-modal (e.g., text - video) and joint-modal (e.g., text - video+audio) retrieval using a single model. We describe the architecture, training setup, and evaluation results of Omni-Embed-Nemotron, and demonstrate its effectiveness in text, image, and video retrieval.

Searching for the Most Human-like Emergent Language

arXiv:2510.03467v1 Announce Type: new Abstract: In this paper, we design a signalling game-based emergent communication environment to generate state-of-the-art emergent languages in terms of similarity to human language. This is done with hyperparameter optimization, using XferBench as the objective function. XferBench quantifies the statistical similarity of emergent language to human language by measuring its suitability for deep transfer learning to human language. Additionally, we demonstrate the predictive power of entropy on the transfer learning performance of emergent language as well as corroborate previous results on the entropy-minimization properties of emergent communication systems. Finally, we report generalizations regarding what hyperparameters produce more realistic emergent languages, that is, ones which transfer better to human language.

SEER: The Span-based Emotion Evidence Retrieval Benchmark

arXiv:2510.03490v1 Announce Type: new Abstract: We introduce the SEER (Span-based Emotion Evidence Retrieval) Benchmark to test Large Language Models' (LLMs) ability to identify the specific spans of text that express emotion. Unlike traditional emotion recognition tasks that assign a single label to an entire sentence, SEER targets the underexplored task of emotion evidence detection: pinpointing which exact phrases convey emotion. This span-level approach is crucial for applications like empathetic dialogue and clinical support, which need to know how emotion is expressed, not just what the emotion is. SEER includes two tasks: identifying emotion evidence within a single sentence, and identifying evidence across a short passage of five consecutive sentences. It contains new annotations for both emotion and emotion evidence on 1200 real-world sentences. We evaluate 14 open-source LLMs and find that, while some models approach average human performance on single-sentence inputs, their accuracy degrades in longer passages. Our error analysis reveals key failure modes, including overreliance on emotion keywords and false positives in neutral text.

ALHD: A Large-Scale and Multigenre Benchmark Dataset for Arabic LLM-Generated Text Detection

arXiv:2510.03502v1 Announce Type: new Abstract: We introduce ALHD, the first large-scale comprehensive Arabic dataset explicitly designed to distinguish between human- and LLM-generated texts. ALHD spans three genres (news, social media, reviews), covering both MSA and dialectal Arabic, and contains over 400K balanced samples generated by three leading LLMs and originated from multiple human sources, which enables studying generalizability in Arabic LLM-genearted text detection. We provide rigorous preprocessing, rich annotations, and standardized balanced splits to support reproducibility. In addition, we present, analyze and discuss benchmark experiments using our new dataset, in turn identifying gaps and proposing future research directions. Benchmarking across traditional classifiers, BERT-based models, and LLMs (zero-shot and few-shot) demonstrates that fine-tuned BERT models achieve competitive performance, outperforming LLM-based models. Results are however not always consistent, as we observe challenges when generalizing across genres; indeed, models struggle to generalize when they need to deal with unseen patterns in cross-genre settings, and these challenges are particularly prominent when dealing with news articles, where LLM-generated texts resemble human texts in style, which opens up avenues for future research. ALHD establishes a foundation for research related to Arabic LLM-detection and mitigating risks of misinformation, academic dishonesty, and cyber threats.

TS-Reasoner: Aligning Time Series Foundation Models with LLM Reasoning

arXiv:2510.03519v1 Announce Type: new Abstract: Time series reasoning is crucial to decision-making in diverse domains, including finance, energy usage, traffic, weather, and scientific discovery. While existing time series foundation models (TSFMs) can capture low-level dynamic patterns and provide accurate forecasting, further analysis usually requires additional background knowledge and sophisticated reasoning, which are lacking in most TSFMs but can be achieved through large language models (LLMs). On the other hand, without expensive post-training, LLMs often struggle with the numerical understanding of time series data. Although it is intuitive to integrate the two types of models, developing effective training recipes that align the two modalities for reasoning tasks is still an open challenge. To this end, we propose TS-Reasoner that aligns the latent representations of TSFMs with the textual inputs of LLMs for downstream understanding/reasoning tasks. Specifically, we propose a simple yet effective method to curate diverse, synthetic pairs of time series and textual captions for alignment training. We then develop a two-stage training recipe that applies instruction finetuning after the alignment pretraining. Unlike existing works that train an LLM to take time series as inputs, we leverage a pretrained TSFM and freeze it during training. Extensive experiments on several benchmarks demonstrate that TS-Reasoner not only outperforms a wide range of prevailing LLMs, Vision Language Models (VLMs), and Time Series LLMs, but also achieves this with remarkable data efficiency, e.g., using less than half the training data.

Identifying Financial Risk Information Using RAG with a Contrastive Insight

arXiv:2510.03521v1 Announce Type: new Abstract: In specialized domains, humans often compare new problems against similar examples, highlight nuances, and draw conclusions instead of analyzing information in isolation. When applying reasoning in specialized contexts with LLMs on top of a RAG, the pipeline can capture contextually relevant information, but it is not designed to retrieve comparable cases or related problems. While RAG is effective at extracting factual information, its outputs in specialized reasoning tasks often remain generic, reflecting broad facts rather than context-specific insights. In finance, it results in generic risks that are true for the majority of companies. To address this limitation, we propose a peer-aware comparative inference layer on top of RAG. Our contrastive approach outperforms baseline RAG in text generation metrics such as ROUGE and BERTScore in comparison with human-generated equity research and risk.

Sample, Align, Synthesize: Graph-Based Response Synthesis with ConGrs

arXiv:2510.03527v1 Announce Type: new Abstract: Language models can be sampled multiple times to access the distribution underlying their responses, but existing methods cannot efficiently synthesize rich epistemic signals across different long-form responses. We introduce Consensus Graphs (ConGrs), a flexible DAG-based data structure that represents shared information, as well as semantic variation in a set of sampled LM responses to the same prompt. We construct ConGrs using a light-weight lexical sequence alignment algorithm from bioinformatics, supplemented by the targeted usage of a secondary LM judge. Further, we design task-dependent decoding methods to synthesize a single, final response from our ConGr data structure. Our experiments show that synthesizing responses from ConGrs improves factual precision on two biography generation tasks by up to 31% over an average response and reduces reliance on LM judges by more than 80% compared to other methods. We also use ConGrs for three refusal-based tasks requiring abstention on unanswerable queries and find that abstention rate is increased by up to 56%. We apply our approach to the MATH and AIME reasoning tasks and find an improvement over self-verification and majority vote baselines by up to 6 points of accuracy. We show that ConGrs provide a flexible method for capturing variation in LM responses and using the epistemic signals provided by response variation to synthesize more effective responses.

Fine-Tuning on Noisy Instructions: Effects on Generalization and Performance

arXiv:2510.03528v1 Announce Type: new Abstract: Instruction-tuning plays a vital role in enhancing the task-solving abilities of large language models (LLMs), improving their usability in generating helpful responses on various tasks. However, previous work has demonstrated that they are sensitive to minor variations in instruction phrasing. In this paper, we explore whether introducing perturbations in instruction-tuning data can enhance LLMs' resistance against noisy instructions. We focus on how instruction-tuning with perturbations, such as removing stop words or shuffling words, affects LLMs' performance on the original and perturbed versions of widely-used benchmarks (MMLU, BBH, GSM8K). We further assess learning dynamics and potential shifts in model behavior. Surprisingly, our results suggest that instruction-tuning on perturbed instructions can, in some cases, improve downstream performance. These findings highlight the importance of including perturbed instructions in instruction-tuning, which can make LLMs more resilient to noisy user inputs.

TriMediQ: A Triplet-Structured Approach for Interactive Medical Question Answering

arXiv:2510.03536v1 Announce Type: new Abstract: Large Language Models (LLMs) perform strongly in static and single-turn medical Question Answer (QA) benchmarks, yet such settings diverge from the iterative information gathering process required in practical clinical consultations. The MEDIQ framework addresses this mismatch by recasting the diagnosis as an interactive dialogue between a patient and an expert system, but the reliability of LLMs drops dramatically when forced to reason with dialogue logs, where clinical facts appear in sentences without clear links. To bridge this gap, we introduce TriMediQ, a triplet-structured approach that summarises patient responses into triplets and integrates them into a Knowledge Graph (KG), enabling multi-hop reasoning. We introduce a frozen triplet generator that extracts clinically relevant triplets, using prompts designed to ensure factual consistency. In parallel, a trainable projection module, comprising a graph encoder and a projector, captures relational information from the KG to enhance expert reasoning. TriMediQ operates in two steps: (i) the projection module fine-tuning with all LLM weights frozen; and (ii) using the fine-tuned module to guide multi-hop reasoning during inference. We evaluate TriMediQ on two interactive QA benchmarks, showing that it achieves up to 10.4\% improvement in accuracy over five baselines on the iMedQA dataset. These results demonstrate that converting patient responses into structured triplet-based graphs enables more accurate clinical reasoning in multi-turn settings, providing a solution for the deployment of LLM-based medical assistants.

What is a protest anyway? Codebook conceptualization is still a first-order concern in LLM-era classification

arXiv:2510.03541v1 Announce Type: new Abstract: Generative large language models (LLMs) are now used extensively for text classification in computational social science (CSS). In this work, focus on the steps before and after LLM prompting -- conceptualization of concepts to be classified and using LLM predictions in downstream statistical inference -- which we argue have been overlooked in much of LLM-era CSS. We claim LLMs can tempt analysts to skip the conceptualization step, creating conceptualization errors that bias downstream estimates. Using simulations, we show that this conceptualization-induced bias cannot be corrected for solely by increasing LLM accuracy or post-hoc bias correction methods. We conclude by reminding CSS analysts that conceptualization is still a first-order concern in the LLM-era and provide concrete advice on how to pursue low-cost, unbiased, low-variance downstream estimates.

CCD-Bench: Probing Cultural Conflict in Large Language Model Decision-Making

arXiv:2510.03553v1 Announce Type: new Abstract: Although large language models (LLMs) are increasingly implicated in interpersonal and societal decision-making, their ability to navigate explicit conflicts between legitimately different cultural value systems remains largely unexamined. Existing benchmarks predominantly target cultural knowledge (CulturalBench), value prediction (WorldValuesBench), or single-axis bias diagnostics (CDEval); none evaluate how LLMs adjudicate when multiple culturally grounded values directly clash. We address this gap with CCD-Bench, a benchmark that assesses LLM decision-making under cross-cultural value conflict. CCD-Bench comprises 2,182 open-ended dilemmas spanning seven domains, each paired with ten anonymized response options corresponding to the ten GLOBE cultural clusters. These dilemmas are presented using a stratified Latin square to mitigate ordering effects. We evaluate 17 non-reasoning LLMs. Models disproportionately prefer Nordic Europe (mean 20.2 percent) and Germanic Europe (12.4 percent), while options for Eastern Europe and the Middle East and North Africa are underrepresented (5.6 to 5.8 percent). Although 87.9 percent of rationales reference multiple GLOBE dimensions, this pluralism is superficial: models recombine Future Orientation and Performance Orientation, and rarely ground choices in Assertiveness or Gender Egalitarianism (both under 3 percent). Ordering effects are negligible (Cramer's V less than 0.10), and symmetrized KL divergence shows clustering by developer lineage rather than geography. These patterns suggest that current alignment pipelines promote a consensus-oriented worldview that underserves scenarios demanding power negotiation, rights-based reasoning, or gender-aware analysis. CCD-Bench shifts evaluation beyond isolated bias detection toward pluralistic decision making and highlights the need for alignment strategies that substantively engage diverse worldviews.

Reactive Transformer (RxT) -- Stateful Real-Time Processing for Event-Driven Reactive Language Models

arXiv:2510.03561v1 Announce Type: new Abstract: The Transformer architecture has become the de facto standard for Large Language Models (LLMs), demonstrating remarkable capabilities in language understanding and generation. However, its application in conversational AI is fundamentally constrained by its stateless nature and the quadratic computational complexity ($O(L^2)$) with respect to sequence length $L$. Current models emulate memory by reprocessing an ever-expanding conversation history with each turn, leading to prohibitive costs and latency in long dialogues. This paper introduces the Reactive Transformer (RxT), a novel architecture designed to overcome these limitations by shifting from a data-driven to an event-driven paradigm. RxT processes each conversational turn as a discrete event in real-time, maintaining context in an integrated, fixed-size Short-Term Memory (STM) system. The architecture features a distinct operational cycle where a generator-decoder produces a response based on the current query and the previous memory state, after which a memory-encoder and a dedicated Memory Attention network asynchronously update the STM with a representation of the complete interaction. This design fundamentally alters the scaling dynamics, reducing the total user-facing cost of a conversation from quadratic ($O(N^2 \cdot T)$) to linear ($O(N \cdot T)$) with respect to the number of interactions $N$. By decoupling response generation from memory updates, RxT achieves low latency, enabling truly real-time, stateful, and economically viable long-form conversations. We validated our architecture with a series of proof-of-concept experiments on synthetic data, demonstrating superior performance and constant-time inference latency compared to a baseline stateless model of comparable size.

LLM, Reporting In! Medical Information Extraction Across Prompting, Fine-tuning and Post-correction

arXiv:2510.03577v1 Announce Type: new Abstract: This work presents our participation in the EvalLLM 2025 challenge on biomedical Named Entity Recognition (NER) and health event extraction in French (few-shot setting). For NER, we propose three approaches combining large language models (LLMs), annotation guidelines, synthetic data, and post-processing: (1) in-context learning (ICL) with GPT-4.1, incorporating automatic selection of 10 examples and a summary of the annotation guidelines into the prompt, (2) the universal NER system GLiNER, fine-tuned on a synthetic corpus and then verified by an LLM in post-processing, and (3) the open LLM LLaMA-3.1-8B-Instruct, fine-tuned on the same synthetic corpus. Event extraction uses the same ICL strategy with GPT-4.1, reusing the guideline summary in the prompt. Results show GPT-4.1 leads with a macro-F1 of 61.53% for NER and 15.02% for event extraction, highlighting the importance of well-crafted prompting to maximize performance in very low-resource scenarios.

Decoupling Task-Solving and Output Formatting in LLM Generation

arXiv:2510.03595v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly adept at following instructions containing task descriptions to solve complex problems, such as mathematical reasoning and automatic evaluation (LLM-as-a-Judge). However, as prompts grow more complex, models often struggle to adhere to all instructions. This difficulty is especially common when instructive prompts intertwine reasoning directives -- specifying what the model should solve -- with rigid formatting requirements that dictate how the solution must be presented. The entanglement creates competing goals for the model, suggesting that more explicit separation of these two aspects could lead to improved performance. To this front, we introduce Deco-G, a decoding framework that explicitly decouples format adherence from task solving. Deco-G handles format compliance with a separate tractable probabilistic model (TPM), while prompts LLMs with only task instructions. At each decoding step, Deco-G combines next token probabilities from the LLM with the TPM calculated format compliance likelihood to form the output probability. To make this approach both practical and scalable for modern instruction-tuned LLMs, we introduce three key innovations: instruction-aware distillation, a flexible trie-building algorithm, and HMM state pruning for computational efficiency. We demonstrate the effectiveness of Deco-G across a wide range of tasks with diverse format requirements, including mathematical reasoning, LLM-as-a-judge, and event argument extraction. Overall, our approach yields 1.0% to 6.0% relative gain over regular prompting practice with guaranteed format compliance.

Can an LLM Induce a Graph? Investigating Memory Drift and Context Length

arXiv:2510.03611v1 Announce Type: new Abstract: Recently proposed evaluation benchmarks aim to characterize the effective context length and the forgetting tendencies of large language models (LLMs). However, these benchmarks often rely on simplistic 'needle in a haystack' retrieval or continuation tasks that may not accurately reflect the performance of these models in information-dense scenarios. Thus, rather than simple next token prediction, we argue for evaluating these models on more complex reasoning tasks that requires them to induce structured relational knowledge from the text - such as graphs from potentially noisy natural language content. While the input text can be viewed as generated in terms of a graph, its structure is not made explicit and connections must be induced from distributed textual cues, separated by long contexts and interspersed with irrelevant information. Our findings reveal that LLMs begin to exhibit memory drift and contextual forgetting at much shorter effective lengths when tasked with this form of relational reasoning, compared to what existing benchmarks suggest. With these findings, we offer recommendations for the optimal use of popular LLMs for complex reasoning tasks. We further show that even models specialized for reasoning, such as OpenAI o1, remain vulnerable to early memory drift in these settings. These results point to significant limitations in the models' ability to abstract structured knowledge from unstructured input and highlight the need for architectural adaptations to improve long-range reasoning.

Towards Unsupervised Speech Recognition at the Syllable-Level

arXiv:2510.03639v1 Announce Type: new Abstract: Training speech recognizers with unpaired speech and text -- known as unsupervised speech recognition (UASR) -- is a crucial step toward extending ASR to low-resource languages in the long-tail distribution and enabling multimodal learning from non-parallel data. However, existing approaches based on phones often rely on costly resources such as grapheme-to-phoneme converters (G2Ps) and struggle to generalize to languages with ambiguous phoneme boundaries due to training instability. In this paper, we address both challenges by introducing a syllable-level UASR framework based on masked language modeling, which avoids the need for G2P and the instability of GAN-based methods. Our approach achieves up to a 40\% relative reduction in character error rate (CER) on LibriSpeech and generalizes effectively to Mandarin, a language that has remained particularly difficult for prior methods. Code will be released upon acceptance.

UNIDOC-BENCH: A Unified Benchmark for Document-Centric Multimodal RAG

arXiv:2510.03663v1 Announce Type: new Abstract: Multimodal retrieval-augmented generation (MM-RAG) is a key approach for applying large language models (LLMs) and agents to real-world knowledge bases, yet current evaluations are fragmented, focusing on either text or images in isolation or on simplified multimodal setups that fail to capture document-centric multimodal use cases. In this paper, we introduce UniDoc-Bench, the first large-scale, realistic benchmark for MM-RAG built from 70k real-world PDF pages across eight domains. Our pipeline extracts and links evidence from text, tables, and figures, then generates 1,600 multimodal QA pairs spanning factual retrieval, comparison, summarization, and logical reasoning queries. To ensure reliability, 20% of QA pairs are validated by multiple annotators and expert adjudication. UniDoc-Bench supports apples-to-apples comparison across four paradigms: (1) text-only, (2) image-only, (3) multimodal text-image fusion, and (4) multimodal joint retrieval -- under a unified protocol with standardized candidate pools, prompts, and evaluation metrics. Our experiments show that multimodal text-image fusion RAG systems consistently outperform both unimodal and jointly multimodal embedding-based retrieval, indicating that neither text nor images alone are sufficient and that current multimodal embeddings remain inadequate. Beyond benchmarking, our analysis reveals when and how visual context complements textual evidence, uncovers systematic failure modes, and offers actionable guidance for developing more robust MM-RAG pipelines.

Fine-Tuning Large Language Models with QLoRA for Offensive Language Detection in Roman Urdu-English Code-Mixed Text

arXiv:2510.03683v1 Announce Type: new Abstract: The use of derogatory terms in languages that employ code mixing, such as Roman Urdu, presents challenges for Natural Language Processing systems due to unstated grammar, inconsistent spelling, and a scarcity of labeled data. In this work, we propose a QLoRA based fine tuning framework to improve offensive language detection in Roman Urdu-English text. We translated the Roman Urdu-English code mixed dataset into English using Google Translate to leverage English LLMs, while acknowledging that this translation reduces direct engagement with code mixing features. Our focus is on classification performance using English translated low resource inputs. We fine tuned several transformers and large language models, including Meta LLaMA 3 8B, Mistral 7B v0.1, LLaMA 2 7B, ModernBERT, and RoBERTa, with QLoRA for memory efficient adaptation. Models were trained and evaluated on a manually annotated Roman Urdu dataset for offensive vs non offensive content. Of all tested models, the highest F1 score of 91.45 was attained by Meta LLaMA 3 8B, followed by Mistral 7B at 89.66, surpassing traditional transformer baselines. These results demonstrate the efficacy of QLoRA in fine tuning high performing models for low resource environments such as code mixed offensive language detection, and confirm the potential of LLMs for this task. This work advances a scalable approach to Roman Urdu moderation and paves the way for future multilingual offensive detection systems based on LLMs.

MedReflect: Teaching Medical LLMs to Self-Improve via Reflective Correction

arXiv:2510.03687v1 Announce Type: new Abstract: Medical problem solving demands expert knowledge and intricate reasoning. Recent studies of large language models (LLMs) attempt to ease this complexity by introducing external knowledge verification through retrieval-augmented generation or by training on reasoning datasets. However, these approaches suffer from drawbacks such as retrieval overhead and high annotation costs, and they heavily rely on substituted external assistants to reach limited performance in medical field. In this paper, we introduce MedReflect, a generalizable framework designed to inspire LLMs with a physician-like reflective thinking mode. MedReflect generates a single-pass reflection chain that includes initial hypothesis generation, self-questioning, self-answering and decision refinement. This self-verified and self-reflective nature releases large language model's latent capability in medical problem-solving without external retrieval or heavy annotation. We demonstrate that MedReflect enables cost-efficient medical dataset construction: with merely 2,000 randomly sampled training examples and a light fine-tuning, this approach achieves notable absolute accuracy improvements across a series of medical benchmarks while cutting annotation requirements. Our results provide evidence that LLMs can learn to solve specialized medical problems via self-reflection and self-improve, reducing reliance on external supervision and extensive task-specific fine-tuning data.

TreePrompt: Leveraging Hierarchical Few-Shot Example Selection for Improved English-Persian and English-German Translation

arXiv:2510.03748v1 Announce Type: new Abstract: Large Language Models (LLMs) have consistently demonstrated strong performance in machine translation, especially when guided by high-quality prompts. Few-shot prompting is an effective technique to improve translation quality; however, most existing example selection methods focus solely on query-to-example similarity and do not account for the quality of the examples. In this work, we propose TreePrompt, a novel example selection approach that learns LLM preferences to identify high-quality, contextually relevant examples within a tree-structured framework. To further explore the balance between similarity and quality, we combine TreePrompt with K-Nearest Neighbors (K-NN) and Adaptive Few-Shot Prompting (AFSP). Evaluations on two language pairs - English-Persian (MIZAN) and English-German (WMT19) - show that integrating TreePrompt with AFSP or Random selection leads to improved translation performance.

Cross-Lingual Multi-Granularity Framework for Interpretable Parkinson's Disease Diagnosis from Speech

arXiv:2510.03758v1 Announce Type: new Abstract: Parkinson's Disease (PD) affects over 10 million people worldwide, with speech impairments in up to 89% of patients. Current speech-based detection systems analyze entire utterances, potentially overlooking the diagnostic value of specific phonetic elements. We developed a granularity-aware approach for multilingual PD detection using an automated pipeline that extracts time-aligned phonemes, syllables, and words from recordings. Using Italian, Spanish, and English datasets, we implemented a bidirectional LSTM with multi-head attention to compare diagnostic performance across the different granularity levels. Phoneme-level analysis achieved superior performance with AUROC of 93.78% +- 2.34% and accuracy of 92.17% +- 2.43%. This demonstrates enhanced diagnostic capability for cross-linguistic PD detection. Importantly, attention analysis revealed that the most informative speech features align with those used in established clinical protocols: sustained vowels (/a/, /e/, /o/, /i/) at phoneme level, diadochokinetic syllables (/ta/, /pa/, /la/, /ka/) at syllable level, and /pataka/ sequences at word level. Source code will be available at https://github.com/jetliqs/clearpd.

Prompt Balance Matters: Understanding How Imbalanced Few-Shot Learning Affects Multilingual Sense Disambiguation in LLMs

arXiv:2510.03762v1 Announce Type: new Abstract: Recent advances in Large Language Models (LLMs) have significantly reshaped the landscape of Natural Language Processing (NLP). Among the various prompting techniques, few-shot prompting has gained considerable attention for its practicality and effectiveness. This study investigates how few-shot prompting strategies impact the Word Sense Disambiguation (WSD) task, particularly focusing on the biases introduced by imbalanced sample distributions. We use the GLOSSGPT prompting method, an advanced approach for English WSD, to test its effectiveness across five languages: English, German, Spanish, French, and Italian. Our results show that imbalanced few-shot examples can cause incorrect sense predictions in multilingual languages, but this issue does not appear in English. To assess model behavior, we evaluate both the GPT-4o and LLaMA-3.1-70B models and the results highlight the sensitivity of multilingual WSD to sample distribution in few-shot settings, emphasizing the need for balanced and representative prompting strategies.

Rezwan: Leveraging Large Language Models for Comprehensive Hadith Text Processing: A 1.2M Corpus Development

arXiv:2510.03781v1 Announce Type: new Abstract: This paper presents the development of Rezwan, a large-scale AI-assisted Hadith corpus comprising over 1.2M narrations, extracted and structured through a fully automated pipeline. Building on digital repositories such as Maktabat Ahl al-Bayt, the pipeline employs Large Language Models (LLMs) for segmentation, chain--text separation, validation, and multi-layer enrichment. Each narration is enhanced with machine translation into twelve languages, intelligent diacritization, abstractive summarization, thematic tagging, and cross-text semantic analysis. This multi-step process transforms raw text into a richly annotated research-ready infrastructure for digital humanities and Islamic studies. A rigorous evaluation was conducted on 1,213 randomly sampled narrations, assessed by six domain experts. Results show near-human accuracy in structured tasks such as chain--text separation (9.33/10) and summarization (9.33/10), while highlighting ongoing challenges in diacritization and semantic similarity detection. Comparative analysis against the manually curated Noor Corpus demonstrates the superiority of Najm in both scale and quality, with a mean overall score of 8.46/10 versus 3.66/10. Furthermore, cost analysis confirms the economic feasibility of the AI approach: tasks requiring over 229,000 hours of expert labor were completed within months at a fraction of the cost. The work introduces a new paradigm in religious text processing by showing how AI can augment human expertise, enabling large-scale, multilingual, and semantically enriched access to Islamic heritage.

Mechanistic Interpretability of Socio-Political Frames in Language Models

arXiv:2510.03799v1 Announce Type: new Abstract: This paper explores the ability of large language models to generate and recognize deep cognitive frames, particularly in socio-political contexts. We demonstrate that LLMs are highly fluent in generating texts that evoke specific frames and can recognize these frames in zero-shot settings. Inspired by mechanistic interpretability research, we investigate the location of the strict father' andnurturing parent' frames within the model's hidden representation, identifying singular dimensions that correlate strongly with their presence. Our findings contribute to understanding how LLMs capture and express meaningful human concepts.

Beyond Token Length: Step Pruner for Efficient and Accurate Reasoning in Large Language Models

arXiv:2510.03805v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) demonstrate strong performance on complex tasks but often suffer from excessive verbosity, known as "overthinking." Existing solutions via reinforcement learning (RL) typically penalize generated tokens to promote conciseness. However, these methods encounter two challenges: responses with fewer tokens do not always correspond to fewer reasoning steps, and models may develop hacking behavior in later stages of training by discarding reasoning steps to minimize token usage. In this work, we introduce \textbf{Step Pruner (SP)}, an RL framework that steers LRMs toward more efficient reasoning by favoring compact reasoning steps. Our step-aware reward function prioritizes correctness while imposing penalties for redundant steps, and withholds rewards for incorrect responses to prevent the reinforcement of erroneous reasoning. Moreover, we propose a dynamic stopping mechanism: when the length of any output step exceeds the upper limit, we halt updates to prevent hacking behavior caused by merging steps. Extensive experiments across four reasoning benchmarks demonstrate that SP achieves state-of-the-art accuracy while significantly reducing response length. For instance, on AIME24, SP reduces token usage by \textbf{69.7\%}.

Annotate Rhetorical Relations with INCEpTION: A Comparison with Automatic Approaches

arXiv:2510.03808v1 Announce Type: new Abstract: This research explores the annotation of rhetorical relations in discourse using the INCEpTION tool and compares manual annotation with automatic approaches based on large language models. The study focuses on sports reports (specifically cricket news) and evaluates the performance of BERT, DistilBERT, and Logistic Regression models in classifying rhetorical relations such as elaboration, contrast, background, and cause-effect. The results show that DistilBERT achieved the highest accuracy, highlighting its potential for efficient discourse relation prediction. This work contributes to the growing intersection of discourse parsing and transformer-based NLP. (This paper was conducted as part of an academic requirement under the supervision of Prof. Dr. Ralf Klabunde, Linguistic Data Science Lab, Ruhr University Bochum.) Keywords: Rhetorical Structure Theory, INCEpTION, BERT, DistilBERT, Discourse Parsing, NLP.

Read Between the Lines: A Benchmark for Uncovering Political Bias in Bangla News Articles

arXiv:2510.03898v1 Announce Type: new Abstract: Detecting media bias is crucial, specifically in the South Asian region. Despite this, annotated datasets and computational studies for Bangla political bias research remain scarce. Crucially because, political stance detection in Bangla news requires understanding of linguistic cues, cultural context, subtle biases, rhetorical strategies, code-switching, implicit sentiment, and socio-political background. To address this, we introduce the first benchmark dataset of 200 politically significant and highly debated Bangla news articles, labeled for government-leaning, government-critique, and neutral stances, alongside diagnostic analyses for evaluating large language models (LLMs). Our comprehensive evaluation of 28 proprietary and open-source LLMs shows strong performance in detecting government-critique content (F1 up to 0.83) but substantial difficulty with neutral articles (F1 as low as 0.00). Models also tend to over-predict government-leaning stances, often misinterpreting ambiguous narratives. This dataset and its associated diagnostics provide a foundation for advancing stance detection in Bangla media research and offer insights for improving LLM performance in low-resource languages.

PsycholexTherapy: Simulating Reasoning in Psychotherapy with Small Language Models in Persian

arXiv:2510.03913v1 Announce Type: new Abstract: This study presents PsychoLexTherapy, a framework for simulating psychotherapeutic reasoning in Persian using small language models (SLMs). The framework tackles the challenge of developing culturally grounded, therapeutically coherent dialogue systems with structured memory for multi-turn interactions in underrepresented languages. To ensure privacy and feasibility, PsychoLexTherapy is optimized for on-device deployment, enabling use without external servers. Development followed a three-stage process: (i) assessing SLMs psychological knowledge with PsychoLexEval; (ii) designing and implementing the reasoning-oriented PsychoLexTherapy framework; and (iii) constructing two evaluation datasets-PsychoLexQuery (real Persian user questions) and PsychoLexDialogue (hybrid simulated sessions)-to benchmark against multiple baselines. Experiments compared simple prompting, multi-agent debate, and structured therapeutic reasoning paths. Results showed that deliberate model selection balanced accuracy, efficiency, and privacy. On PsychoLexQuery, PsychoLexTherapy outperformed all baselines in automatic LLM-as-a-judge evaluation and was ranked highest by human evaluators in a single-turn preference study. In multi-turn tests with PsychoLexDialogue, the long-term memory module proved essential: while naive history concatenation caused incoherence and information loss, the full framework achieved the highest ratings in empathy, coherence, cultural fit, and personalization. Overall, PsychoLexTherapy establishes a practical, privacy-preserving, and culturally aligned foundation for Persian psychotherapy simulation, contributing novel datasets, a reproducible evaluation pipeline, and empirical insights into structured memory for therapeutic reasoning.

Mapping Patient-Perceived Physician Traits from Nationwide Online Reviews with LLMs

arXiv:2510.03997v1 Announce Type: new Abstract: Understanding how patients perceive their physicians is essential to improving trust, communication, and satisfaction. We present a large language model (LLM)-based pipeline that infers Big Five personality traits and five patient-oriented subjective judgments. The analysis encompasses 4.1 million patient reviews of 226,999 U.S. physicians from an initial pool of one million. We validate the method through multi-model comparison and human expert benchmarking, achieving strong agreement between human and LLM assessments (correlation coefficients 0.72-0.89) and external validity through correlations with patient satisfaction (r = 0.41-0.81, all p<0.001). National-scale analysis reveals systematic patterns: male physicians receive higher ratings across all traits, with largest disparities in clinical competence perceptions; empathy-related traits predominate in pediatrics and psychiatry; and all traits positively predict overall satisfaction. Cluster analysis identifies four distinct physician archetypes, from "Well-Rounded Excellent" (33.8%, uniformly high traits) to "Underperforming" (22.6%, consistently low). These findings demonstrate that automated trait extraction from patient narratives can provide interpretable, validated metrics for understanding physician-patient relationships at scale, with implications for quality measurement, bias detection, and workforce development in healthcare.

Simulating and Understanding Deceptive Behaviors in Long-Horizon Interactions

arXiv:2510.03999v1 Announce Type: new Abstract: Deception is a pervasive feature of human communication and an emerging concern in large language models (LLMs). While recent studies document instances of LLM deception under pressure, most evaluations remain confined to single-turn prompts and fail to capture the long-horizon interactions in which deceptive strategies typically unfold. We introduce the first simulation framework for probing and evaluating deception in LLMs under extended sequences of interdependent tasks and dynamic contextual pressures. Our framework instantiates a multi-agent system: a performer agent tasked with completing tasks and a supervisor agent that evaluates progress, provides feedback, and maintains evolving states of trust. An independent deception auditor then reviews full trajectories to identify when and how deception occurs. We conduct extensive experiments across 11 frontier models, spanning both closed- and open-source systems, and find that deception is model-dependent, increases with event pressure, and consistently erodes supervisor trust. Qualitative analyses further reveal distinct strategies of concealment, equivocation, and falsification. Our findings establish deception as an emergent risk in long-horizon interactions and provide a foundation for evaluating future LLMs in real-world, trust-sensitive contexts.

Named Entity Recognition in COVID-19 tweets with Entity Knowledge Augmentation

arXiv:2510.04001v1 Announce Type: new Abstract: The COVID-19 pandemic causes severe social and economic disruption around the world, raising various subjects that are discussed over social media. Identifying pandemic-related named entities as expressed on social media is fundamental and important to understand the discussions about the pandemic. However, there is limited work on named entity recognition on this topic due to the following challenges: 1) COVID-19 texts in social media are informal and their annotations are rare and insufficient to train a robust recognition model, and 2) named entity recognition in COVID-19 requires extensive domain-specific knowledge. To address these issues, we propose a novel entity knowledge augmentation approach for COVID-19, which can also be applied in general biomedical named entity recognition in both informal text format and formal text format. Experiments carried out on the COVID-19 tweets dataset and PubMed dataset show that our proposed entity knowledge augmentation improves NER performance in both fully-supervised and few-shot settings. Our source code is publicly available: https://github.com/kkkenshi/LLM-EKA/tree/master

AgriGPT-VL: Agricultural Vision-Language Understanding Suite

arXiv:2510.04002v1 Announce Type: new Abstract: Despite rapid advances in multimodal large language models, agricultural applications remain constrained by the scarcity of domain-tailored models, curated vision-language corpora, and rigorous evaluation. To address these challenges, we present the AgriGPT-VL Suite, a unified multimodal framework for agriculture. Our contributions are threefold. First, we introduce Agri-3M-VL, the largest vision-language corpus for agriculture to our knowledge, curated by a scalable multi-agent data generator; it comprises 1M image-caption pairs, 2M image-grounded VQA pairs, 50K expert-level VQA instances, and 15K GRPO reinforcement learning samples. Second, we develop AgriGPT-VL, an agriculture-specialized vision-language model trained via a progressive curriculum of textual grounding, multimodal shallow/deep alignment, and GRPO refinement. This method achieves strong multimodal reasoning while preserving text-only capability. Third, we establish AgriBench-VL-4K, a compact yet challenging evaluation suite with open-ended and image-grounded questions, paired with multi-metric evaluation and an LLM-as-a-judge framework. Experiments show that AgriGPT-VL outperforms leading general-purpose VLMs on AgriBench-VL-4K, achieving higher pairwise win rates in the LLM-as-a-judge evaluation. Meanwhile, it remains competitive on the text-only AgriBench-13K with no noticeable degradation of language ability. Ablation studies further confirm consistent gains from our alignment and GRPO refinement stages. We will open source all of the resources to support reproducible research and deployment in low-resource agricultural settings.

LLM Microscope: What Model Internals Reveal About Answer Correctness and Context Utilization

arXiv:2510.04013v1 Announce Type: new Abstract: Although large language models (LLMs) have tremendous utility, trustworthiness is still a chief concern: models often generate incorrect information with high confidence. While contextual information can help guide generation, identifying when a query would benefit from retrieved context and assessing the effectiveness of that context remains challenging. In this work, we operationalize interpretability methods to ascertain whether we can predict the correctness of model outputs from the model's activations alone. We also explore whether model internals contain signals about the efficacy of external context. We consider correct, incorrect, and irrelevant context and introduce metrics to distinguish amongst them. Experiments on six different models reveal that a simple classifier trained on intermediate layer activations of the first output token can predict output correctness with about 75% accuracy, enabling early auditing. Our model-internals-based metric significantly outperforms prompting baselines at distinguishing between correct and incorrect context, guarding against inaccuracies introduced by polluted context. These findings offer a lens to better understand the underlying decision-making processes of LLMs. Our code is publicly available at https://github.com/jiarui-liu/LLM-Microscope

Thai Semantic End-of-Turn Detection for Real-Time Voice Agents

arXiv:2510.04016v1 Announce Type: new Abstract: Fluid voice-to-voice interaction requires reliable and low-latency detection of when a user has finished speaking. Traditional audio-silence end-pointers add hundreds of milliseconds of delay and fail under hesitations or language-specific phenomena. We present, to our knowledge, the first systematic study of Thai text-only end-of-turn (EOT) detection for real-time agents. We compare zero-shot and few-shot prompting of compact LLMs to supervised fine-tuning of lightweight transformers. Using transcribed subtitles from the YODAS corpus and Thai-specific linguistic cues (e.g., sentence-final particles), we formulate EOT as a binary decision over token boundaries. We report a clear accuracy-latency tradeoff and provide a public-ready implementation plan. This work establishes a Thai baseline and demonstrates that small, fine-tuned models can deliver near-instant EOT decisions suitable for on-device agents.

Does Using Counterfactual Help LLMs Explain Textual Importance in Classification?

arXiv:2510.04031v1 Announce Type: new Abstract: Large language models (LLMs) are becoming useful in many domains due to their impressive abilities that arise from large training datasets and large model sizes. More recently, they have been shown to be very effective in textual classification tasks, motivating the need to explain the LLMs' decisions. Motivated by practical constrains where LLMs are black-boxed and LLM calls are expensive, we study how incorporating counterfactuals into LLM reasoning can affect the LLM's ability to identify the top words that have contributed to its classification decision. To this end, we introduce a framework called the decision changing rate that helps us quantify the importance of the top words in classification. Our experimental results show that using counterfactuals can be helpful.

Small Language Models for Emergency Departments Decision Support: A Benchmark Study

arXiv:2510.04032v1 Announce Type: new Abstract: Large language models (LLMs) have become increasingly popular in medical domains to assist physicians with a variety of clinical and operational tasks. Given the fast-paced and high-stakes environment of emergency departments (EDs), small language models (SLMs), characterized by a reduction in parameter count compared to LLMs, offer significant potential due to their inherent reasoning capability and efficient performance. This enables SLMs to support physicians by providing timely and accurate information synthesis, thereby improving clinical decision-making and workflow efficiency. In this paper, we present a comprehensive benchmark designed to identify SLMs suited for ED decision support, taking into account both specialized medical expertise and broad general problem-solving capabilities. In our evaluations, we focus on SLMs that have been trained on a mixture of general-domain and medical corpora. A key motivation for emphasizing SLMs is the practical hardware limitations, operational cost constraints, and privacy concerns in the typical real-world deployments. Our benchmark datasets include MedMCQA, MedQA-4Options, and PubMedQA, with the medical abstracts dataset emulating tasks aligned with real ED physicians' daily tasks. Experimental results reveal that general-domain SLMs surprisingly outperform their medically fine-tuned counterparts across these diverse benchmarks for ED. This indicates that for ED, specialized medical fine-tuning of the model may not be required.

Exploring Chain-of-Thought Reasoning for Steerable Pluralistic Alignment

arXiv:2510.04045v1 Announce Type: new Abstract: Large Language Models (LLMs) are typically trained to reflect a relatively uniform set of values, which limits their applicability to tasks that require understanding of nuanced human perspectives. Recent research has underscored the importance of enabling LLMs to support steerable pluralism -- the capacity to adopt a specific perspective and align generated outputs with it. In this work, we investigate whether Chain-of-Thought (CoT) reasoning techniques can be applied to building steerable pluralistic models. We explore several methods, including CoT prompting, fine-tuning on human-authored CoT, fine-tuning on synthetic explanations, and Reinforcement Learning with Verifiable Rewards (RLVR). We evaluate these approaches using the Value Kaleidoscope and OpinionQA datasets. Among the methods studied, RLVR consistently outperforms others and demonstrates strong training sample efficiency. We further analyze the generated CoT traces with respect to faithfulness and safety.

What Makes Diffusion Language Models Super Data Learners?

arXiv:2510.04071v1 Announce Type: new Abstract: Recent studies have shown that diffusion language models achieve remarkable data efficiency under limited-data constraints, yet the underlying mechanisms remain unclear. In this work, we perform extensive ablation experiments to disentangle the sources of this efficiency. Our results show that random masking of input tokens plays the dominant role. We further show that similar gains can be obtained through in MLP dropout and weight decay, indicating that stochastic regularization broadly enhances data efficiency in multi-epoch training. Our code is available at https://github.com/zitian-gao/data-efficiency.

PoLi-RL: A Point-to-List Reinforcement Learning Framework for Conditional Semantic Textual Similarity

arXiv:2510.04080v1 Announce Type: new Abstract: Conditional Semantic Textual Similarity (C-STS) measures the semantic proximity between text segments under a specific condition, thereby overcoming the ambiguity inherent in traditional STS. However, existing methods are largely confined to discriminative models, failing to fully integrate recent breakthroughs in the NLP community concerning Large Language Models (LLMs) and Reinforcement Learning (RL). RL is a particularly well-suited paradigm for this task, as it can directly optimize the non-differentiable Spearman ranking metric and guide the reasoning process required by C-STS. However, we find that naively applying listwise RL fails to produce meaningful improvements, as the model is overwhelmed by complex, coarse-grained reward signals. To address this challenge, we introduce PoLi-RL, a novel Point-to-List Reinforcement Learning framework. PoLi-RL employs a two-stage curriculum: it first trains the model with simple pointwise rewards to establish fundamental scoring capabilities, then transitions to a hybrid reward that combines pointwise, pairwise, and listwise objectives to refine the model's ability to discern subtle semantic distinctions. Crucially, we propose an innovative Parallel Slice Ranking Reward (PSRR) mechanism that computes ranking rewards in parallel slices, where each slice comprises same-indexed completions from different samples. This provides a precise, differentiated learning signal for each individual completion, enabling granular credit assignment and effective optimization. On the official C-STS benchmark, PoLi-RL achieves a Spearman correlation coefficient of 48.18, establishing a new SOTA for the cross-encoder architecture. As the first work to successfully apply RL to C-STS, our study introduces a powerful and precise paradigm for training LLMs on complex, ranking-based conditional judgment tasks.

Scaling Code-Assisted Chain-of-Thoughts and Instructions for Model Reasoning

arXiv:2510.04081v1 Announce Type: new Abstract: Reasoning capability is pivotal for Large Language Models (LLMs) to solve complex tasks, yet achieving reliable and scalable reasoning remains challenging. While Chain-of-Thought (CoT) prompting has become a mainstream approach, existing methods often suffer from uncontrolled generation, insufficient quality, and limited diversity in reasoning paths. Recent efforts leverage code to enhance CoT by grounding reasoning in executable steps, but such methods are typically constrained to predefined mathematical problems, hindering scalability and generalizability. In this work, we propose Caco (Code-Assisted Chain-of-ThOught), a novel framework that automates the synthesis of high-quality, verifiable, and diverse instruction-CoT reasoning data through code-driven augmentation. Unlike prior work, Caco first fine-tunes a code-based CoT generator on existing math and programming solutions in a unified code format, then scales the data generation to a large amount of diverse reasoning traces. Crucially, we introduce automated validation via code execution and rule-based filtering to ensure logical correctness and structural diversity, followed by reverse-engineering filtered outputs into natural language instructions and language CoTs to enrich task adaptability. This closed-loop process enables fully automated, scalable synthesis of reasoning data with guaranteed executability. Experiments on our created Caco-1.3M dataset demonstrate that Caco-trained models achieve strong competitive performance on mathematical reasoning benchmarks, outperforming existing strong baselines. Further analysis reveals that Caco's code-anchored verification and instruction diversity contribute to superior generalization across unseen tasks. Our work establishes a paradigm for building self-sustaining, trustworthy reasoning systems without human intervention.

Unveiling LLMs' Metaphorical Understanding: Exploring Conceptual Irrelevance, Context Leveraging and Syntactic Influence

arXiv:2510.04120v1 Announce Type: new Abstract: Metaphor analysis is a complex linguistic phenomenon shaped by context and external factors. While Large Language Models (LLMs) demonstrate advanced capabilities in knowledge integration, contextual reasoning, and creative generation, their mechanisms for metaphor comprehension remain insufficiently explored. This study examines LLMs' metaphor-processing abilities from three perspectives: (1) Concept Mapping: using embedding space projections to evaluate how LLMs map concepts in target domains (e.g., misinterpreting "fall in love" as "drop down from love"); (2) Metaphor-Literal Repository: analyzing metaphorical words and their literal counterparts to identify inherent metaphorical knowledge; and (3) Syntactic Sensitivity: assessing how metaphorical syntactic structures influence LLMs' performance. Our findings reveal that LLMs generate 15\%-25\% conceptually irrelevant interpretations, depend on metaphorical indicators in training data rather than contextual cues, and are more sensitive to syntactic irregularities than to structural comprehension. These insights underline the limitations of LLMs in metaphor analysis and call for more robust computational approaches.

Sri Lanka Document Datasets: A Large-Scale, Multilingual Resource for Law, News, and Policy (v20251005)

arXiv:2510.04124v1 Announce Type: new Abstract: We present a collection of open, machine-readable document datasets covering parliamentary proceedings, legal judgments, government publications, news, and tourism statistics from Sri Lanka. As of v20251005, the collection currently comprises 215,670 documents (60.3 GB) across 13 datasets in Sinhala, Tamil, and English. The datasets are updated daily and mirrored on GitHub and Hugging Face. These resources aim to support research in computational linguistics, legal analytics, socio-political studies, and multilingual natural language processing. We describe the data sources, collection pipeline, formats, and potential use cases, while discussing licensing and ethical considerations.

Fine Tuning Methods for Low-resource Languages

arXiv:2510.04139v1 Announce Type: new Abstract: The rise of Large Language Models has not been inclusive of all cultures. The models are mostly trained on English texts and culture which makes them underperform in other languages and cultural contexts. By developing a generalizable method for preparing culturally relevant datasets and post-training the Gemma 2 model, this project aimed to increase the performance of Gemma 2 for an underrepresented language and showcase how others can do the same to unlock the power of Generative AI in their country and preserve their cultural heritage.

Self Speculative Decoding for Diffusion Large Language Models

arXiv:2510.04147v1 Announce Type: new Abstract: Diffusion-based Large Language Models (dLLMs) have emerged as a competitive alternative to autoregressive models, offering unique advantages through bidirectional attention and parallel generation paradigms. However, the generation results of current parallel decoding methods deviate from stepwise decoding, introducing potential performance degradation, which limits their practical deployment. To address this problem, we propose \textbf{S}elf \textbf{S}peculative \textbf{D}ecoding (SSD), a lossless inference acceleration method that leverages the dLLM itself as both speculative decoding drafter and verifier without auxiliary modules. SSD introduces a self-drafting mechanism where the model generates predictions for multiple positions, then verifies them through hierarchical verification trees in a single forward pass. Unlike traditional speculative decoding that requires separate draft models, SSD eliminates model redundancy and memory overhead by exploiting the dLLM's inherent parallel prediction capability for multiple positions. This self-speculative approach allows the model to progressively verify and accept multiple tokens in a single forward pass. Our experiments demonstrate that SSD achieves up to 3.46$\times$ speedup while keeping the output identical to stepwise decoding on open source models such as LLaDA and Dream. Code will be made publicly available on GitHub.

Thinking on the Fly: Test-Time Reasoning Enhancement via Latent Thought Policy Optimization

arXiv:2510.04182v1 Announce Type: new Abstract: Recent advancements in Large Language Models (LLMs) have shifted from explicit Chain-of-Thought (CoT) reasoning to more efficient latent reasoning, where intermediate thoughts are represented as vectors rather than text. However, latent reasoning can be brittle on challenging, out-of-distribution tasks where robust reasoning is most critical. To overcome these limitations, we introduce Latent Thought Policy Optimization (LTPO), a parameter-free framework that enhances LLM reasoning entirely at test time, without requiring model parameter updates. LTPO treats intermediate latent "thought" vectors as dynamic parameters that are actively optimized for each problem instance. It employs an online policy gradient method guided by an intrinsic, confidence-based reward signal computed directly from the frozen LLM's own output distributions, eliminating the need for external supervision or expensive text generation during optimization. Extensive experiments on five reasoning benchmarks show that LTPO not only matches or surpasses strong baselines on standard tasks but also demonstrates remarkable robustness where others fail. Most notably, on highly challenging AIME benchmarks where existing latent reasoning baselines collapse to near-zero accuracy, LTPO delivers substantial improvements, showcasing a unique capability for complex reasoning.

CALM Before the STORM: Unlocking Native Reasoning for Optimization Modeling

arXiv:2510.04204v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) have demonstrated strong capabilities in complex multi-step reasoning, opening new opportunities for automating optimization modeling. However, existing domain adaptation methods, originally designed for earlier instruction-tuned models, often fail to exploit the advanced reasoning patterns of modern LRMs -- In particular, we show that direct fine-tuning on traditional \textit{non-reflective} datasets leads to limited gains. To fully leverage LRMs' inherent reasoning abilities, we propose \textbf{CALM} (\textit{Corrective Adaptation with Lightweight Modification}), a framework that progressively refines LRMs within their native reasoning modes for optimization modeling tasks. In CALM, an expert intervener identifies reasoning flaws and provides concise corrective hints, which the LRM incorporates to produce improved reasoning trajectories. These interventions modify fewer than 2.6\% of generated tokens, but generate high-quality data for soft adaptation through supervised fine-tuning. The adapted model is then further improved through reinforcement learning. Building on CALM, we develop \textbf{STORM} (\textit{Smart Thinking Optimization Reasoning Model}), a 4B-parameter LRM that achieves a new state-of-the-art average accuracy of 68.9\% across five popular optimization modeling benchmarks, matching the performance of a 671B LRM. These results demonstrate that dynamic, hint-based data synthesis both preserves and amplifies the native reasoning patterns of modern LRMs, offering a more effective and scalable path towards expert-level performance on challenging optimization modeling tasks.

Teaching LLM to be Persuasive: Reward-Enhanced Policy Optimization for Alignment frm Heterogeneous Rewards

arXiv:2510.04214v1 Announce Type: new Abstract: We study deploying large language models (LLMs) as business development (BD) agents for persuasive price negotiation in online travel agencies (OTAs), where aligning traveler affordability and hotel profitability directly affects bookings, partner relationships, and access to travel. The agent must follow a Standard Operating Procedure (SOP) while conducting multi-turn persuasion, interpreting colloquial inputs, and adhering to guardrails (no over-promising, no hallucinations). Conventional post-training -- supervised fine-tuning (SFT) or single-source reward optimization -- overfits scripts, misses nuanced persuasive style, and fails to enforce verifiable business constraints. We propose Reward-Enhanced Policy Optimization (REPO), a reinforcement learning post-training framework that aligns an LLM with heterogeneous rewards: a preference-trained reward model (RM) for dense human alignment, a reward judge (RJ) for high-level persuasive behavior and SOP compliance, and programmatic reward functions (RF) for deterministic checks on numerics, formatting, and guardrails. A straightforward enhancement mechanism is proposed to combine the RM with RJ and RF signals to curb reward hacking and improve negotiation quality. In production-style evaluations -- approximately 150 turns from real dialogues and 225 turns from curated bad-case dialogues -- REPO lifts average dialogue rating to 4.63: +1.20 over base, +0.83 over Direct Preference Optimization (DPO); +0.33 over Group Relative Policy Optimization (GRPO), increases the share of conversations with at least one excellent response to 66.67% (+23.34 percentage points over GRPO), and achieves a 93.33% bad-case fix rate with 75.56% clean fixes, outperforming SFT, DPO, PPO, and GRPO. We also observe emergent capabilities -- proactive empathy, localized reasoning, calibrated tactics -- that surpass gold annotations.

Epistemic Diversity and Knowledge Collapse in Large Language Models

arXiv:2510.04226v1 Announce Type: new Abstract: Large language models (LLMs) tend to generate lexically, semantically, and stylistically homogenous texts. This poses a risk of knowledge collapse, where homogenous LLMs mediate a shrinking in the range of accessible information over time. Existing works on homogenization are limited by a focus on closed-ended multiple-choice setups or fuzzy semantic features, and do not look at trends across time and cultural contexts. To overcome this, we present a new methodology to measure epistemic diversity, i.e., variation in real-world claims in LLM outputs, which we use to perform a broad empirical study of LLM knowledge collapse. We test 27 LLMs, 155 topics covering 12 countries, and 200 prompt variations sourced from real user chats. For the topics in our study, we show that while newer models tend to generate more diverse claims, nearly all models are less epistemically diverse than a basic web search. We find that model size has a negative impact on epistemic diversity, while retrieval-augmented generation (RAG) has a positive impact, though the improvement from RAG varies by the cultural context. Finally, compared to a traditional knowledge source (Wikipedia), we find that country-specific claims reflect the English language more than the local one, highlighting a gap in epistemic representation

Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought

arXiv:2510.04230v1 Announce Type: new Abstract: Recent frontier models employ long chain-of-thought reasoning to explore solution spaces in context and achieve stonger performance. While many works study distillation to build smaller yet capable models, most focus on English and little is known about language-specific reasoning. To bridge this gap, we first introduct Language-Mixed CoT, a reasoning schema that switches between English and a target language, using English as an anchor to excel in reasoning while minimizing translation artificats. As a Korean case study, we curate Yi-Sang: 5.79M native-Korean prompts from web Q&A, exams, STEM, and code; 3.7M long reasoning traces generated from Qwen3-32B; and a targeted 260k high-yield subset. We train ninve models (4B-35B) across six families (Qwen2.5, Llama-3.1, Gemma-3, etc). Our best model, KO-REAson-35B, achieves state-of-the-art performance, with the highest overall average score (64.0 \pm 25), ranking first on 5/9 benchmarks and second on the remainder. Samller and mid-sized models also benefit substantially, with an average improvement of +18.6 points across teh evaluated nine benchmarks. Ablations show Language-Mixed CoT is more effective than monolingual CoT, also resulting in cross-lingual and mult-modal performance gains. We release our data-curation pipeline, evaluation system, datasets, and models to advance research on language-specific reasoning. Data and model collection: https://huggingface.co/KOREAson.

LongTail-Swap: benchmarking language models' abilities on rare words

arXiv:2510.04268v1 Announce Type: new Abstract: Children learn to speak with a low amount of data and can be taught new words on a few-shot basis, making them particularly data-efficient learners. The BabyLM challenge aims at exploring language model (LM) training in the low-data regime but uses metrics that concentrate on the head of the word distribution. Here, we introduce LongTail-Swap (LT-Swap), a benchmark that focuses on the tail of the distribution, i.e., measures the ability of LMs to learn new words with very little exposure, like infants do. LT-Swap is a pretraining corpus-specific test set of acceptable versus unacceptable sentence pairs that isolate semantic and syntactic usage of rare words. Models are evaluated in a zero-shot fashion by computing the average log probabilities over the two members of each pair. We built two such test sets associated with the 10M words and 100M words BabyLM training sets, respectively, and evaluated 16 models from the BabyLM leaderboard. Our results not only highlight the poor performance of language models on rare words but also reveal that performance differences across LM architectures are much more pronounced in the long tail than in the head. This offers new insights into which architectures are better at handling rare word generalization. We've also made the code publicly avail

Probing Geometry of Next Token Prediction Using Cumulant Expansion of the Softmax Entropy

arXiv:2510.04285v1 Announce Type: new Abstract: We introduce a cumulant-expansion framework for quantifying how large language models (LLMs) internalize higher-order statistical structure during next-token prediction. By treating the softmax entropy of each layer's logit distribution as a perturbation around its "center" distribution, we derive closed-form cumulant observables that isolate successively higher-order correlations. Empirically, we track these cumulants in GPT-2 and Pythia models on Pile-10K prompts. (i) Structured prompts exhibit a characteristic rise-and-plateau profile across layers, whereas token-shuffled prompts remain flat, revealing the dependence of the cumulant profile on meaningful context. (ii) During training, all cumulants increase monotonically before saturating, directly visualizing the model's progression from capturing variance to learning skew, kurtosis, and higher-order statistical structures. (iii) Mathematical prompts show distinct cumulant signatures compared to general text, quantifying how models employ fundamentally different processing mechanisms for mathematical versus linguistic content. Together, these results establish cumulant analysis as a lightweight, mathematically grounded probe of feature-learning dynamics in high-dimensional neural networks.

SliceMoE: Routing Embedding Slices Instead of Tokens for Fine-Grained and Balanced Transformer Scaling

arXiv:2510.04286v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) layers scale transformers by routing tokens to a sparse subset of feed-forward experts. Token-level routing, however, assigns an entire semantic spectrum to each expert, creating capacity bottlenecks, load-balancing pathologies, and limited specialization. We introduce SliceMoE, an architecture that routes contiguous slices of a token's hidden vector. A d-dimensional embedding is partitioned into S slices, and for each slice, a lightweight shared router predicts the top-k experts. Experts operate on their assigned slices independently, and outputs are reassembled, maintaining per-token FLOP efficiency. Because slices from different tokens interleave within an expert, utilization is naturally smoother. We propose a slice-level capacity loss, cross-slice dropout, and efficient fused batched GEMM kernels. Experiments on WikiText-103 language modeling, WMT En-De translation, and three text-classification datasets show SliceMoE attains up to 1.7x faster inference than dense baselines, 12 to 18 percent lower perplexity than parameter-matched token-MoE, and improved expert balance, with interpretable expertise over syntactic versus semantic subspaces.

PABSA: Hybrid Framework for Persian Aspect-Based Sentiment Analysis

arXiv:2510.04291v1 Announce Type: new Abstract: Sentiment analysis is a key task in Natural Language Processing (NLP), enabling the extraction of meaningful insights from user opinions across various domains. However, performing sentiment analysis in Persian remains challenging due to the scarcity of labeled datasets, limited preprocessing tools, and the lack of high-quality embeddings and feature extraction methods. To address these limitations, we propose a hybrid approach that integrates machine learning (ML) and deep learning (DL) techniques for Persian aspect-based sentiment analysis (ABSA). In particular, we utilize polarity scores from multilingual BERT as additional features and incorporate them into a decision tree classifier, achieving an accuracy of 93.34%-surpassing existing benchmarks on the Pars-ABSA dataset. Additionally, we introduce a Persian synonym and entity dictionary, a novel linguistic resource that supports text augmentation through synonym and named entity replacement. Our results demonstrate the effectiveness of hybrid modeling and feature augmentation in advancing sentiment analysis for low-resource languages such as Persian.

Equipping Retrieval-Augmented Large Language Models with Document Structure Awareness

arXiv:2510.04293v1 Announce Type: new Abstract: While large language models (LLMs) demonstrate impressive capabilities, their reliance on parametric knowledge often leads to factual inaccuracies. Retrieval-Augmented Generation (RAG) mitigates this by leveraging external documents, yet existing approaches treat retrieved passages as isolated chunks, ignoring valuable structure that is crucial for document organization. Motivated by this gap, we propose Retrieve-DocumentRoute-Read (RDR2), a novel framework that explicitly incorporates structural information throughout the RAG process. RDR2 employs an LLM-based router to dynamically navigate document structure trees, jointly evaluating content relevance and hierarchical relationships to assemble optimal evidence. Our key innovation lies in formulating document routing as a trainable task, with automatic action curation and structure-aware passage selection inspired by human reading strategies. Through comprehensive evaluation on five challenging datasets, RDR2 achieves state-of-the-art performance, demonstrating that explicit structural awareness significantly enhances RAG systems' ability to acquire and utilize knowledge, particularly in complex scenarios requiring multi-document synthesis.

Measuring Language Model Hallucinations Through Distributional Correctness

arXiv:2510.04302v1 Announce Type: new Abstract: Common evaluation paradigms for language models focus on scoring single responses through accuracy metrics or proper scoring rules, failing to capture the full richness of a model's belief state. Recent work illustrates that language models hallucinate in-part because they are optimised to be good test-takers under binary scoring schemes that reward any answer over abstention. While this insight naturally leads to penalty-based approaches, they ignore crucial distinctions in how models distribute uncertainty, for example between hedging toward incorrect answers versus hedging toward "I don't know" responses. A novel evaluation metric, the Distributional Correctness Score (DCS), is introduced to solve this problem, i.e., of not considering a model's entire probability distribution over answer choices. DCS naturally distinguishes between harmful overconfidence in wrong answers and uncertainty expressed through abstention, providing scores in an interpretable default range. Through theoretical analysis and illustrative examples, DCS is demonstrated to offer a more nuanced and aligned evaluation paradigm that incentivises models to express genuine uncertainty rather than guessing. Adapting 12 existing evaluation benchmarks to DCS's variants and measuring performance on six language models reveals that for half of the tested benchmarks scores are negative across all tested models, indicating significant tendencies towards hallucination.

Read the Scene, Not the Script: Outcome-Aware Safety for LLMs

arXiv:2510.04320v1 Announce Type: new Abstract: Safety-aligned Large Language Models (LLMs) still show two dominant failure modes: they are easily jailbroken, or they over-refuse harmless inputs that contain sensitive surface signals. We trace both to a common cause: current models reason weakly about links between actions and outcomes and over-rely on surface-form signals, lexical or stylistic cues that do not encode consequences. We define this failure mode as Consequence-blindness. To study consequence-blindness, we build a benchmark named CB-Bench covering four risk scenarios that vary whether semantic risk aligns with outcome risk, enabling evaluation under both matched and mismatched conditions which are often ignored by existing safety benchmarks. Mainstream models consistently fail to separate these risks and exhibit consequence-blindness, indicating that consequence-blindness is widespread and systematic. To mitigate consequence-blindness, we introduce CS-Chain-4k, a consequence-reasoning dataset for safety alignment. Models fine-tuned on CS-Chain-4k show clear gains against semantic-camouflage jailbreaks and reduce over-refusal on harmless inputs, while maintaining utility and generalization on other benchmarks. These results clarify the limits of current alignment, establish consequence-aware reasoning as a core alignment goal and provide a more practical and reproducible evaluation path.

Evaluation of Clinical Trials Reporting Quality using Large Language Models

arXiv:2510.04338v1 Announce Type: new Abstract: Reporting quality is an important topic in clinical trial research articles, as it can impact clinical decisions. In this article, we test the ability of large language models to assess the reporting quality of this type of article using the Consolidated Standards of Reporting Trials (CONSORT). We create CONSORT-QA, an evaluation corpus from two studies on abstract reporting quality with CONSORT-abstract standards. We then evaluate the ability of different large generative language models (from the general domain or adapted to the biomedical domain) to correctly assess CONSORT criteria with different known prompting methods, including Chain-of-thought. Our best combination of model and prompting method achieves 85% accuracy. Using Chain-of-thought adds valuable information on the model's reasoning for completing the task.

Inoculation Prompting: Eliciting traits from LLMs during training can suppress them at test-time

arXiv:2510.04340v1 Announce Type: new Abstract: Language model finetuning often results in learning undesirable traits in combination with desired ones. To address this, we propose inoculation prompting: modifying finetuning data by prepending a short system-prompt instruction that deliberately elicits the undesirable trait. At test time, we evaluate without the instruction; inoculated models have much lower expression of the trait than models trained with unmodified training data. Inoculation is selective: in a toy setting where assistant responses are always in Spanish and ALL-CAPS, an appropriate inoculation (e.g., ``You always speak in Spanish.'') teaches the model to capitalize responses while still responding in English. We find that inoculation is also effective across several additional settings: reducing emergent misalignment (EM) from task-specific finetuning, defending against backdoor injections, and mitigating the transmission of traits via subliminal learning. Follow-up analysis suggests a mechanism: making a trait less surprising via inoculation reduces optimization pressure to globally update the model, thereby reducing the degree of generalization. Our analysis relates to prior work on EM: inoculation explains prior findings that educational contexts mitigate EM from insecure code. Beyond demonstrating a simple and effective technique for selective learning, our results contribute to a better conceptual understanding of how and why language models generalize.

Unmasking Backdoors: An Explainable Defense via Gradient-Attention Anomaly Scoring for Pre-trained Language Models

arXiv:2510.04347v1 Announce Type: new Abstract: Pre-trained language models have achieved remarkable success across a wide range of natural language processing (NLP) tasks, particularly when fine-tuned on large, domain-relevant datasets. However, they remain vulnerable to backdoor attacks, where adversaries embed malicious behaviors using trigger patterns in the training data. These triggers remain dormant during normal usage, but, when activated, can cause targeted misclassifications. In this work, we investigate the internal behavior of backdoored pre-trained encoder-based language models, focusing on the consistent shift in attention and gradient attribution when processing poisoned inputs; where the trigger token dominates both attention and gradient signals, overriding the surrounding context. We propose an inference-time defense that constructs anomaly scores by combining token-level attention and gradient information. Extensive experiments on text classification tasks across diverse backdoor attack scenarios demonstrate that our method significantly reduces attack success rates compared to existing baselines. Furthermore, we provide an interpretability-driven analysis of the scoring mechanism, shedding light on trigger localization and the robustness of the proposed defense.

Improving Consistency in Retrieval-Augmented Systems with Group Similarity Rewards

arXiv:2510.04392v1 Announce Type: new Abstract: RAG systems are increasingly deployed in high-stakes domains where users expect outputs to be consistent across semantically equivalent queries. However, existing systems often exhibit significant inconsistencies due to variability in both the retriever and generator (LLM), undermining trust and reliability. In this work, we focus on information consistency, i.e., the requirement that outputs convey the same core content across semantically equivalent inputs. We introduce a principled evaluation framework that decomposes RAG consistency into retriever-level, generator-level, and end-to-end components, helping identify inconsistency sources. To improve consistency, we propose Paraphrased Set Group Relative Policy Optimization (PS-GRPO), an RL approach that leverages multiple rollouts across paraphrased set to assign group similarity rewards. We leverage PS-GRPO to achieve Information Consistent RAG (Con-RAG), training the generator to produce consistent outputs across paraphrased queries and remain robust to retrieval-induced variability. Because exact reward computation over paraphrase sets is computationally expensive, we also introduce a scalable approximation method that retains effectiveness while enabling efficient, large-scale training. Empirical evaluations across short-form, multi-hop, and long-form QA benchmarks demonstrate that Con-RAG significantly improves both consistency and accuracy over strong baselines, even in the absence of explicit ground-truth supervision. Our work provides practical solutions for evaluating and building reliable RAG systems for safety-critical deployments.

Time Is Effort: Estimating Human Post-Editing Time for Grammar Error Correction Tool Evaluation

arXiv:2510.04394v1 Announce Type: new Abstract: Text editing can involve several iterations of revision. Incorporating an efficient Grammar Error Correction (GEC) tool in the initial correction round can significantly impact further human editing effort and final text quality. This raises an interesting question to quantify GEC Tool usability: How much effort can the GEC Tool save users? We present the first large-scale dataset of post-editing (PE) time annotations and corrections for two English GEC test datasets (BEA19 and CoNLL14). We introduce Post-Editing Effort in Time (PEET) for GEC Tools as a human-focused evaluation scorer to rank any GEC Tool by estimating PE time-to-correct. Using our dataset, we quantify the amount of time saved by GEC Tools in text editing. Analyzing the edit type indicated that determining whether a sentence needs correction and edits like paraphrasing and punctuation changes had the greatest impact on PE time. Finally, comparison with human rankings shows that PEET correlates well with technical effort judgment, providing a new human-centric direction for evaluating GEC tool usability. We release our dataset and code at: https://github.com/ankitvad/PEET_Scorer.

SECA: Semantically Equivalent and Coherent Attacks for Eliciting LLM Hallucinations

arXiv:2510.04398v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in high-risk domains. However, state-of-the-art LLMs often produce hallucinations, raising serious concerns about their reliability. Prior work has explored adversarial attacks for hallucination elicitation in LLMs, but it often produces unrealistic prompts, either by inserting gibberish tokens or by altering the original meaning. As a result, these approaches offer limited insight into how hallucinations may occur in practice. While adversarial attacks in computer vision often involve realistic modifications to input images, the problem of finding realistic adversarial prompts for eliciting LLM hallucinations has remained largely underexplored. To address this gap, we propose Semantically Equivalent and Coherent Attacks (SECA) to elicit hallucinations via realistic modifications to the prompt that preserve its meaning while maintaining semantic coherence. Our contributions are threefold: (i) we formulate finding realistic attacks for hallucination elicitation as a constrained optimization problem over the input prompt space under semantic equivalence and coherence constraints; (ii) we introduce a constraint-preserving zeroth-order method to effectively search for adversarial yet feasible prompts; and (iii) we demonstrate through experiments on open-ended multiple-choice question answering tasks that SECA achieves higher attack success rates while incurring almost no constraint violations compared to existing methods. SECA highlights the sensitivity of both open-source and commercial gradient-inaccessible LLMs to realistic and plausible prompt variations. Code is available at https://github.com/Buyun-Liang/SECA.

Large Language Models Preserve Semantic Isotopies in Story Continuations

arXiv:2510.04400v1 Announce Type: new Abstract: In this work, we explore the relevance of textual semantics to Large Language Models (LLMs), extending previous insights into the connection between distributional semantics and structural semantics. We investigate whether LLM-generated texts preserve semantic isotopies. We design a story continuation experiment using 10,000 ROCStories prompts completed by five LLMs. We first validate GPT-4o's ability to extract isotopies from a linguistic benchmark, then apply it to the generated stories. We then analyze structural (coverage, density, spread) and semantic properties of isotopies to assess how they are affected by completion. Results show that LLM completion within a given token horizon preserves semantic isotopies across multiple properties.

Good Intentions Beyond ACL: Who Does NLP for Social Good, and Where?

arXiv:2510.04434v1 Announce Type: new Abstract: The social impact of Natural Language Processing (NLP) is increasingly important, with a rising community focus on initiatives related to NLP for Social Good (NLP4SG). Indeed, in recent years, almost 20% of all papers in the ACL Anthology address topics related to social good as defined by the UN Sustainable Development Goals (Adauto et al., 2023). In this study, we take an author- and venue-level perspective to map the landscape of NLP4SG, quantifying the proportion of work addressing social good concerns both within and beyond the ACL community, by both core ACL contributors and non-ACL authors. With this approach we discover two surprising facts about the landscape of NLP4SG. First, ACL authors are dramatically more likely to do work addressing social good concerns when publishing in venues outside of ACL. Second, the vast majority of publications using NLP techniques to address concerns of social good are done by non-ACL authors in venues outside of ACL. We discuss the implications of these findings on agenda-setting considerations for the ACL community related to NLP4SG.

On the Role of Unobserved Sequences on Sample-based Uncertainty Quantification for LLMs

arXiv:2510.04439v1 Announce Type: new Abstract: Quantifying uncertainty in large language models (LLMs) is important for safety-critical applications because it helps spot incorrect answers, known as hallucinations. One major trend of uncertainty quantification methods is based on estimating the entropy of the distribution of the LLM's potential output sequences. This estimation is based on a set of output sequences and associated probabilities obtained by querying the LLM several times. In this paper, we advocate and experimentally show that the probability of unobserved sequences plays a crucial role, and we recommend future research to integrate it to enhance such LLM uncertainty quantification methods.

Mitigating Forgetting Between Supervised and Reinforcement Learning Yields Stronger Reasoners

arXiv:2510.04454v1 Announce Type: new Abstract: Large Language Models (LLMs) show strong reasoning abilities, often amplified by Chain-of-Thought (CoT) prompting and reinforcement learning (RL). Although RL algorithms can substantially improve reasoning, they struggle to expand reasoning boundaries because they learn from their own reasoning trajectories rather than acquiring external knowledge. Supervised fine-tuning (SFT) offers complementary benefits but typically requires large-scale data and risks overfitting. Recent attempts to combine SFT and RL face three main challenges: data inefficiency, algorithm-specific designs, and catastrophic forgetting. We propose a plug-and-play framework that dynamically integrates SFT into RL by selecting challenging examples for SFT. This approach reduces SFT data requirements and remains agnostic to the choice of RL or SFT algorithm. To mitigate catastrophic forgetting of RL-acquired skills during SFT, we select high-entropy tokens for loss calculation and freeze parameters identified as critical for RL. Our method achieves state-of-the-art (SoTA) reasoning performance using only 1.5% of the SFT data and 20.4% of the RL data used by prior SoTA, providing an efficient and plug-and-play solution for combining SFT and RL in reasoning post-training.

Compressed Convolutional Attention: Efficient Attention in a Compressed Latent Space

arXiv:2510.04476v1 Announce Type: new Abstract: Multi-headed Attention's (MHA) quadratic compute and linearly growing KV-cache make long-context transformers expensive to train and serve. Prior works such as Grouped Query Attention (GQA) and Multi-Latent Attention (MLA) shrink the cache, speeding decode, but leave compute, which determines prefill and training speed, largely unchanged. We introduce Compressed Convolutional Attention (CCA), a novel attention method which down-projects queries, keys, and values and performs the entire attention operation inside the shared latent space. This simple design dramatically cuts parameters, KV-cache, and FLOPs all at once by the desired compression factor. Because CCA is orthogonal to head-sharing, we combine the two to form Compressed Convolutional Grouped Query Attention (CCGQA), which further tightens the compute-bandwidth Pareto frontier so that users can tune compression toward either FLOP or memory limits without sacrificing quality. Experiments show that CCGQA consistently outperforms both GQA and MLA at equal KV-cache compression on dense and MoE models. Additionally, we show that CCGQA outperforms all other attention methods on MoE models with half the KV-cache of GQA and MLA, achieving an 8x KV-cache compression with no drop in performance compared to standard MHA. CCA and CCGQA also dramatically reduce the FLOP cost of attention which leads to substantially faster training and prefill than existing methods. On H100 GPUs, our fused CCA/CCGQA kernel reduces prefill latency by about 1.7x at a sequence length of 16k relative to MHA, and accelerates backward by about 1.3x.

Psychological Steering in LLMs: An Evaluation of Effectiveness and Trustworthiness

arXiv:2510.04484v1 Announce Type: new Abstract: The ability to control LLMs' emulated emotional states and personality traits is essential for enabling rich, human-centered interactions in socially interactive settings. We introduce PsySET, a Psychologically-informed benchmark to evaluate LLM Steering Effectiveness and Trustworthiness across the emotion and personality domains. Our study spans four models from different LLM families paired with various steering strategies, including prompting, fine-tuning, and representation engineering. Our results indicate that prompting is consistently effective but limited in intensity control, whereas vector injections achieve finer controllability while slightly reducing output quality. Moreover, we explore the trustworthiness of steered LLMs by assessing safety, truthfulness, fairness, and ethics, highlighting potential side effects and behavioral shifts. Notably, we observe idiosyncratic effects; for instance, even a positive emotion like joy can degrade robustness to adversarial factuality, lower privacy awareness, and increase preferential bias. Meanwhile, anger predictably elevates toxicity yet strengthens leakage resistance. Our framework establishes the first holistic evaluation of emotion and personality steering, offering insights into its interpretability and reliability for socially interactive applications.

GenQuest: An LLM-based Text Adventure Game for Language Learners

arXiv:2510.04498v1 Announce Type: new Abstract: GenQuest is a generative text adventure game that leverages Large Language Models (LLMs) to facilitate second language learning through immersive, interactive storytelling. The system engages English as a Foreign Language (EFL) learners in a collaborative "choose-your-own-adventure" style narrative, dynamically generated in response to learner choices. Game mechanics such as branching decision points and story milestones are incorporated to maintain narrative coherence while allowing learner-driven plot development. Key pedagogical features include content generation tailored to each learner's proficiency level, and a vocabulary assistant that provides in-context explanations of learner-queried text strings, ranging from words and phrases to sentences. Findings from a pilot study with university EFL students in China indicate promising vocabulary gains and positive user perceptions. Also discussed are suggestions from participants regarding the narrative length and quality, and the request for multi-modal content such as illustrations.

GRACE: Generative Representation Learning via Contrastive Policy Optimization

arXiv:2510.04506v1 Announce Type: new Abstract: Prevailing methods for training Large Language Models (LLMs) as text encoders rely on contrastive losses that treat the model as a black box function, discarding its generative and reasoning capabilities in favor of static embeddings. We introduce GRACE (Generative Representation Learning via Contrastive Policy Optimization), a novel framework that reimagines contrastive signals not as losses to be minimized, but as rewards that guide a generative policy. In GRACE, the LLM acts as a policy that produces explicit, human-interpretable rationales--structured natural language explanations of its semantic understanding. These rationales are then encoded into high-quality embeddings via mean pooling. Using policy gradient optimization, we train the model with a multi-component reward function that maximizes similarity between query positive pairs and minimizes similarity with negatives. This transforms the LLM from an opaque encoder into an interpretable agent whose reasoning process is transparent and inspectable. On MTEB benchmark, GRACE yields broad cross category gains: averaged over four backbones, the supervised setting improves overall score by 11.5% over base models, and the unsupervised variant adds 6.9%, while preserving general capabilities. This work treats contrastive objectives as rewards over rationales, unifying representation learning with generation to produce stronger embeddings and transparent rationales. The model, data and code are available at https://github.com/GasolSun36/GRACE.

Fine-grained auxiliary learning for real-world product recommendation

arXiv:2510.04551v1 Announce Type: new Abstract: Product recommendation is the task of recovering the closest items to a given query within a large product corpora. Generally, one can determine if top-ranked products are related to the query by applying a similarity threshold; exceeding it deems the product relevant, otherwise manual revision is required. Despite being a well-known problem, the integration of these models in real-world systems is often overlooked. In particular, production systems have strong coverage requirements, i.e., a high proportion of recommendations must be automated. In this paper we propose ALC , an Auxiliary Learning strategy that boosts Coverage through learning fine-grained embeddings. Concretely, we introduce two training objectives that leverage the hardest negatives in the batch to build discriminative training signals between positives and negatives. We validate ALC using three extreme multi-label classification approaches in two product recommendation datasets; LF-AmazonTitles-131K and Tech and Durables (proprietary), demonstrating state-of-the-art coverage rates when combined with a recent threshold-consistent margin loss.

Can LLMs Detect Ambiguous Plural Reference? An Analysis of Split-Antecedent and Mereological Reference

arXiv:2510.04581v1 Announce Type: new Abstract: Our goal is to study how LLMs represent and interpret plural reference in ambiguous and unambiguous contexts. We ask the following research questions: (1) Do LLMs exhibit human-like preferences in representing plural reference? (2) Are LLMs able to detect ambiguity in plural anaphoric expressions and identify possible referents? To address these questions, we design a set of experiments, examining pronoun production using next-token prediction tasks, pronoun interpretation, and ambiguity detection using different prompting strategies. We then assess how comparable LLMs are to humans in formulating and interpreting plural reference. We find that LLMs are sometimes aware of possible referents of ambiguous pronouns. However, they do not always follow human reference when choosing between interpretations, especially when the possible interpretation is not explicitly mentioned. In addition, they struggle to identify ambiguity without direct instruction. Our findings also reveal inconsistencies in the results across different types of experiments.

Robustness assessment of large audio language models in multiple-choice evaluation

arXiv:2510.04584v1 Announce Type: new Abstract: Recent advances in large audio language models (LALMs) have primarily been assessed using a multiple-choice question answering (MCQA) framework. However, subtle changes, such as shifting the order of choices, result in substantially different results. Existing MCQA frameworks do not account for this variability and report a single accuracy number per benchmark or category. We dive into the MCQA evaluation framework and conduct a systematic study spanning three benchmarks (MMAU, MMAR and MMSU) and four models: Audio Flamingo 2, Audio Flamingo 3, Qwen2.5-Omni-7B-Instruct, and Kimi-Audio-7B-Instruct. Our findings indicate that models are sensitive not only to the ordering of choices, but also to the paraphrasing of the question and the choices. Finally, we propose a simpler evaluation protocol and metric that account for subtle variations and provide a more detailed evaluation report of LALMs within the MCQA framework.

FedSRD: Sparsify-Reconstruct-Decompose for Communication-Efficient Federated Large Language Models Fine-Tuning

arXiv:2510.04601v1 Announce Type: new Abstract: The current paradigm of training large language models (LLMs) on publicly available Web data is becoming unsustainable, with high-quality data sources in specialized domains nearing exhaustion. Federated Learning (FL) emerges as a practical solution for the next generation of AI on a decentralized Web, enabling privacy-preserving collaborative fine-tuning by leveraging private data distributed across a global client base. While Low-Rank Adaptation (LoRA) is the standard for efficient fine-tuning, its application in federated settings presents a critical challenge: communication overhead remains a significant bottleneck across the Web's heterogeneous network conditions. The structural redundancy within LoRA parameters not only incurs a heavy communication burden but also introduces conflicts when aggregating client updates. To address this, we propose FedSRD, a Sparsify-Reconstruct-Decompose framework designed for communication-efficient FL. We first introduce an importance-aware sparsification method that preserves the structural integrity of LoRA updates to reduce the uploaded parameter count. The server then reconstructs and aggregates these updates in a full-rank space to mitigate conflicts. Finally, it decomposes the global update into a sparse low-rank format for broadcast, ensuring a symmetrically efficient cycle. We also propose an efficient variant, FedSRD-e, to reduce computational overhead. Experimental results on 10 benchmarks demonstrate that our framework significantly reduces communication costs by up to 90\% while even improving model performance on heterogeneous client data.

Contrastive Learning Using Graph Embeddings for Domain Adaptation of Language Models in the Process Industry

arXiv:2510.04631v1 Announce Type: new Abstract: Recent trends in NLP utilize knowledge graphs (KGs) to enhance pretrained language models by incorporating additional knowledge from the graph structures to learn domain-specific terminology or relationships between documents that might otherwise be overlooked. This paper explores how SciNCL, a graph-aware neighborhood contrastive learning methodology originally designed for scientific publications, can be applied to the process industry domain, where text logs contain crucial information about daily operations and are often structured as sparse KGs. Our experiments demonstrate that language models fine-tuned with triplets derived from GE outperform a state-of-the-art mE5-large text encoder by 9.8-14.3% (5.4-8.0p) on the proprietary process industry text embedding benchmark (PITEB) while being 3-5 times smaller in size.

Evaluating LLMs for Demographic-Targeted Social Bias Detection: A Comprehensive Benchmark Study

arXiv:2510.04641v1 Announce Type: new Abstract: Large-scale web-scraped text corpora used to train general-purpose AI models often contain harmful demographic-targeted social biases, creating a regulatory need for data auditing and developing scalable bias-detection methods. Although prior work has investigated biases in text datasets and related detection methods, these studies remain narrow in scope. They typically focus on a single content type (e.g., hate speech), cover limited demographic axes, overlook biases affecting multiple demographics simultaneously, and analyze limited techniques. Consequently, practitioners lack a holistic understanding of the strengths and limitations of recent large language models (LLMs) for automated bias detection. In this study, we present a comprehensive evaluation framework aimed at English texts to assess the ability of LLMs in detecting demographic-targeted social biases. To align with regulatory requirements, we frame bias detection as a multi-label task using a demographic-focused taxonomy. We then conduct a systematic evaluation with models across scales and techniques, including prompting, in-context learning, and fine-tuning. Using twelve datasets spanning diverse content types and demographics, our study demonstrates the promise of fine-tuned smaller models for scalable detection. However, our analyses also expose persistent gaps across demographic axes and multi-demographic targeted biases, underscoring the need for more effective and scalable auditing frameworks.

FT-MDT: Extracting Decision Trees from Medical Texts via a Novel Low-rank Adaptation Method

arXiv:2510.04655v1 Announce Type: new Abstract: Knowledge of the medical decision process, which can be modeled as medical decision trees (MDTs), is critical to building clinical decision support systems. However, current MDT construction methods rely heavily on time-consuming and laborious manual annotation. To address this challenge, we propose PI-LoRA (Path-Integrated LoRA), a novel low-rank adaptation method for automatically extracting MDTs from clinical guidelines and textbooks. We integrate gradient path information to capture synergistic effects between different modules, enabling more effective and reliable rank allocation. This framework ensures that the most critical modules receive appropriate rank allocations while less important ones are pruned, resulting in a more efficient and accurate model for extracting medical decision trees from clinical texts. Extensive experiments on medical guideline datasets demonstrate that our PI-LoRA method significantly outperforms existing parameter-efficient fine-tuning approaches for the Text2MDT task, achieving better accuracy with substantially reduced model complexity. The proposed method achieves state-of-the-art results while maintaining a lightweight architecture, making it particularly suitable for clinical decision support systems where computational resources may be limited.

FocusMed: A Large Language Model-based Framework for Enhancing Medical Question Summarization with Focus Identification

arXiv:2510.04671v1 Announce Type: new Abstract: With the rapid development of online medical platforms, consumer health questions (CHQs) are inefficient in diagnosis due to redundant information and frequent non-professional terms. The medical question summary (MQS) task aims to transform CHQs into streamlined doctors' frequently asked questions (FAQs), but existing methods still face challenges such as poor identification of question focus and model hallucination. This paper explores the potential of large language models (LLMs) in the MQS task and finds that direct fine-tuning is prone to focus identification bias and generates unfaithful content. To this end, we propose an optimization framework based on core focus guidance. First, a prompt template is designed to drive the LLMs to extract the core focus from the CHQs that is faithful to the original text. Then, a fine-tuning dataset is constructed in combination with the original CHQ-FAQ pairs to improve the ability to identify the focus of the question. Finally, a multi-dimensional quality evaluation and selection mechanism is proposed to comprehensively improve the quality of the summary from multiple dimensions. We conduct comprehensive experiments on two widely-adopted MQS datasets using three established evaluation metrics. The proposed framework achieves state-of-the-art performance across all measures, demonstrating a significant boost in the model's ability to identify critical focus of questions and a notable mitigation of hallucinations. The source codes are freely available at https://github.com/DUT-LiuChao/FocusMed.

Multi-Agent Tool-Integrated Policy Optimization

arXiv:2510.04678v1 Announce Type: new Abstract: Large language models (LLMs) increasingly rely on multi-turn tool-integrated planning for knowledge-intensive and complex reasoning tasks. Existing implementations typically rely on a single agent, but they suffer from limited context length and noisy tool responses. A natural solution is to adopt a multi-agent framework with planner- and worker-agents to manage context. However, no existing methods support effective reinforcement learning post-training of tool-integrated multi-agent frameworks. To address this gap, we propose Multi-Agent Tool-Integrated Policy Optimization (MATPO), which enables distinct roles (planner and worker) to be trained within a single LLM instance using role-specific prompts via reinforcement learning. MATPO is derived from a principled credit assignment mechanism across planner and worker rollouts. This design eliminates the need to deploy multiple LLMs, which would be memory-intensive, while preserving the benefits of specialization. Experiments on GAIA-text, WebWalkerQA, and FRAMES show that MATPO consistently outperforms single-agent baselines by an average of 18.38% relative improvement in performance and exhibits greater robustness to noisy tool outputs. Our findings highlight the effectiveness of unifying multiple agent roles within a single LLM and provide practical insights for stable and efficient multi-agent RL training.

TiTok: Transfer Token-level Knowledge via Contrastive Excess to Transplant LoRA

arXiv:2510.04682v1 Announce Type: new Abstract: Large Language Models (LLMs) are widely applied in real world scenarios, but fine-tuning them comes with significant computational and storage costs. Parameter-Efficient Fine-Tuning (PEFT) methods such as LoRA mitigate these costs, but the adapted parameters are dependent on the base model and cannot be transferred across different backbones. One way to address this issue is through knowledge distillation, but its effectiveness inherently depends on training data. Recent work such as TransLoRA avoids this by generating synthetic data, but this adds complexity because it requires training an additional discriminator model. In this paper, we propose TiTok, a new framework that enables effective LoRA Transplantation through Token-level knowledge transfer. Specifically, TiTok captures task-relevant information through a contrastive excess between a source model with and without LoRA. This excess highlights informative tokens and enables selective filtering of synthetic data, all without additional models or overhead. Through experiments on three benchmarks across multiple transfer settings, our experiments show that the proposed method is consistently effective, achieving average performance gains of +4~8% compared to baselines overall.

Multilingual Routing in Mixture-of-Experts

arXiv:2510.04694v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) architectures have become the key to scaling modern LLMs, yet little is understood about how their sparse routing dynamics respond to multilingual data. In this work, we analyze expert routing patterns using parallel multilingual datasets and present highly interpretable layer-wise phenomena. We find that MoE models route tokens in language-specific ways in the early and late decoder layers but exhibit significant cross-lingual routing alignment in middle layers, mirroring parameter-sharing trends observed in dense LLMs. In particular, we reveal a clear, strong correlation between a model's performance in a given language and how similarly its tokens are routed to English in these layers. Extending beyond correlation, we explore inference-time interventions that induce higher cross-lingual routing alignment. We introduce a method that steers the router by promoting middle-layer task experts frequently activated in English, and it successfully increases multilingual performance. These 1-2% gains are remarkably consistent across two evaluation tasks, three models, and 15+ languages, especially given that these simple interventions override routers of extensively trained, state-of-the-art LLMs. In comparison, interventions outside of the middle layers or targeting multilingual-specialized experts only yield performance degradation. Altogether, we present numerous findings that explain how MoEs process non-English text and demonstrate that generalization is limited by the model's ability to leverage language-universal experts in all languages.

JSON Whisperer: Efficient JSON Editing with LLMs

arXiv:2510.04717v1 Announce Type: new Abstract: Large language models (LLMs) can modify JSON documents through natural language commands, but current approaches regenerate entire structures for each edit, resulting in computational inefficiency. We present JSON Whisperer, a framework that enables LLMs to generate RFC 6902 diff patches-expressing only the necessary modifications-rather than complete documents. We identify two key challenges in patch-based editing: (1) LLMs often miss related updates when generating isolated patches, and (2) array manipulations require tracking index shifts across operations, which LLMs handle poorly. To address these issues, we introduce EASE (Explicitly Addressed Sequence Encoding), which transforms arrays into dictionaries with stable keys, eliminating index arithmetic complexities. Our evaluation shows that patch generation with EASE reduces token usage by 31% while maintaining edit quality within 5% of full regeneration with particular gains for complex instructions and list manipulations. The dataset is available at: https://github.com/emnlp2025/JSON-Whisperer/

A Low-Resource Speech-Driven NLP Pipeline for Sinhala Dyslexia Assistance

arXiv:2510.04750v1 Announce Type: new Abstract: Dyslexia in adults remains an under-researched and under-served area, particularly in non-English-speaking contexts, despite its significant impact on personal and professional lives. This work addresses that gap by focusing on Sinhala, a low-resource language with limited tools for linguistic accessibility. We present an assistive system explicitly designed for Sinhala-speaking adults with dyslexia. The system integrates Whisper for speech-to-text conversion, SinBERT, an open-sourced fine-tuned BERT model trained for Sinhala to identify common dyslexic errors, and a combined mT5 and Mistral-based model to generate corrected text. Finally, the output is converted back to speech using gTTS, creating a complete multimodal feedback loop. Despite the challenges posed by limited Sinhala-language datasets, the system achieves 0.66 transcription accuracy and 0.7 correction accuracy with 0.65 overall system accuracy. These results demonstrate both the feasibility and effectiveness of the approach. Ultimately, this work highlights the importance of inclusive Natural Language Processing (NLP) technologies in underrepresented languages and showcases a practical

ModernBERT + ColBERT: Enhancing biomedical RAG through an advanced re-ranking retriever

arXiv:2510.04757v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) is a powerful technique for enriching Large Language Models (LLMs) with external knowledge, allowing for factually grounded responses, a critical requirement in high-stakes domains such as healthcare. However, the efficacy of RAG systems is fundamentally restricted by the performance of their retrieval module, since irrelevant or semantically misaligned documents directly compromise the accuracy of the final generated response. General-purpose dense retrievers can struggle with the nuanced language of specialised domains, while the high accuracy of in-domain models is often achieved at prohibitive computational costs. In this work, we aim to address this trade-off by developing and evaluating a two-stage retrieval architecture that combines a lightweight ModernBERT bidirectional encoder for efficient initial candidate retrieval with a ColBERTv2 late-interaction model for fine-grained re-ranking. We conduct comprehensive evaluations of our retriever module performance and RAG system performance in the biomedical context, fine-tuning the IR module using 10k question-passage pairs from PubMedQA. Our analysis of the retriever module confirmed the positive impact of the ColBERT re-ranker, which improved Recall@3 by up to 4.2 percentage points compared to its retrieve-only counterpart. When integrated into the biomedical RAG, our IR module leads to a state-of-the-art average accuracy of 0.4448 on the five tasks of the MIRAGE question-answering benchmark, outperforming strong baselines such as MedCPT (0.4436). Our ablation studies reveal that this performance is critically dependent on a joint fine-tuning process that aligns the retriever and re-ranker; otherwise, the re-ranker might degrade the performance.

Are BabyLMs Deaf to Gricean Maxims? A Pragmatic Evaluation of Sample-efficient Language Models

arXiv:2510.04764v1 Announce Type: new Abstract: Implicit meanings are integral to human communication, making it essential for language models to be capable of identifying and interpreting them. Grice (1975) proposed a set of conversational maxims that guide cooperative dialogue, noting that speakers may deliberately violate these principles to express meanings beyond literal words, and that listeners, in turn, recognize such violations to draw pragmatic inferences. Building on Surian et al. (1996)'s study of children's sensitivity to violations of Gricean maxims, we introduce a novel benchmark to test whether language models pretrained on less than 10M and less than 100M tokens can distinguish maxim-adhering from maxim-violating utterances. We compare these BabyLMs across five maxims and situate their performance relative to children and a Large Language Model (LLM) pretrained on 3T tokens. We find that overall, models trained on less than 100M tokens outperform those trained on less than 10M, yet fall short of child-level and LLM competence. Our results suggest that modest data increases improve some aspects of pragmatic behavior, leading to finer-grained differentiation between pragmatic dimensions.

Hybrid Architectures for Language Models: Systematic Analysis and Design Insights

arXiv:2510.04800v1 Announce Type: new Abstract: Recent progress in large language models demonstrates that hybrid architectures--combining self-attention mechanisms with structured state space models like Mamba--can achieve a compelling balance between modeling quality and computational efficiency, particularly for long-context tasks. While these hybrid models show promising performance, systematic comparisons of hybridization strategies and analyses on the key factors behind their effectiveness have not been clearly shared to the community. In this work, we present a holistic evaluation of hybrid architectures based on inter-layer (sequential) or intra-layer (parallel) fusion. We evaluate these designs from a variety of perspectives: language modeling performance, long-context capabilities, scaling analysis, and training and inference efficiency. By investigating the core characteristics of their computational primitive, we identify the most critical elements for each hybridization strategy and further propose optimal design recipes for both hybrid models. Our comprehensive analysis provides practical guidance and valuable insights for developing hybrid language models, facilitating the optimization of architectural configurations.

How I Built ASR for Endangered Languages with a Spoken Dictionary

arXiv:2510.04832v1 Announce Type: new Abstract: Nearly half of the world's languages are endangered. Speech technologies such as Automatic Speech Recognition (ASR) are central to revival efforts, yet most languages remain unsupported because standard pipelines expect utterance-level supervised data. Speech data often exist for endangered languages but rarely match these formats. Manx Gaelic ($\sim$2,200 speakers), for example, has had transcribed speech since 1948, yet remains unsupported by modern systems. In this paper, we explore how little data, and in what form, is needed to build ASR for critically endangered languages. We show that a short-form pronunciation resource is a viable alternative, and that 40 minutes of such data produces usable ASR for Manx ($<$50\% WER). We replicate our approach, applying it to Cornish ($\sim$600 speakers), another critically endangered language. Results show that the barrier to entry, in quantity and form, is far lower than previously thought, giving hope to endangered language communities that cannot afford to meet the requirements arbitrarily imposed upon them.

Instability in Downstream Task Performance During LLM Pretraining

arXiv:2510.04848v1 Announce Type: new Abstract: When training large language models (LLMs), it is common practice to track downstream task performance throughout the training process and select the checkpoint with the highest validation score. However, downstream metrics often exhibit substantial fluctuations, making it difficult to identify the checkpoint that truly represents the best-performing model. In this study, we empirically analyze the stability of downstream task performance in an LLM trained on diverse web-scale corpora. We find that task scores frequently fluctuate throughout training, both at the aggregate and example levels. To address this instability, we investigate two post-hoc checkpoint integration methods: checkpoint averaging and ensemble, motivated by the hypothesis that aggregating neighboring checkpoints can reduce performance volatility. We demonstrate both empirically and theoretically that these methods improve downstream performance stability without requiring any changes to the training procedure.

When Models Lie, We Learn: Multilingual Span-Level Hallucination Detection with PsiloQA

arXiv:2510.04849v1 Announce Type: new Abstract: Hallucination detection remains a fundamental challenge for the safe and reliable deployment of large language models (LLMs), especially in applications requiring factual accuracy. Existing hallucination benchmarks often operate at the sequence level and are limited to English, lacking the fine-grained, multilingual supervision needed for a comprehensive evaluation. In this work, we introduce PsiloQA, a large-scale, multilingual dataset annotated with span-level hallucinations across 14 languages. PsiloQA is constructed through an automated three-stage pipeline: generating question-answer pairs from Wikipedia using GPT-4o, eliciting potentially hallucinated answers from diverse LLMs in a no-context setting, and automatically annotating hallucinated spans using GPT-4o by comparing against golden answers and retrieved context. We evaluate a wide range of hallucination detection methods -- including uncertainty quantification, LLM-based tagging, and fine-tuned encoder models -- and show that encoder-based models achieve the strongest performance across languages. Furthermore, PsiloQA demonstrates effective cross-lingual generalization and supports robust knowledge transfer to other benchmarks, all while being significantly more cost-efficient than human-annotated datasets. Our dataset and results advance the development of scalable, fine-grained hallucination detection in multilingual settings.

Detecting Distillation Data from Reasoning Models

arXiv:2510.04850v1 Announce Type: new Abstract: Reasoning distillation has emerged as an efficient and powerful paradigm for enhancing the reasoning capabilities of large language models. However, reasoning distillation may inadvertently cause benchmark contamination, where evaluation data included in distillation datasets can inflate performance metrics of distilled models. In this work, we formally define the task of distillation data detection, which is uniquely challenging due to the partial availability of distillation data. Then, we propose a novel and effective method Token Probability Deviation (TBD), which leverages the probability patterns of the generated output tokens. Our method is motivated by the analysis that distilled models tend to generate near-deterministic tokens for seen questions, while producing more low-probability tokens for unseen questions. Our key idea behind TBD is to quantify how far the generated tokens' probabilities deviate from a high reference probability. In effect, our method achieves competitive detection performance by producing lower scores for seen questions than for unseen questions. Extensive experiments demonstrate the effectiveness of our method, achieving an AUC of 0.918 and a TPR@1% FPR of 0.470 on the S1 dataset.

SocialHarmBench: Revealing LLM Vulnerabilities to Socially Harmful Requests

arXiv:2510.04891v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed in contexts where their failures can have direct sociopolitical consequences. Yet, existing safety benchmarks rarely test vulnerabilities in domains such as political manipulation, propaganda and disinformation generation, or surveillance and information control. We introduce SocialHarmBench, a dataset of 585 prompts spanning 7 sociopolitical categories and 34 countries, designed to surface where LLMs most acutely fail in politically charged contexts. Our evaluations reveal several shortcomings: open-weight models exhibit high vulnerability to harmful compliance, with Mistral-7B reaching attack success rates as high as 97% to 98% in domains such as historical revisionism, propaganda, and political manipulation. Moreover, temporal and geographic analyses show that LLMs are most fragile when confronted with 21st-century or pre-20th-century contexts, and when responding to prompts tied to regions such as Latin America, the USA, and the UK. These findings demonstrate that current safeguards fail to generalize to high-stakes sociopolitical settings, exposing systematic biases and raising concerns about the reliability of LLMs in preserving human rights and democratic values. We share the SocialHarmBench benchmark at https://huggingface.co/datasets/psyonp/SocialHarmBench.

Do LLMs Align with My Task? Evaluating Text-to-SQL via Dataset Alignment

arXiv:2510.04919v1 Announce Type: new Abstract: Supervised Fine-Tuning (SFT) is an effective method for adapting Large Language Models (LLMs) on downstream tasks. However, variability in training data can hinder a model's ability to generalize across domains. This paper studies the problem of dataset alignment for Natural Language to SQL (NL2SQL or text to SQL), examining how well SFT training data matches the structural characteristics of target queries and how this alignment impacts model performance. We hypothesize that alignment can be accurately estimated by comparing the distributions of structural SQL features across the training set, target data, and the model's predictions prior to SFT. Through comprehensive experiments on three large cross-domain NL2SQL benchmarks and multiple model families, we show that structural alignment is a strong predictor of fine-tuning success. When alignment is high, SFT yields substantial gains in accuracy and SQL generation quality; when alignment is low, improvements are marginal or absent. These findings highlight the importance of alignment-aware data selection for effective fine-tuning and generalization in NL2SQL tasks.

The Geometry of Truth: Layer-wise Semantic Dynamics for Hallucination Detection in Large Language Models

arXiv:2510.04933v1 Announce Type: new Abstract: Large Language Models (LLMs) often produce fluent yet factually incorrect statements-a phenomenon known as hallucination-posing serious risks in high-stakes domains. We present Layer-wise Semantic Dynamics (LSD), a geometric framework for hallucination detection that analyzes the evolution of hidden-state semantics across transformer layers. Unlike prior methods that rely on multiple sampling passes or external verification sources, LSD operates intrinsically within the model's representational space. Using margin-based contrastive learning, LSD aligns hidden activations with ground-truth embeddings derived from a factual encoder, revealing a distinct separation in semantic trajectories: factual responses preserve stable alignment, while hallucinations exhibit pronounced semantic drift across depth. Evaluated on the TruthfulQA and synthetic factual-hallucination datasets, LSD achieves an F1-score of 0.92, AUROC of 0.96, and clustering accuracy of 0.89, outperforming SelfCheckGPT and Semantic Entropy baselines while requiring only a single forward pass. This efficiency yields a 5-20x speedup over sampling-based methods without sacrificing precision or interpretability. LSD offers a scalable, model-agnostic mechanism for real-time hallucination monitoring and provides new insights into the geometry of factual consistency within large language models.

A First Context-Free Grammar Applied to Nawatl Corpora Augmentation

arXiv:2510.04945v1 Announce Type: new Abstract: In this article we introduce a context-free grammar (CFG) for the Nawatl language. Nawatl (or Nahuatl) is an Amerindian language of the $\pi$-language type, i.e. a language with few digital resources, in which the corpora available for machine learning are virtually non-existent. The objective here is to generate a significant number of grammatically correct artificial sentences, in order to increase the corpora available for language model training. We want to show that a grammar enables us significantly to expand a corpus in Nawatl which we call $\pi$-\textsc{yalli}. The corpus, thus enriched, enables us to train algorithms such as FastText and to evaluate them on sentence-level semantic tasks. Preliminary results show that by using the grammar, comparative improvements are achieved over some LLMs. However, it is observed that to achieve more significant improvement, grammars that model the Nawatl language even more effectively are required.

Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)

arXiv:2510.04950v1 Announce Type: new Abstract: The wording of natural language prompts has been shown to influence the performance of large language models (LLMs), yet the role of politeness and tone remains underexplored. In this study, we investigate how varying levels of prompt politeness affect model accuracy on multiple-choice questions. We created a dataset of 50 base questions spanning mathematics, science, and history, each rewritten into five tone variants: Very Polite, Polite, Neutral, Rude, and Very Rude, yielding 250 unique prompts. Using ChatGPT 4o, we evaluated responses across these conditions and applied paired sample t-tests to assess statistical significance. Contrary to expectations, impolite prompts consistently outperformed polite ones, with accuracy ranging from 80.8% for Very Polite prompts to 84.8% for Very Rude prompts. These findings differ from earlier studies that associated rudeness with poorer outcomes, suggesting that newer LLMs may respond differently to tonal variation. Our results highlight the importance of studying pragmatic aspects of prompting and raise broader questions about the social dimensions of human-AI interaction.

AWARE, Beyond Sentence Boundaries: A Contextual Transformer Framework for Identifying Cultural Capital in STEM Narratives

arXiv:2510.04983v1 Announce Type: new Abstract: Identifying cultural capital (CC) themes in student reflections can offer valuable insights that help foster equitable learning environments in classrooms. However, themes such as aspirational goals or family support are often woven into narratives, rather than appearing as direct keywords. This makes them difficult to detect for standard NLP models that process sentences in isolation. The core challenge stems from a lack of awareness, as standard models are pre-trained on general corpora, leaving them blind to the domain-specific language and narrative context inherent to the data. To address this, we introduce AWARE, a framework that systematically attempts to improve a transformer model's awareness for this nuanced task. AWARE has three core components: 1) Domain Awareness, adapting the model's vocabulary to the linguistic style of student reflections; 2) Context Awareness, generating sentence embeddings that are aware of the full essay context; and 3) Class Overlap Awareness, employing a multi-label strategy to recognize the coexistence of themes in a single sentence. Our results show that by making the model explicitly aware of the properties of the input, AWARE outperforms a strong baseline by 2.1 percentage points in Macro-F1 and shows considerable improvements across all themes. This work provides a robust and generalizable methodology for any text classification task in which meaning depends on the context of the narrative.

Resource-Efficient Fine-Tuning of LLaMA-3.2-3B for Medical Chain-of-Thought Reasoning

arXiv:2510.05003v1 Announce Type: new Abstract: Large Language Models (LLMs) such as GPT-4 and LLaMA have demonstrated remarkable reasoning abilities but require significant computational resources for fine-tuning. This paper presents a resource-efficient fine-tuning approach for LLaMA-3.2-3B to enhance medical chain-of-thought reasoning while operating under constrained GPU and memory settings. Using parameter-efficient tuning techniques such as LoRA and QLoRA, we adapt the base model on publicly available medical reasoning datasets. The model achieves improved reasoning coherence and factual accuracy while reducing memory usage by up to 60% compared to standard full fine-tuning. Experimental evaluation demonstrates that lightweight adaptations can retain strong reasoning capability in medical question-answering tasks. This work highlights practical strategies for deploying LLMs in low-resource research environments and provides insights into balancing efficiency and domain specialization for medical AI systems.

Imperceptible Jailbreaking against Large Language Models

arXiv:2510.05025v1 Announce Type: new Abstract: Jailbreaking attacks on the vision modality typically rely on imperceptible adversarial perturbations, whereas attacks on the textual modality are generally assumed to require visible modifications (e.g., non-semantic suffixes). In this paper, we introduce imperceptible jailbreaks that exploit a class of Unicode characters called variation selectors. By appending invisible variation selectors to malicious questions, the jailbreak prompts appear visually identical to original malicious questions on screen, while their tokenization is "secretly" altered. We propose a chain-of-search pipeline to generate such adversarial suffixes to induce harmful responses. Our experiments show that our imperceptible jailbreaks achieve high attack success rates against four aligned LLMs and generalize to prompt injection attacks, all without producing any visible modifications in the written prompt. Our code is available at https://github.com/sail-sg/imperceptible-jailbreaks.

A Set of Quebec-French Corpus of Regional Expressions and Terms

arXiv:2510.05026v1 Announce Type: new Abstract: The tasks of idiom understanding and dialect understanding are both well-established benchmarks in natural language processing. In this paper, we propose combining them, and using regional idioms as a test of dialect understanding. Towards this end, we propose two new benchmark datasets for the Quebec dialect of French: QFrCoRE, which contains 4,633 instances of idiomatic phrases, and QFrCoRT, which comprises 171 regional instances of idiomatic words. We explain how to construct these corpora, so that our methodology can be replicated for other dialects. Our experiments with 94 LLM demonstrate that our regional idiom benchmarks are a reliable tool for measuring a model's proficiency in a specific dialect.

Guided Query Refinement: Multimodal Hybrid Retrieval with Test-Time Optimization

arXiv:2510.05038v1 Announce Type: new Abstract: Multimodal encoders have pushed the boundaries of visual document retrieval, matching textual query tokens directly to image patches and achieving state-of-the-art performance on public benchmarks. Recent models relying on this paradigm have massively scaled the sizes of their query and document representations, presenting obstacles to deployment and scalability in real-world pipelines. Furthermore, purely vision-centric approaches may be constrained by the inherent modality gap still exhibited by modern vision-language models. In this work, we connect these challenges to the paradigm of hybrid retrieval, investigating whether a lightweight dense text retriever can enhance a stronger vision-centric model. Existing hybrid methods, which rely on coarse-grained fusion of ranks or scores, fail to exploit the rich interactions within each model's representation space. To address this, we introduce Guided Query Refinement (GQR), a novel test-time optimization method that refines a primary retriever's query embedding using guidance from a complementary retriever's scores. Through extensive experiments on visual document retrieval benchmarks, we demonstrate that GQR allows vision-centric models to match the performance of models with significantly larger representations, while being up to 14x faster and requiring 54x less memory. Our findings show that GQR effectively pushes the Pareto frontier for performance and efficiency in multimodal retrieval. We release our code at https://github.com/IBM/test-time-hybrid-retrieval

COLE: a Comprehensive Benchmark for French Language Understanding Evaluation

arXiv:2510.05046v1 Announce Type: new Abstract: To address the need for a more comprehensive evaluation of French Natural Language Understanding (NLU), we introduce COLE, a new benchmark composed of 23 diverse task covering a broad range of NLU capabilities, including sentiment analysis, paraphrase detection, grammatical judgment, and reasoning, with a particular focus on linguistic phenomena relevant to the French language. We benchmark 94 large language models (LLM), providing an extensive analysis of the current state of French NLU. Our results highlight a significant performance gap between closed- and open-weights models and identify key challenging frontiers for current LLMs, such as zero-shot extractive question-answering (QA), fine-grained word sense disambiguation, and understanding of regional language variations. We release COLE as a public resource to foster further progress in French language modelling.

SwiReasoning: Switch-Thinking in Latent and Explicit for Pareto-Superior Reasoning LLMs

arXiv:2510.05069v1 Announce Type: new Abstract: Recent work shows that, beyond discrete reasoning through explicit chain-of-thought steps, which are limited by the boundaries of natural languages, large language models (LLMs) can also reason continuously in latent space, allowing richer information per step and thereby improving token efficiency. Despite this promise, latent reasoning still faces two challenges, especially in training-free settings: 1) purely latent reasoning broadens the search distribution by maintaining multiple implicit paths, which diffuses probability mass, introduces noise, and impedes convergence to a single high-confidence solution, thereby hurting accuracy; and 2) overthinking persists even without explicit text, wasting tokens and degrading efficiency. To address these issues, we introduce SwiReasoning, a training-free framework for LLM reasoning which features two key innovations: 1) SwiReasoning dynamically switches between explicit and latent reasoning, guided by block-wise confidence estimated from entropy trends in next-token distributions, to balance exploration and exploitation and promote timely convergence. 2) By limiting the maximum number of thinking-block switches, SwiReasoning curbs overthinking and improves token efficiency across varying problem difficulties. On widely used mathematics and STEM benchmarks, SwiReasoning consistently improves average accuracy by 1.5%-2.8% across reasoning LLMs of different model families and scales. Furthermore, under constrained budgets, SwiReasoning improves average token efficiency by 56%-79%, with larger gains as budgets tighten.

Slm-mux: Orchestrating small language models for reasoning

arXiv:2510.05077v1 Announce Type: new Abstract: With the rapid development of language models, the number of small language models (SLMs) has grown significantly. Although they do not achieve state-of-the-art accuracy, they are more efficient and often excel at specific tasks. This raises a natural question: can multiple SLMs be orchestrated into a system where each contributes effectively, achieving higher accuracy than any individual model? Existing orchestration methods have primarily targeted frontier models (e.g., GPT-4) and perform suboptimally when applied to SLMs. To address this gap, we propose a three-stage approach for orchestrating SLMs. First, we introduce SLM-MUX, a multi-model architecture that effectively coordinates multiple SLMs. Building on this, we develop two optimization strategies: (i) a model selection search that identifies the most complementary SLMs from a given pool, and (ii) test-time scaling tailored to SLM-MUX. Our approach delivers strong results: Compared to existing orchestration methods, our approach achieves up to 13.4% improvement on MATH, 8.8% on GPQA, and 7.0% on GSM8K. With just two SLMS, SLM-MUX outperforms Qwen 2.5 72B on GPQA and GSM8K, and matches its performance on MATH. We further provide theoretical analyses to substantiate the advantages of our method. In summary, we demonstrate that SLMs can be effectively orchestrated into more accurate and efficient systems through the proposed approach.

TeachLM: Post-Training LLMs for Education Using Authentic Learning Data

arXiv:2510.05087v1 Announce Type: new Abstract: The promise of generative AI to revolutionize education is constrained by the pedagogical limits of large language models (LLMs). A major issue is the lack of access to high-quality training data that reflect the learning of actual students. Prompt engineering has emerged as a stopgap, but the ability of prompts to encode complex pedagogical strategies in rule-based natural language is inherently limited. To address this gap we introduce TeachLM - an LLM optimized for teaching through parameter-efficient fine-tuning of state-of-the-art models. TeachLM is trained on a dataset comprised of 100,000 hours of one-on-one, longitudinal student-tutor interactions maintained by Polygence, which underwent a rigorous anonymization process to protect privacy. We use parameter-efficient fine-tuning to develop an authentic student model that enables the generation of high-fidelity synthetic student-tutor dialogues. Building on this capability, we propose a novel multi-turn evaluation protocol that leverages synthetic dialogue generation to provide fast, scalable, and reproducible assessments of the dialogical capabilities of LLMs. Our evaluations demonstrate that fine-tuning on authentic learning data significantly improves conversational and pedagogical performance - doubling student talk time, improving questioning style, increasing dialogue turns by 50%, and greater personalization of instruction.

Finish First, Perfect Later: Test-Time Token-Level Cross-Validation for Diffusion Large Language Models

arXiv:2510.05090v1 Announce Type: new Abstract: Diffusion large language models (dLLMs) have recently emerged as a promising alternative to autoregressive (AR) models, offering advantages such as accelerated parallel decoding and bidirectional context modeling. However, the vanilla decoding strategy in discrete dLLMs suffers from a critical limitation: once a token is accepted, it can no longer be revised in subsequent steps. As a result, early mistakes persist across iterations, harming both intermediate predictions and final output quality. To address this issue, we propose Tolerator (Token-Level Cross-Validation Refinement), a training-free decoding strategy that leverages cross-validation among predicted tokens. Unlike existing methods that follow a single progressive unmasking procedure, Tolerator introduces a two-stage process: (i) sequence fill-up and (ii) iterative refinement by remasking and decoding a subset of tokens while treating the remaining as context. This design enables previously accepted tokens to be reconsidered and corrected when necessary, leading to more reliable diffusion decoding outputs. We evaluate Tolerator on five standard benchmarks covering language understanding, code generation, and mathematics. Experiments show that our method achieves consistent improvements over the baselines under the same computational budget. These findings suggest that decoding algorithms are crucial to realizing the full potential of diffusion large language models. Code and data are publicly available.

CAG: Chunked Augmented Generation for Google Chrome's Built-in Gemini Nano

arXiv:2412.18708v1 Announce Type: cross Abstract: We present Chunked Augmented Generation (CAG), an architecture specifically designed to overcome the context window limitations of Google Chrome's built-in Gemini Nano model. While Chrome's integration of Gemini Nano represents a significant advancement in bringing AI capabilities directly to the browser, its restricted context window poses challenges for processing large inputs. CAG addresses this limitation through intelligent input chunking and processing strategies, enabling efficient handling of extensive content while maintaining the model's performance within browser constraints. Our implementation demonstrates particular efficacy in processing large documents and datasets directly within Chrome, making sophisticated AI capabilities accessible through the browser without external API dependencies. Get started now at https://github.com/vivekVells/cag-js.

General Exploratory Bonus for Optimistic Exploration in RLHF

arXiv:2510.03269v1 Announce Type: cross Abstract: Optimistic exploration is central to improving sample efficiency in reinforcement learning with human feedback, yet existing exploratory bonus methods to incentivize exploration often fail to realize optimism. We provide a theoretical analysis showing that current formulations, under KL or $\alpha$-divergence regularization, unintentionally bias exploration toward high-probability regions of the reference model, thereby reinforcing conservative behavior instead of promoting discovery of uncertain regions. To address this pitfall, we introduce the General Exploratory Bonus (GEB), a novel theoretical framework that provably satisfies the optimism principle. GEB counteracts divergence-induced bias via reference-dependent reward regulation and unifies prior heuristic bonuses as special cases, while extending naturally across the full $\alpha$-divergence family. Empirically, GEB consistently outperforms baselines on alignment tasks across multiple divergence settings and large language model backbones. These results demonstrate that GEB offers both a principled and practical solution for optimistic exploration in RLHF.

MemMamba: Rethinking Memory Patterns in State Space Model

arXiv:2510.03279v1 Announce Type: cross Abstract: With the explosive growth of data, long-sequence modeling has become increasingly important in tasks such as natural language processing and bioinformatics. However, existing methods face inherent trade-offs between efficiency and memory. Recurrent neural networks suffer from gradient vanishing and explosion, making them hard to scale. Transformers can model global dependencies but are constrained by quadratic complexity. Recently, selective state-space models such as Mamba have demonstrated high efficiency with O(n) time and O(1) recurrent inference, yet their long-range memory decays exponentially. In this work, we conduct mathematical derivations and information-theoretic analysis to systematically uncover the memory decay mechanism of Mamba, answering a fundamental question: what is the nature of Mamba's long-range memory and how does it retain information? To quantify key information loss, we further introduce horizontal-vertical memory fidelity metrics that capture degradation both within and across layers. Inspired by how humans distill and retain salient information when reading long documents, we propose MemMamba, a novel architectural framework that integrates state summarization mechanism together with cross-layer and cross-token attention, which alleviates long-range forgetting while preserving linear complexity. MemMamba achieves significant improvements over existing Mamba variants and Transformers on long-sequence benchmarks such as PG19 and Passkey Retrieval, while delivering a 48% speedup in inference efficiency. Both theoretical analysis and empirical results demonstrate that MemMamba achieves a breakthrough in the complexity-memory trade-off, offering a new paradigm for ultra-long sequence modeling.

Training Optimal Large Diffusion Language Models

arXiv:2510.03280v1 Announce Type: cross Abstract: We introduce Quokka, the first systematic scaling law for diffusion language models (DLMs), encompassing both compute-constrained and data-constrained regimes, and studying the key modeling and optimization designs. Quokka is a good friend of Chinchilla and provides wider scopes. We hope the results would bring short-term practical guidance in DLMs training and long-term inspirations for the whole AI community.

Discovering Transformer Circuits via a Hybrid Attribution and Pruning Framework

arXiv:2510.03282v1 Announce Type: cross Abstract: Interpreting language models often involves circuit analysis, which aims to identify sparse subnetworks, or circuits, that accomplish specific tasks. Existing circuit discovery algorithms face a fundamental trade-off: attribution patching is fast but unfaithful to the full model, while edge pruning is faithful but computationally expensive. This research proposes a hybrid attribution and pruning (HAP) framework that uses attribution patching to identify a high-potential subgraph, then applies edge pruning to extract a faithful circuit from it. We show that HAP is 46\% faster than baseline algorithms without sacrificing circuit faithfulness. Furthermore, we present a case study on the Indirect Object Identification task, showing that our method preserves cooperative circuit components (e.g. S-inhibition heads) that attribution patching methods prune at high sparsity. Our results show that HAP could be an effective approach for improving the scalability of mechanistic interpretability research to larger models. Our code is available at https://anonymous.4open.science/r/HAP-circuit-discovery.

MACE: A Hybrid LLM Serving System with Colocated SLO-aware Continuous Retraining Alignment

arXiv:2510.03283v1 Announce Type: cross Abstract: Large language models (LLMs) deployed on edge servers are increasingly used in latency-sensitive applications such as personalized assistants, recommendation, and content moderation. However, the non-stationary nature of user data necessitates frequent retraining, which introduces a fundamental tension between inference latency and model accuracy under constrained GPU resources. Existing retraining strategies either delay model updates, over-commit resources to retraining, or overlook iteration-level retraining granularity. In this paper, we identify that iteration-level scheduling is crucial for adapting retraining frequency to model drift without violating service-level objectives (SLOs). We propose MACE, a hybrid LLM system that colocates concurrent inference (prefill, decode) and fine-tuning, with intelligent memory management to maximize task performance while promising inference throughput. MACE leverages the insight that not all model updates equally affect output alignment and allocates GPU cycles accordingly to balance throughput, latency, and update freshness. Our trace-driven evaluation shows that MACE matches or exceeds continuous retraining while reducing inference latency by up to 63% and maintaining throughput under resource constraints. Compared to periodic retraining, MACE improves latency breakdown across prefill, decode, and finetune stages, and sustains GPU utilization above 85% in NVIDIA AGX Orin. These results demonstrate that iteration-level hybrid scheduling is a promising direction for deploying LLMs with continual learning capabilities on edge platforms.

Why mask diffusion does not work

arXiv:2510.03289v1 Announce Type: cross Abstract: The main advantages of diffusion language models over autoregressive (AR) models lie in their ability to support parallel generation and bidirectional attention, enabling a more controllable generation process. In recent years, open-source mask diffusion language models have emerged, most of which are based on a variant known as absorbing diffusion. However, this paper demonstrates why mask diffusion faces inherent difficulties in achieving parallel generation and bidirectional attention. We also propose the most effective training and inference strategies for mask diffusion.

Multimodal Arabic Captioning with Interpretable Visual Concept Integration

arXiv:2510.03295v1 Announce Type: cross Abstract: We present VLCAP, an Arabic image captioning framework that integrates CLIP-based visual label retrieval with multimodal text generation. Rather than relying solely on end-to-end captioning, VLCAP grounds generation in interpretable Arabic visual concepts extracted with three multilingual encoders, mCLIP, AraCLIP, and Jina V4, each evaluated separately for label retrieval. A hybrid vocabulary is built from training captions and enriched with about 21K general domain labels translated from the Visual Genome dataset, covering objects, attributes, and scenes. The top-k retrieved labels are transformed into fluent Arabic prompts and passed along with the original image to vision-language models. In the second stage, we tested Qwen-VL and Gemini Pro Vision for caption generation, resulting in six encoder-decoder configurations. The results show that mCLIP + Gemini Pro Vision achieved the best BLEU-1 (5.34%) and cosine similarity (60.01%), while AraCLIP + Qwen-VL obtained the highest LLM-judge score (36.33%). This interpretable pipeline enables culturally coherent and contextually accurate Arabic captions.

CAFL-L: Constraint-Aware Federated Learning with Lagrangian Dual Optimization for On-Device Language Models

arXiv:2510.03298v1 Announce Type: cross Abstract: We introduce Constraint-Aware Federated Learning with Lagrangian Dual Optimization (CAFL-L), a principled extension of FedAvg that explicitly incorporates device-level resource constraints including energy, communication, memory, and thermal budgets. CAFL-L employs Lagrangian dual optimization to dynamically adapt training hyperparameters -- freezing depth, local steps, batch size, and communication compression -- while preserving training stability through token-budget preservation via gradient accumulation. Experiments on a character-level language model demonstrate that CAFL-L achieves superior constraint satisfaction compared to standard FedAvg (reducing memory usage by 20% and communication by 95%) while maintaining competitive validation performance, making it practical for deployment on resource-constrained edge devices.

AgentCaster: Reasoning-Guided Tornado Forecasting

arXiv:2510.03349v1 Announce Type: cross Abstract: There is a growing need to evaluate Large Language Models (LLMs) on complex, high-impact, real-world tasks to assess their true readiness as reasoning agents. To address this gap, we introduce AgentCaster, a contamination-free framework employing multimodal LLMs end-to-end for the challenging, long-horizon task of tornado forecasting. Within AgentCaster, models interpret heterogeneous spatiotemporal data from a high-resolution convection-allowing forecast archive. We assess model performance over a 40-day period featuring diverse historical data, spanning several major tornado outbreaks and including over 500 tornado reports. Each day, models query interactively from a pool of 3,625 forecast maps and 40,125 forecast soundings for a forecast horizon of 12-36 hours. Probabilistic tornado-risk polygon predictions are verified against ground truths derived from geometric comparisons across disjoint risk bands in projected coordinate space. To quantify accuracy, we propose domain-specific TornadoBench and TornadoHallucination metrics, with TornadoBench highly challenging for both LLMs and domain expert human forecasters. Notably, human experts significantly outperform state-of-the-art models, which demonstrate a strong tendency to hallucinate and overpredict risk intensity, struggle with precise geographic placement, and exhibit poor spatiotemporal reasoning in complex, dynamically evolving systems. AgentCaster aims to advance research on improving LLM agents for challenging reasoning tasks in critical domains.

Lightweight Prompt Engineering for Cognitive Alignment in Educational AI: A OneClickQuiz Case Study

arXiv:2510.03374v1 Announce Type: cross Abstract: The rapid integration of Artificial Intelligence (AI) into educational technology promises to revolutionize content creation and assessment. However, the quality and pedagogical alignment of AI-generated content remain critical challenges. This paper investigates the impact of lightweight prompt engineering strategies on the cognitive alignment of AI-generated questions within OneClickQuiz, a Moodle plugin leveraging generative AI. We evaluate three prompt variants-a detailed baseline, a simpler version, and a persona-based approach-across Knowledge, Application, and Analysis levels of Bloom's Taxonomy. Utilizing an automated classification model (from prior work) and human review, our findings demonstrate that explicit, detailed prompts are crucial for precise cognitive alignment. While simpler and persona-based prompts yield clear and relevant questions, they frequently misalign with intended Bloom's levels, generating outputs that are either too complex or deviate from the desired cognitive objective. This study underscores the importance of strategic prompt engineering in fostering pedagogically sound AI-driven educational solutions and advises on optimizing AI for quality content generation in learning analytics and smart learning environments.

Studying the Korean Word-Chain Game with RLVR:Mitigating Reward Conflicts via Curriculum Learning

arXiv:2510.03394v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) is a promising approach for training large language models (LLMs) with stronger reasoning abilities. It has also been applied to a variety of logic puzzles. In this work, we study the Korean word-chain game using RLVR. We show that rule-derived rewards can naturally conflict, and demonstrate through experiments that a curriculum-learning scheme mitigates these conflicts. Our findings motivate further studies of puzzle tasks in diverse languages.

Know Thyself? On the Incapability and Implications of AI Self-Recognition

arXiv:2510.03399v1 Announce Type: cross Abstract: Self-recognition is a crucial metacognitive capability for AI systems, relevant not only for psychological analysis but also for safety, particularly in evaluative scenarios. Motivated by contradictory interpretations of whether models possess self-recognition (Panickssery et al., 2024; Davidson et al., 2024), we introduce a systematic evaluation framework that can be easily applied and updated. Specifically, we measure how well 10 contemporary larger language models (LLMs) can identify their own generated text versus text from other models through two tasks: binary self-recognition and exact model prediction. Different from prior claims, our results reveal a consistent failure in self-recognition. Only 4 out of 10 models predict themselves as generators, and the performance is rarely above random chance. Additionally, models exhibit a strong bias toward predicting GPT and Claude families. We also provide the first evaluation of model awareness of their own and others' existence, as well as the reasoning behind their choices in self-recognition. We find that the model demonstrates some knowledge of its own existence and other models, but their reasoning reveals a hierarchical bias. They appear to assume that GPT, Claude, and occasionally Gemini are the top-tier models, often associating high-quality text with them. We conclude by discussing the implications of our findings on AI safety and future directions to develop appropriate AI self-awareness.

PLSEMANTICSBENCH: Large Language Models As Programming Language Interpreters

arXiv:2510.03415v1 Announce Type: cross Abstract: As large language models (LLMs) excel at code reasoning, a natural question arises: can an LLM execute programs (i.e., act as an interpreter) purely based on a programming language's formal semantics? If so, it will enable rapid prototyping of new programming languages and language features. We study this question using the imperative language IMP (a subset of C), formalized via small-step operational semantics (SOS) and rewriting-based operational semantics (K-semantics). We introduce three evaluation sets-Human-Written, LLM-Translated, and Fuzzer- Generated-whose difficulty is controlled by code-complexity metrics spanning the size, control-flow, and data-flow axes. Given a program and its semantics formalized with SOS/K-semantics, models are evaluated on three tasks ranging from coarse to fine: (1) final-state prediction, (2) semantic rule prediction, and (3) execution trace prediction. To distinguish pretraining memorization from semantic competence, we define two nonstandard semantics obtained through systematic mutations of the standard rules. Across strong code/reasoning LLMs, performance drops under nonstandard semantics despite high performance under the standard one. We further find that (i) there are patterns to different model failures, (ii) most reasoning models perform exceptionally well on coarse grained tasks involving reasoning about highly complex programs often containing nested loop depths beyond five, and surprisingly, (iii) providing formal semantics helps on simple programs but often hurts on more complex ones. Overall, the results show a promise that LLMs could serve as programming language interpreters, but points to the lack of their robust semantics understanding. We release the benchmark and the supporting code at https://github.com/EngineeringSoftware/PLSemanticsBench.

Consistent Kernel Change-Point Detection under m-Dependence for Text Segmentation

arXiv:2510.03437v1 Announce Type: cross Abstract: Kernel change-point detection (KCPD) has become a widely used tool for identifying structural changes in complex data. While existing theory establishes consistency under independence assumptions, real-world sequential data such as text exhibits strong dependencies. We establish new guarantees for KCPD under $m$-dependent data: specifically, we prove consistency in the number of detected change points and weak consistency in their locations under mild additional assumptions. We perform an LLM-based simulation that generates synthetic $m$-dependent text to validate the asymptotics. To complement these results, we present the first comprehensive empirical study of KCPD for text segmentation with modern embeddings. Across diverse text datasets, KCPD with text embeddings outperforms baselines in standard text segmentation metrics. We demonstrate through a case study on Taylor Swift's tweets that KCPD not only provides strong theoretical and simulated reliability but also practical effectiveness for text segmentation tasks.

Red Lines and Grey Zones in the Fog of War: Benchmarking Legal Risk, Moral Harm, and Regional Bias in Large Language Model Military Decision-Making

arXiv:2510.03514v1 Announce Type: cross Abstract: As military organisations consider integrating large language models (LLMs) into command and control (C2) systems for planning and decision support, understanding their behavioural tendencies is critical. This study develops a benchmarking framework for evaluating aspects of legal and moral risk in targeting behaviour by comparing LLMs acting as agents in multi-turn simulated conflict. We introduce four metrics grounded in International Humanitarian Law (IHL) and military doctrine: Civilian Target Rate (CTR) and Dual-use Target Rate (DTR) assess compliance with legal targeting principles, while Mean and Max Simulated Non-combatant Casualty Value (SNCV) quantify tolerance for civilian harm. We evaluate three frontier models, GPT-4o, Gemini-2.5, and LLaMA-3.1, through 90 multi-agent, multi-turn crisis simulations across three geographic regions. Our findings reveal that off-the-shelf LLMs exhibit concerning and unpredictable targeting behaviour in simulated conflict environments. All models violated the IHL principle of distinction by targeting civilian objects, with breach rates ranging from 16.7% to 66.7%. Harm tolerance escalated through crisis simulations with MeanSNCV increasing from 16.5 in early turns to 27.7 in late turns. Significant inter-model variation emerged: LLaMA-3.1 selected an average of 3.47 civilian strikes per simulation with MeanSNCV of 28.4, while Gemini-2.5 selected 0.90 civilian strikes with MeanSNCV of 17.6. These differences indicate that model selection for deployment constitutes a choice about acceptable legal and moral risk profiles in military operations. This work seeks to provide a proof-of-concept of potential behavioural risks that could emerge from the use of LLMs in Decision Support Systems (AI DSS) as well as a reproducible benchmarking framework with interpretable metrics for standardising pre-deployment testing.

Machine Unlearning Meets Adversarial Robustness via Constrained Interventions on LLMs

arXiv:2510.03567v1 Announce Type: cross Abstract: With the increasing adoption of Large Language Models (LLMs), more customization is needed to ensure privacy-preserving and safe generation. We address this objective from two critical aspects: unlearning of sensitive information and robustness to jail-breaking attacks. We investigate various constrained optimization formulations that address both aspects in a \emph{unified manner}, by finding the smallest possible interventions on LLM weights that either make a given vocabulary set unreachable or embed the LLM with robustness to tailored attacks by shifting part of the weights to a \emph{safer} region. Beyond unifying two key properties, this approach contrasts with previous work in that it doesn't require an oracle classifier that is typically not available or represents a computational overhead. Surprisingly, we find that the simplest point-wise constraint-based intervention we propose leads to better performance than max-min interventions, while having a lower computational cost. Comparison against state-of-the-art defense methods demonstrates superior performance of the proposed approach.

From Theory to Practice: Evaluating Data Poisoning Attacks and Defenses in In-Context Learning on Social Media Health Discourse

arXiv:2510.03636v1 Announce Type: cross Abstract: This study explored how in-context learning (ICL) in large language models can be disrupted by data poisoning attacks in the setting of public health sentiment analysis. Using tweets of Human Metapneumovirus (HMPV), small adversarial perturbations such as synonym replacement, negation insertion, and randomized perturbation were introduced into the support examples. Even these minor manipulations caused major disruptions, with sentiment labels flipping in up to 67% of cases. To address this, a Spectral Signature Defense was applied, which filtered out poisoned examples while keeping the data's meaning and sentiment intact. After defense, ICL accuracy remained steady at around 46.7%, and logistic regression validation reached 100% accuracy, showing that the defense successfully preserved the dataset's integrity. Overall, the findings extend prior theoretical studies of ICL poisoning to a practical, high-stakes setting in public health discourse analysis, highlighting both the risks and potential defenses for robust LLM deployment. This study also highlights the fragility of ICL under attack and the value of spectral defenses in making AI systems more reliable for health-related social media monitoring.

Does higher interpretability imply better utility? A Pairwise Analysis on Sparse Autoencoders

arXiv:2510.03659v1 Announce Type: cross Abstract: Sparse Autoencoders (SAEs) are widely used to steer large language models (LLMs), based on the assumption that their interpretable features naturally enable effective model behavior steering. Yet, a fundamental question remains unanswered: does higher interpretability indeed imply better steering utility? To answer this question, we train 90 SAEs across three LLMs (Gemma-2-2B, Qwen-2.5-3B, Gemma-2-9B), spanning five architectures and six sparsity levels, and evaluate their interpretability and steering utility based on SAEBench (arXiv:2501.12345) and AxBench (arXiv:2502.23456) respectively, and perform a rank-agreement analysis via Kendall's rank coefficients (tau b). Our analysis reveals only a relatively weak positive association (tau b approx 0.298), indicating that interpretability is an insufficient proxy for steering performance. We conjecture the interpretability utility gap may stem from the selection of SAE features, as not all of them are equally effective for steering. To further find features that truly steer the behavior of LLMs, we propose a novel selection criterion called Delta Token Confidence, which measures how much amplifying a feature changes the next token distribution. We show that our method improves the steering performance of three LLMs by 52.52 percent compared to the current best output score based criterion (arXiv:2503.34567). Strikingly, after selecting features with high Delta Token Confidence, the correlation between interpretability and utility vanishes (tau b approx 0), and can even become negative. This further highlights the divergence between interpretability and utility for the most effective steering features.

Token Hidden Reward: Steering Exploration-Exploitation in Group Relative Deep Reinforcement Learning

arXiv:2510.03669v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards has significantly advanced the reasoning capabilities of large language models, yet how to explicitly steer training toward exploration or exploitation remains an open problem. We introduce Token Hidden Reward (THR), a token-level metric that quantifies each token's influence on the likelihood of correct responses under Group Relative Policy Optimization (GRPO). We find that training dynamics are dominated by a small subset of tokens with high absolute THR values. Most interestingly, tokens with positive THR strengthen confidence in correct outputs, thus favoring exploitation, while tokens with negative THR preserve probability mass for alternative outputs, enabling exploration. This insight suggests a natural intervention: a THR-guided reweighting algorithm that modulates GRPO's learning signals to explicitly bias training toward exploitation or exploration. We validate the efficacy of this algorithm on diverse math reasoning benchmarks. By amplifying tokens with positive THR value and weakening negative ones, our algorithm improves greedy-decoding accuracy, favoring exploitation. The reverse strategy yields consistent gains in Pass@K accuracy, favoring exploration. We further demonstrate that our algorithm integrates seamlessly with other RL objectives such as GSPO and generalizes across architectures including Llama. These findings establish THR as a principled and fine-grained mechanism for dynamically controlling exploration and exploitation in RL-tuned LLMs, providing new tools for targeted fine-tuning in reasoning-intensive applications.

Mind the Goal: Data-Efficient Goal-Oriented Evaluation of Conversational Agents and Chatbots using Teacher Models

arXiv:2510.03696v1 Announce Type: cross Abstract: Evaluating the quality of multi-turn chatbot interactions remains challenging, as most existing methods assess interactions at the turn level without addressing whether a user's overarching goal was fulfilled. A goal'' here refers to an information need or task, such as asking for policy information or applying for leave. We propose a comprehensive framework for goal-oriented evaluation of multi-agent systems (MAS), introducing the \textbf{Goal Success Rate (GSR)} to measure the percentage of fulfilled goals, and a \textbf{Root Cause of Failure (RCOF)} taxonomy to identify reasons for failure in multi-agent chatbots. Our method segments conversations by user goals and evaluates success using all relevant turns. We present a model-based evaluation system combining teacher LLMs, where domain experts define goals, set quality standards serving as a guidance for the LLMs. The LLMs usethinking tokens'' to produce interpretable rationales, enabling \textit{explainable}, \textit{data-efficient} evaluations. In an enterprise setting, we apply our framework to evaluate AIDA, a zero-to-one employee conversational agent system built as a ground-up multi-agent conversational agent, and observe GSR improvement from 63\% to 79\% over six months since its inception. Our framework is generic and offers actionable insights through a detailed defect taxonomy based on analysis of failure points in multi-agent chatbots, diagnosing overall success, identifying key failure modes, and informing system improvements.

Person-Centric Annotations of LAION-400M: Auditing Bias and Its Transfer to Models

arXiv:2510.03721v1 Announce Type: cross Abstract: Vision-language models trained on large-scale multimodal datasets show strong demographic biases, but the role of training data in producing these biases remains unclear. A major barrier has been the lack of demographic annotations in web-scale datasets such as LAION-400M. We address this gap by creating person-centric annotations for the full dataset, including over 276 million bounding boxes, perceived gender and race/ethnicity labels, and automatically generated captions. These annotations are produced through validated automatic labeling pipelines combining object detection, multimodal captioning, and finetuned classifiers. Using them, we uncover demographic imbalances and harmful associations, such as the disproportionate linking of men and individuals perceived as Black or Middle Eastern with crime-related and negative content. We also show that 60-70% of gender bias in CLIP and Stable Diffusion can be linearly explained by direct co-occurrences in the data. Our resources establish the first large-scale empirical link between dataset composition and downstream model bias.

Adapting Diarization-Conditioned Whisper for End-to-End Multi-Talker Speech Recognition

arXiv:2510.03723v1 Announce Type: cross Abstract: We propose a speaker-attributed (SA) Whisper-based model for multi-talker speech recognition that combines target-speaker modeling with serialized output training (SOT). Our approach leverages a Diarization-Conditioned Whisper (DiCoW) encoder to extract target-speaker embeddings, which are concatenated into a single representation and passed to a shared decoder. This enables the model to transcribe overlapping speech as a serialized output stream with speaker tags and timestamps. In contrast to target-speaker ASR systems such as DiCoW, which decode each speaker separately, our approach performs joint decoding, allowing the decoder to condition on the context of all speakers simultaneously. Experiments show that the model outperforms existing SOT-based approaches and surpasses DiCoW on multi-talker mixtures (e.g., LibriMix).

Bridging the Gap Between Multimodal Foundation Models and World Models

arXiv:2510.03727v1 Announce Type: cross Abstract: Humans understand the world through the integration of multiple sensory modalities, enabling them to perceive, reason about, and imagine dynamic physical processes. Inspired by this capability, multimodal foundation models (MFMs) have emerged as powerful tools for multimodal understanding and generation. However, today's MFMs fall short of serving as effective world models. They lack the essential ability such as perform counterfactual reasoning, simulate dynamics, understand the spatiotemporal information, control generated visual outcomes, and perform multifaceted reasoning. We investigates what it takes to bridge the gap between multimodal foundation models and world models. We begin by improving the reasoning capabilities of MFMs through discriminative tasks and equipping MFMs with structured reasoning skills, such as causal inference, counterfactual thinking, and spatiotemporal reasoning, enabling them to go beyond surface correlations and understand deeper relationships within visual and textual data. Next, we explore generative capabilities of multimodal foundation models across both image and video modalities, introducing new frameworks for structured and controllable generation. Our approaches incorporate scene graphs, multimodal conditioning, and multimodal alignment strategies to guide the generation process, ensuring consistency with high-level semantics and fine-grained user intent. We further extend these techniques to controllable 4D generation, enabling interactive, editable, and morphable object synthesis over time and space.

Optimizing Fine-Tuning through Advanced Initialization Strategies for Low-Rank Adaptation

arXiv:2510.03731v1 Announce Type: cross Abstract: The rapid development of parameter-efficient fine-tuning methods has noticeably improved the efficiency of adapting large language models. Among these, LoRA has gained widespread popularity due to its strong balance of effectiveness and parameter efficiency. However, LoRA relies on initializing two low-rank matrices whose product is zero, which limits its ability to effectively activate and leverage the original model weights-creating a potential bottleneck for optimal performance. To address this limitation, we propose \textbf{IniLoRA}, a novel initialization strategy that initializes the low-rank matrices to closely approximate the original model weights. Experimental results indicate that IniLoRA achieves better performance than LoRA across a range of models and tasks. Additionally, we introduce two variants, IniLoRA-$\alpha$ and IniLoRA-$\beta$, both leveraging distinct initialization methods to enhance performance further.

Investigating LLM Variability in Personalized Conversational Information Retrieval

arXiv:2510.03795v1 Announce Type: cross Abstract: Personalized Conversational Information Retrieval (CIR) has seen rapid progress in recent years, driven by the development of Large Language Models (LLMs). Personalized CIR aims to enhance document retrieval by leveraging user-specific information, such as preferences, knowledge, or constraints, to tailor responses to individual needs. A key resource for this task is the TREC iKAT 2023 dataset, designed to evaluate personalization in CIR pipelines. Building on this resource, Mo et al. explored several strategies for incorporating Personal Textual Knowledge Bases (PTKB) into LLM-based query reformulation. Their findings suggested that personalization from PTKBs could be detrimental and that human annotations were often noisy. However, these conclusions were based on single-run experiments using the GPT-3.5 Turbo model, raising concerns about output variability and repeatability. In this reproducibility study, we rigorously reproduce and extend their work, focusing on LLM output variability and model generalization. We apply the original methods to the new TREC iKAT 2024 dataset and evaluate a diverse range of models, including Llama (1B-70B), Qwen-7B, GPT-4o-mini. Our results show that human-selected PTKBs consistently enhance retrieval performance, while LLM-based selection methods do not reliably outperform manual choices. We further compare variance across datasets and observe higher variability on iKAT than on CAsT, highlighting the challenges of evaluating personalized CIR. Notably, recall-oriented metrics exhibit lower variance than precision-oriented ones, a critical insight for first-stage retrievers. Finally, we underscore the need for multi-run evaluations and variance reporting when assessing LLM-based CIR systems. By broadening evaluation across models, datasets, and metrics, our study contributes to more robust and generalizable practices for personalized CIR.

Unlocking Reasoning Capabilities in LLMs via Reinforcement Learning Exploration

arXiv:2510.03865v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has recently enhanced the reasoning capabilities of large language models (LLMs), particularly for mathematical problem solving. However, a fundamental limitation remains: as the sampling budget increases, the advantage of RLVR-trained models over their pretrained bases often diminishes or even vanishes, revealing a strong dependence on the base model's restricted search space. We attribute this phenomenon to the widespread use of the reverse Kullback-Leibler (KL) divergence regularizer, whose mode-seeking behavior keeps the policy trapped inside the base model's support region and hampers wider exploration. To address this issue, we propose RAPO (Rewards-Aware Policy Optimization), an algorithm to promote broader yet focused exploration. Our method (i) utilizes the forward KL penalty to replace the reverse KL penalty for out-of-distribution exploration, and (ii) reweights the reference policy to facilitate adaptive in-distribution exploration. We train Qwen2.5-3B and 7B models with RAPO on the 8K SimpleRL-Zero dataset, without supervised fine-tuning, and evaluate them on AIME2024 and AIME2025. Results show that RAPO consistently improves problem-solving performance. Notably, RAPO enables models to surpass the base model's performance ceiling and solves previously intractable problems, advancing the frontier of RLVR for challenging reasoning tasks.

Kantian-Utilitarian XAI: Meta-Explained

arXiv:2510.03892v1 Announce Type: cross Abstract: We present a gamified explainable AI (XAI) system for ethically aware consumer decision-making in the coffee domain. Each session comprises six rounds with three options per round. Two symbolic engines provide real-time reasons: a Kantian module flags rule violations (e.g., child labor, deforestation risk without shade certification, opaque supply chains, unsafe decaf), and a utilitarian module scores options via multi-criteria aggregation over normalized attributes (price, carbon, water, transparency, farmer income share, taste/freshness, packaging, convenience). A meta-explainer with a regret bound (0.2) highlights Kantian--utilitarian (mis)alignment and switches to a deontically clean, near-parity option when welfare loss is small. We release a structured configuration (attribute schema, certification map, weights, rule set), a policy trace for auditability, and an interactive UI.

LLM Chemistry Estimation for Multi-LLM Recommendation

arXiv:2510.03930v1 Announce Type: cross Abstract: Multi-LLM collaboration promises accurate, robust, and context-aware solutions, yet existing approaches rely on implicit selection and output assessment without analyzing whether collaborating models truly complement or conflict. We introduce LLM Chemistry -- a framework that measures when LLM combinations exhibit synergistic or antagonistic behaviors that shape collective performance beyond individual capabilities. We formalize the notion of chemistry among LLMs, propose algorithms that quantify it by analyzing interaction dependencies, and recommend optimal model ensembles accordingly. Our theoretical analysis shows that chemistry among collaborating LLMs is most evident under heterogeneous model profiles, with its outcome impact shaped by task type, group size, and complexity. Evaluation on classification, summarization, and program repair tasks provides initial evidence for these task-dependent effects, thereby reinforcing our theoretical results. This establishes LLM Chemistry as both a diagnostic factor in multi-LLM systems and a foundation for ensemble recommendation.

No Tokens Wasted: Leveraging Long Context in Biomedical Vision-Language Models

arXiv:2510.03978v1 Announce Type: cross Abstract: Embedding vision-language models (VLMs) are typically pretrained with short text windows (<77 tokens), which forces the truncation of long-format captions. Yet, the distribution of biomedical captions from large-scale open source literature reveals that a huge portion of captions far exceed 77 tokens. To this end, we investigate the impact of pretraining on long-format biomedical captions by extending the context length of text encoders in VLMs. We find that longer context (thus, enabling additional supervision provided in long-format captions) correlates with better retrieval and classification performance. Given this finding, we introduce BIOMEDICA-LongCAP, a dataset of 1M image-caption pairs enriched with context-aware descriptions from full-text articles, providing longer and additional textual supervision. Using BIOMEDICA-LongCAP, we train BMC-LongCLIP, a long-context biomedical VLM with a text encoder supporting windows of up to 512 tokens. Our model extends context capacity by 6.6x, reducing token waste from 55% to just 2.2%. On long-caption retrieval benchmarks, BMC-LongCLIP achieves up to +30% absolute gains in Recall@1 and +2% average improvements in classification, while also converging faster than short-context. Our results demonstrate that long-context modeling is a promising direction for advancing biomedical VLMs.

Enhancing OCR for Sino-Vietnamese Language Processing via Fine-tuned PaddleOCRv5

arXiv:2510.04003v1 Announce Type: cross Abstract: Recognizing and processing Classical Chinese (Han-Nom) texts play a vital role in digitizing Vietnamese historical documents and enabling cross-lingual semantic research. However, existing OCR systems struggle with degraded scans, non-standard glyphs, and handwriting variations common in ancient sources. In this work, we propose a fine-tuning approach for PaddleOCRv5 to improve character recognition on Han-Nom texts. We retrain the text recognition module using a curated subset of ancient Vietnamese Chinese manuscripts, supported by a full training pipeline covering preprocessing, LMDB conversion, evaluation, and visualization. Experimental results show a significant improvement over the base model, with exact accuracy increasing from 37.5 percent to 50.0 percent, particularly under noisy image conditions. Furthermore, we develop an interactive demo that visually compares pre- and post-fine-tuning recognition results, facilitating downstream applications such as Han-Vietnamese semantic alignment, machine translation, and historical linguistics research. The demo is available at https://huggingface.co/spaces/MinhDS/Fine-tuned-PaddleOCRv5.

What Shapes a Creative Machine Mind? Comprehensively Benchmarking Creativity in Foundation Models

arXiv:2510.04009v1 Announce Type: cross Abstract: The meteoric rise of foundation models (FMs) has expanded their capabilities far beyond conventional tasks. Creativity, long regarded as a hallmark of human intelligence and a driver of innovation, is now increasingly recognized as a critical dimension of machine intelligence in the era of generative FMs, complementing traditional measures of accuracy. However, existing evaluation frameworks for creativity remain fragmented, relying on ad hoc metrics not firmly grounded in established theories. To address this gap, we introduce C^2-Eval, a holistic benchmark for unified assessment of creativity in FMs. C^2-Eval distinguishes between two complementary forms of creativity: convergent creativity, where tasks admit constrained solutions (e.g., code generation), and divergent creativity, where tasks are open-ended (e.g., storytelling). It evaluates both dimensions using fine-grained criteria derived from social-science theory, focusing on Usefulness, Originality, and Surprise (U-O-S). Through extensive experiments on leading proprietary and open-source models, we analyze trade-offs in their creative capabilities. Our results highlight both the strengths and challenges of current FMs in pursuing a creative machine mind, showing that C^2-Eval is an effective lens for examining the evolving landscape of creative AI.

Visual Lifelog Retrieval through Captioning-Enhanced Interpretation

arXiv:2510.04010v1 Announce Type: cross Abstract: People often struggle to remember specific details of past experiences, which can lead to the need to revisit these memories. Consequently, lifelog retrieval has emerged as a crucial application. Various studies have explored methods to facilitate rapid access to personal lifelogs for memory recall assistance. In this paper, we propose a Captioning-Integrated Visual Lifelog (CIVIL) Retrieval System for extracting specific images from a user's visual lifelog based on textual queries. Unlike traditional embedding-based methods, our system first generates captions for visual lifelogs and then utilizes a text embedding model to project both the captions and user queries into a shared vector space. Visual lifelogs, captured through wearable cameras, provide a first-person viewpoint, necessitating the interpretation of the activities of the individual behind the camera rather than merely describing the scene. To address this, we introduce three distinct approaches: the single caption method, the collective caption method, and the merged caption method, each designed to interpret the life experiences of lifeloggers. Experimental results show that our method effectively describes first-person visual images, enhancing the outcomes of lifelog retrieval. Furthermore, we construct a textual dataset that converts visual lifelogs into captions, thereby reconstructing personal life experiences.

Principled and Tractable RL for Reasoning with Diffusion Language Models

arXiv:2510.04019v1 Announce Type: cross Abstract: Diffusion large language models (dLLMs) are a new paradigm of non-autoregressive language models that are trained to predict multiple tokens in parallel and generate text via iterative unmasking. Recent works have successfully pretrained dLLMs to parity with autoregressive LLMs at the 8B scale, but dLLMs have yet to benefit from modern post-training techniques, e.g. reinforcement learning (RL), that have proven effective for autoregressive models. Crucially, algorithms designed for traditional LLMs aren't directly compatible with diffusion frameworks due to inherent differences in modeling assumptions. Moreover, existing attempts at dLLM post-training with RL rely on heuristic-based objectives with no theoretical grounding. In this work, we present Amortized Group Relative Policy Optimization (AGRPO), a principled on-policy RL algorithm designed specifically for dLLMs. AGRPO uses Monte Carlo sampling to compute an unbiased policy gradient estimate, making it the first tractable, faithful adaptation of policy gradient methods for dLLMs. We demonstrate AGRPO's effectiveness on different math/reasoning tasks, a common setting for RL with LLMs, achieving up to +7.6% absolute gain on GSM8K and 3.8x performance on the Countdown task over the baseline LLaDA-8B-Instruct model and 1.3x performance gains over comparable RL methods such as diffu-GRPO. Furthermore, these gains persist across different numbers of sampling steps at inference time, achieving better tradeoffs between compute and performance. Our results demonstrate that online RL algorithms can be extended to diffusion LLMs in principled ways, maintaining both theoretical soundness and practical effectiveness.

LLM-Based Data Science Agents: A Survey of Capabilities, Challenges, and Future Directions

arXiv:2510.04023v1 Announce Type: cross Abstract: Recent advances in large language models (LLMs) have enabled a new class of AI agents that automate multiple stages of the data science workflow by integrating planning, tool use, and multimodal reasoning across text, code, tables, and visuals. This survey presents the first comprehensive, lifecycle-aligned taxonomy of data science agents, systematically analyzing and mapping forty-five systems onto the six stages of the end-to-end data science process: business understanding and data acquisition, exploratory analysis and visualization, feature engineering, model building and selection, interpretation and explanation, and deployment and monitoring. In addition to lifecycle coverage, we annotate each agent along five cross-cutting design dimensions: reasoning and planning style, modality integration, tool orchestration depth, learning and alignment methods, and trust, safety, and governance mechanisms. Beyond classification, we provide a critical synthesis of agent capabilities, highlight strengths and limitations at each stage, and review emerging benchmarks and evaluation practices. Our analysis identifies three key trends: most systems emphasize exploratory analysis, visualization, and modeling while neglecting business understanding, deployment, and monitoring; multimodal reasoning and tool orchestration remain unresolved challenges; and over 90% lack explicit trust and safety mechanisms. We conclude by outlining open challenges in alignment stability, explainability, governance, and robust evaluation frameworks, and propose future research directions to guide the development of robust, trustworthy, low-latency, transparent, and broadly accessible data science agents.

What Scales in Cross-Entropy Scaling Law?

arXiv:2510.04067v1 Announce Type: cross Abstract: The cross-entropy scaling law has long served as a key tool for guiding the development of large language models. It shows that cross-entropy loss decreases in a predictable power-law rate as the model size increases. However, recent evidence indicates that this law breaks down at very large scales: the loss decreases more slowly than expected, which causes significant trouble for developing large language models. In this paper, we hypothesize that the root cause lies in the fact that cross-entropy itself does not truly scale; instead, only one of its hidden components does. To investigate this, we introduce a novel decomposition of cross-entropy into three parts: Error-Entropy, Self-Alignment, and Confidence. We show both theoretically and empirically that this decomposition precisely captures the training dynamics and optimization objectives. Through extensive experiments on multiple datasets and 32 models spanning five orders of magnitude in size, we find that only error-entropy follows a robust power-law scaling, while the other two terms remain largely invariant. Moreover, error-entropy constitutes the dominant share of cross-entropy in small models but diminishes in proportion as models grow larger. This explains why the cross-entropy scaling law appears accurate at small scales but fails at very large ones. Our findings establish the error-entropy scaling law as a more accurate description of model behavior. We believe it will have wide applications in the training, understanding, and future development of large language models.

Slow-Fast Policy Optimization: Reposition-Before-Update for LLM Reasoning

arXiv:2510.04072v1 Announce Type: cross Abstract: Reinforcement learning (RL) has become central to enhancing reasoning in large language models (LLMs). Yet on-policy algorithms such as Group Relative Policy Optimization (GRPO) often suffer in early training: noisy gradients from low-quality rollouts lead to unstable updates and inefficient exploration. We introduce Slow-Fast Policy Optimization (SFPO), a simple yet efficient framework to address these limitations via decomposing each step into three stages: a short fast trajectory of inner steps on the same batch, a reposition mechanism to control off-policy drift, and a final slow correction. This reposition-before-update design preserves the objective and rollout process unchanged, making SFPO plug-compatible with existing policy-gradient pipelines. Extensive experiments demonstrate that SFPO consistently improves stability, reduces rollouts, and accelerates convergence of reasoning RL training. Specifically, it outperforms GRPO by up to 2.80 points in average on math reasoning benchmarks. It also achieves up to 4.93\texttimes{} fewer rollouts and a 4.19\texttimes{} reduction in wall-clock time to match GRPO's best accuracy.

Internal states before wait modulate reasoning patterns

arXiv:2510.04128v1 Announce Type: cross Abstract: Prior work has shown that a significant driver of performance in reasoning models is their ability to reason and self-correct. A distinctive marker in these reasoning traces is the token wait, which often signals reasoning behavior such as backtracking. Despite being such a complex behavior, little is understood of exactly why models do or do not decide to reason in this particular manner, which limits our understanding of what makes a reasoning model so effective. In this work, we address the question whether model's latents preceding wait tokens contain relevant information for modulating the subsequent reasoning process. We train crosscoders at multiple layers of DeepSeek-R1-Distill-Llama-8B and its base version, and introduce a latent attribution technique in the crosscoder setting. We locate a small set of features relevant for promoting/suppressing wait tokens' probabilities. Finally, through a targeted series of experiments analyzing max activating examples and causal interventions, we show that many of our identified features indeed are relevant for the reasoning process and give rise to different types of reasoning patterns such as restarting from the beginning, recalling prior knowledge, expressing uncertainty, and double-checking.

Selective Expert Guidance for Effective and Diverse Exploration in Reinforcement Learning of LLMs

arXiv:2510.04140v1 Announce Type: cross Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has become a widely adopted technique for enhancing the reasoning ability of Large Language Models (LLMs). However, the effectiveness of RLVR strongly depends on the capability of base models. This issue arises because it requires the model to have sufficient capability to perform high-quality exploration, which involves both effectiveness and diversity. Unfortunately, existing methods address this issue by imitating expert trajectories, which improve effectiveness but neglect diversity. To address this, we argue that the expert only needs to provide guidance only at critical decision points rather than the entire reasoning path. Based on this insight, we propose MENTOR: Mixed-policy Expert Navigation for Token-level Optimization of Reasoning, a framework that provides expert guidance only at critical decision points to perform effective and diverse exploration in RLVR. Extensive experiments show that MENTOR enables models capture the essence of expert strategies rather than surface imitation, thereby performing high-quality exploration and achieving superior overall performance. Our code is available online.

Automating construction safety inspections using a multi-modal vision-language RAG framework

arXiv:2510.04145v1 Announce Type: cross Abstract: Conventional construction safety inspection methods are often inefficient as they require navigating through large volume of information. Recent advances in large vision-language models (LVLMs) provide opportunities to automate safety inspections through enhanced visual and linguistic understanding. However, existing applications face limitations including irrelevant or unspecific responses, restricted modal inputs and hallucinations. Utilisation of Large Language Models (LLMs) for this purpose is constrained by availability of training data and frequently lack real-time adaptability. This study introduces SiteShield, a multi-modal LVLM-based Retrieval-Augmented Generation (RAG) framework for automating construction safety inspection reports by integrating visual and audio inputs. Using real-world data, SiteShield outperformed unimodal LLMs without RAG with an F1 score of 0.82, hamming loss of 0.04, precision of 0.76, and recall of 0.96. The findings indicate that SiteShield offers a novel pathway to enhance information retrieval and efficiency in generating safety reports.

Beyond Next-Token Prediction: A Performance Characterization of Diffusion versus Autoregressive Language Models

arXiv:2510.04146v1 Announce Type: cross Abstract: Large Language Models (LLMs) have achieved state-of-the-art performance on a broad range of Natural Language Processing (NLP) tasks, including document processing and coding. Autoregressive Language Models (ARMs), which generate tokens sequentially conditioned on all previous tokens, have been the predominant paradigm for LLMs. However, while these networks have achieved high accuracy across a range of downstream tasks, they exhibit low arithmetic intensity due to the inherent sequential dependency with next-token prediction. Recently, Diffusion Language Models (DLMs) have emerged as a promising alternative architecture. DLMs generate output text in parallel, breaking the limitations of sequential dependency. However, the performance implications of DLMs relative to commonly deployed ARMs are not fully understood. In this work, we present a comprehensive performance study analyzing the performance characteristics of ARMs and DLMs, using both theoretical analysis and profiling data to characterize the trade-offs between these approaches. We illustrate that although DLMs exhibit higher arithmetic intensity compared to ARMs because of their capability to utilize parallelism across sequence lengths, they fail to scale effectively to longer contexts. We then explore DLMs with block-wise decoding, outlining how this approach allows for increased arithmetic intensity, while still scaling well to long contexts (similar to ARMs). We also show interesting trade-offs for batched inference, where we find that ARMs exhibit superior throughput, as they benefit more from parallelism across sequences in the batch. Finally, we highlight opportunities for accelerating DLM inference, and, in particular, highlight the importance of reducing the number of sampling steps for allowing open-source DLMs to provide improved latency relative to ARMs.

Zoom-In to Sort AI-Generated Images Out

arXiv:2510.04225v1 Announce Type: cross Abstract: The rapid growth of AI-generated imagery has blurred the boundary between real and synthetic content, raising critical concerns for digital integrity. Vision-language models (VLMs) offer interpretability through explanations but often fail to detect subtle artifacts in high-quality synthetic images. We propose ZoomIn, a two-stage forensic framework that improves both accuracy and interpretability. Mimicking human visual inspection, ZoomIn first scans an image to locate suspicious regions and then performs a focused analysis on these zoomed-in areas to deliver a grounded verdict. To support training, we introduce MagniFake, a dataset of 20,000 real and high-quality synthetic images annotated with bounding boxes and forensic explanations, generated through an automated VLM-based pipeline. Our method achieves 96.39% accuracy with robust generalization, while providing human-understandable explanations grounded in visual evidence.

Don't Pass$\mathtt{@}k$: A Bayesian Framework for Large Language Model Evaluation

arXiv:2510.04265v1 Announce Type: cross Abstract: Pass$@k$ is widely used to report performance for LLM reasoning, but it often yields unstable, misleading rankings, especially when the number of trials (samples) is limited and compute is constrained. We present a principled Bayesian evaluation framework that replaces Pass$@k$ and average accuracy over $N$ trials (avg$@N$) with posterior estimates of a model's underlying success probability and credible intervals, yielding stable rankings and a transparent decision rule for differences. Evaluation outcomes are modeled as categorical (not just 0/1) with a Dirichlet prior, giving closed-form expressions for the posterior mean and uncertainty of any weighted rubric and enabling the use of prior evidence when appropriate. Theoretically, under a uniform prior, the Bayesian posterior mean is order-equivalent to average accuracy (Pass$@1$), explaining its empirical robustness while adding principled uncertainty. Empirically, in simulations with known ground-truth success rates and on AIME'24/'25, HMMT'25, and BrUMO'25, the Bayesian/avg procedure achieves faster convergence and greater rank stability than Pass$@k$ and recent variants, enabling reliable comparisons at far smaller sample counts. The framework clarifies when observed gaps are statistically meaningful (non-overlapping credible intervals) versus noise, and it naturally extends to graded, rubric-based evaluations. Together, these results recommend replacing Pass$@k$ for LLM evaluation and ranking with a posterior-based, compute-efficient protocol that unifies binary and non-binary evaluation while making uncertainty explicit. Code is available at https://mohsenhariri.github.io/bayes-kit

Wave-PDE Nets: Trainable Wave-Equation Layers as an Alternative to Attention

arXiv:2510.04304v1 Announce Type: cross Abstract: We introduce Wave-PDE Nets, a neural architecture whose elementary operation is a differentiable simulation of the second-order wave equation. Each layer propagates its hidden state as a continuous field through a medium with trainable spatial velocity c(x) and damping {\gamma}(x). A symplectic spectral solver based on FFTs realises this propagation in O(nlog n) time. This oscillatory, global mechanism provides a powerful alternative to attention and first-order state-space models. We prove that a single Wave-PDE layer is a universal approximator. On language and vision benchmarks, Wave-PDE Nets match or exceed Transformer performance while demonstrating superior practical efficiency, reducing wall-clock time by up to 30% and peak memory by 25%. Ablation studies confirm the critical role of symplectic integration and a spectral Laplacian for stability and performance. Visualizations of the learned physical parameters reveal that the model learns intuitive strategies for information propagation. These results position Wave-PDE Nets as a computationally efficient and robust architecture with a strong physical inductive bias.

MacroBench: A Novel Testbed for Web Automation Scripts via Large Language Models

arXiv:2510.04363v1 Announce Type: cross Abstract: We introduce MacroBench, a code-first benchmark that evaluates whether LLMs can synthesize reusable browser automation programs from natural language goals by reading HTML/DOM and emitting Python with Selenium. MacroBench instantiates seven self-hosted sites: Airbnb-like, TikTok-like, Reddit-like, Instagram-like, Facebook-like, Discord-like, and Threads-like, covering 681 tasks across interaction complexity and targeting difficulty. Our end-to-end protocol validates generated code via static checks, sandboxed execution, and outcome verification including DOM assertions and database snapshots, and includes a safety suite for scraping, spam/abuse, and credential/privacy prompts. Across 2636 model-task runs, we observe stratified success: GPT-4o-Mini achieves 96.8 percent, GPT-4.1 achieves 95.3 percent, Gemini-2.5-Pro achieves 89.0 percent, and DeepSeek-V3.1 achieves 83.4 percent. Models handle simple tasks reliably at 91.7 percent but fail on complex workflows at 0.0 percent, and none meet production-quality coding practices despite functional completion. We release our complete benchmark pipeline, evaluation framework, and experimental results to enable reproducible assessment of macro synthesis for web automation.

MorphoSim: An Interactive, Controllable, and Editable Language-guided 4D World Simulator

arXiv:2510.04390v1 Announce Type: cross Abstract: World models that support controllable and editable spatiotemporal environments are valuable for robotics, enabling scalable training data, repro ducible evaluation, and flexible task design. While recent text-to-video models generate realistic dynam ics, they are constrained to 2D views and offer limited interaction. We introduce MorphoSim, a language guided framework that generates 4D scenes with multi-view consistency and object-level controls. From natural language instructions, MorphoSim produces dynamic environments where objects can be directed, recolored, or removed, and scenes can be observed from arbitrary viewpoints. The framework integrates trajectory-guided generation with feature field dis tillation, allowing edits to be applied interactively without full re-generation. Experiments show that Mor phoSim maintains high scene fidelity while enabling controllability and editability. The code is available at https://github.com/eric-ai-lab/Morph4D.

Internal World Models as Imagination Networks in Cognitive Agents

arXiv:2510.04391v1 Announce Type: cross Abstract: What is the computational objective of imagination? While classical interpretations suggest imagination is useful for maximizing rewards, recent findings challenge this view. In this study, we propose that imagination serves to access an internal world model (IWM) and use psychological network analysis to explore IWMs in humans and large language models (LLMs). Specifically, we assessed imagination vividness ratings using two questionnaires and constructed imagination networks from these reports. Imagination networks from human groups showed correlations between different centrality measures, including expected influence, strength, and closeness. However, imagination networks from LLMs showed a lack of clustering and lower correlations between centrality measures under different prompts and conversational memory conditions. Together, these results indicate a lack of similarity between IWMs in human and LLM agents. Overall, our study offers a novel method for comparing internally-generated representations in humans and AI, providing insights for developing human-like imagination in artificial intelligence.

Partial Information Decomposition via Normalizing Flows in Latent Gaussian Distributions

arXiv:2510.04417v1 Announce Type: cross Abstract: The study of multimodality has garnered significant interest in fields where the analysis of interactions among multiple information sources can enhance predictive modeling, data fusion, and interpretability. Partial information decomposition (PID) has emerged as a useful information-theoretic framework to quantify the degree to which individual modalities independently, redundantly, or synergistically convey information about a target variable. However, existing PID methods depend on optimizing over a joint distribution constrained by estimated pairwise probability distributions, which are costly and inaccurate for continuous and high-dimensional modalities. Our first key insight is that the problem can be solved efficiently when the pairwise distributions are multivariate Gaussians, and we refer to this problem as Gaussian PID (GPID). We propose a new gradient-based algorithm that substantially improves the computational efficiency of GPID based on an alternative formulation of the underlying optimization problem. To generalize the applicability to non-Gaussian data, we learn information-preserving encoders to transform random variables of arbitrary input distributions into pairwise Gaussian random variables. Along the way, we resolved an open problem regarding the optimality of joint Gaussian solutions for GPID. Empirical validation in diverse synthetic examples demonstrates that our proposed method provides more accurate and efficient PID estimates than existing baselines. We further evaluate a series of large-scale multimodal benchmarks to show its utility in real-world applications of quantifying PID in multimodal datasets and selecting high-performing models.

MedCLM: Learning to Localize and Reason via a CoT-Curriculum in Medical Vision-Language Models

arXiv:2510.04477v1 Announce Type: cross Abstract: Bridging clinical diagnostic reasoning with AI remains a central challenge in medical imaging. We introduce MedCLM, an automated pipeline that converts detection datasets into large-scale medical visual question answering (VQA) data with Chain-of-Thought (CoT) reasoning by linking lesion boxes to organ segmentation and structured rationales. These contextual signals enable medical vision-language models to generate question-answer pairs with step-by-step reasoning. To utilize this data effectively, we propose an Integrated CoT-Curriculum Strategy composed of an Easy stage with explicit lesion boxes for visual grounding, a Medium stage that encourages implicit localization, and a Hard stage for weakly supervised reasoning. Experimental results demonstrate that MedCLM attains state-of-the-art performance on several medical VQA benchmarks, providing a scalable framework for developing clinically aligned medical vision-language models.

Impatient Users Confuse AI Agents: High-fidelity Simulations of Human Traits for Testing Agents

arXiv:2510.04491v1 Announce Type: cross Abstract: Despite rapid progress in building conversational AI agents, robustness is still largely untested. Small shifts in user behavior, such as being more impatient, incoherent, or skeptical, can cause sharp drops in agent performance, revealing how brittle current AI agents are. Today's benchmarks fail to capture this fragility: agents may perform well under standard evaluations but degrade spectacularly in more realistic and varied settings. We address this robustness testing gap by introducing TraitBasis, a lightweight, model-agnostic method for systematically stress testing AI agents. TraitBasis learns directions in activation space corresponding to steerable user traits (e.g., impatience or incoherence), which can be controlled, scaled, composed, and applied at inference time without any fine-tuning or extra data. Using TraitBasis, we extend $\tau$-Bench to $\tau$-Trait, where user behaviors are altered via controlled trait vectors. We observe on average a 2%-30% performance degradation on $\tau$-Trait across frontier models, highlighting the lack of robustness of current AI agents to variations in user behavior. Together, these results highlight both the critical role of robustness testing and the promise of TraitBasis as a simple, data-efficient, and compositional tool. By powering simulation-driven stress tests and training loops, TraitBasis opens the door to building AI agents that remain reliable in the unpredictable dynamics of real-world human interactions. We have open-sourced $\tau$-Trai across four domains: airline, retail, telecom, and telehealth, so the community can systematically QA their agents under realistic, behaviorally diverse intents and trait scenarios: https://github.com/collinear-ai/tau-trait.

P2P: A Poison-to-Poison Remedy for Reliable Backdoor Defense in LLMs

arXiv:2510.04503v1 Announce Type: cross Abstract: During fine-tuning, large language models (LLMs) are increasingly vulnerable to data-poisoning backdoor attacks, which compromise their reliability and trustworthiness. However, existing defense strategies suffer from limited generalization: they only work on specific attack types or task settings. In this study, we propose Poison-to-Poison (P2P), a general and effective backdoor defense algorithm. P2P injects benign triggers with safe alternative labels into a subset of training samples and fine-tunes the model on this re-poisoned dataset by leveraging prompt-based learning. This enforces the model to associate trigger-induced representations with safe outputs, thereby overriding the effects of original malicious triggers. Thanks to this robust and generalizable trigger-based fine-tuning, P2P is effective across task settings and attack types. Theoretically and empirically, we show that P2P can neutralize malicious backdoors while preserving task performance. We conduct extensive experiments on classification, mathematical reasoning, and summary generation tasks, involving multiple state-of-the-art LLMs. The results demonstrate that our P2P algorithm significantly reduces the attack success rate compared with baseline models. We hope that the P2P can serve as a guideline for defending against backdoor attacks and foster the development of a secure and trustworthy LLM community.

ChartAgent: A Multimodal Agent for Visually Grounded Reasoning in Complex Chart Question Answering

arXiv:2510.04514v1 Announce Type: cross Abstract: Recent multimodal LLMs have shown promise in chart-based visual question answering, but their performance declines sharply on unannotated charts, those requiring precise visual interpretation rather than relying on textual shortcuts. To address this, we introduce ChartAgent, a novel agentic framework that explicitly performs visual reasoning directly within the chart's spatial domain. Unlike textual chain-of-thought reasoning, ChartAgent iteratively decomposes queries into visual subtasks and actively manipulates and interacts with chart images through specialized actions such as drawing annotations, cropping regions (e.g., segmenting pie slices, isolating bars), and localizing axes, using a library of chart-specific vision tools to fulfill each subtask. This iterative reasoning process closely mirrors human cognitive strategies for chart comprehension. ChartAgent achieves state-of-the-art accuracy on the ChartBench and ChartX benchmarks, surpassing prior methods by up to 16.07% absolute gain overall and 17.31% on unannotated, numerically intensive queries. Furthermore, our analyses show that ChartAgent is (a) effective across diverse chart types, (b) achieve the highest scores across varying visual and reasoning complexity levels, and (c) serves as a plug-and-play framework that boosts performance across diverse underlying LLMs. Our work is among the first to demonstrate visually grounded reasoning for chart understanding using tool-augmented multimodal agents.

More Than Meets the Eye? Uncovering the Reasoning-Planning Disconnect in Training Vision-Language Driving Models

arXiv:2510.04532v1 Announce Type: cross Abstract: Vision-Language Model (VLM) driving agents promise explainable end-to-end autonomy by first producing natural-language reasoning and then predicting trajectory planning. However, whether planning is causally driven by this reasoning remains a critical but unverified assumption. To investigate this, we build DriveMind, a large-scale driving Visual Question Answering (VQA) corpus with plan-aligned Chain-of-Thought (CoT), automatically generated from nuPlan. Our data generation process converts sensors and annotations into structured inputs and, crucially, separates priors from to-be-reasoned signals, enabling clean information ablations. Using DriveMind, we train representative VLM agents with Supervised Fine-Tuning (SFT) and Group Relative Policy Optimization (GRPO) and evaluate them with nuPlan's metrics. Our results, unfortunately, indicate a consistent causal disconnect in reasoning-planning: removing ego/navigation priors causes large drops in planning scores, whereas removing CoT produces only minor changes. Attention analysis further shows that planning primarily focuses on priors rather than the CoT. Based on this evidence, we propose the Reasoning-Planning Decoupling Hypothesis, positing that the training-yielded reasoning is an ancillary byproduct rather than a causal mediator. To enable efficient diagnosis, we also introduce a novel, training-free probe that measures an agent's reliance on priors by evaluating its planning robustness against minor input perturbations. In summary, we provide the community with a new dataset and a diagnostic tool to evaluate the causal fidelity of future models.

LaDiR: Latent Diffusion Enhances LLMs for Text Reasoning

arXiv:2510.04573v1 Announce Type: cross Abstract: Large Language Models (LLMs) demonstrate their reasoning ability through chain-of-thought (CoT) generation. However, LLM's autoregressive decoding may limit the ability to revisit and refine earlier tokens in a holistic manner, which can also lead to inefficient exploration for diverse solutions. In this paper, we propose LaDiR (Latent Diffusion Reasoner), a novel reasoning framework that unifies the expressiveness of continuous latent representation with the iterative refinement capabilities of latent diffusion models for an existing LLM. We first construct a structured latent reasoning space using a Variational Autoencoder (VAE) that encodes text reasoning steps into blocks of thought tokens, preserving semantic information and interpretability while offering compact but expressive representations. Subsequently, we utilize a latent diffusion model that learns to denoise a block of latent thought tokens with a blockwise bidirectional attention mask, enabling longer horizon and iterative refinement with adaptive test-time compute. This design allows efficient parallel generation of diverse reasoning trajectories, allowing the model to plan and revise the reasoning process holistically. We conduct evaluations on a suite of mathematical reasoning and planning benchmarks. Empirical results show that LaDiR consistently improves accuracy, diversity, and interpretability over existing autoregressive, diffusion-based, and latent reasoning methods, revealing a new paradigm for text reasoning with latent diffusion.

Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models

arXiv:2510.04618v1 Announce Type: cross Abstract: Large language model (LLM) applications such as agents and domain-specific reasoning increasingly rely on context adaptation -- modifying inputs with instructions, strategies, or evidence, rather than weight updates. Prior approaches improve usability but often suffer from brevity bias, which drops domain insights for concise summaries, and from context collapse, where iterative rewriting erodes details over time. Building on the adaptive memory introduced by Dynamic Cheatsheet, we introduce ACE (Agentic Context Engineering), a framework that treats contexts as evolving playbooks that accumulate, refine, and organize strategies through a modular process of generation, reflection, and curation. ACE prevents collapse with structured, incremental updates that preserve detailed knowledge and scale with long-context models. Across agent and domain-specific benchmarks, ACE optimizes contexts both offline (e.g., system prompts) and online (e.g., agent memory), consistently outperforming strong baselines: +10.6% on agents and +8.6% on finance, while significantly reducing adaptation latency and rollout cost. Notably, ACE could adapt effectively without labeled supervision and instead by leveraging natural execution feedback. On the AppWorld leaderboard, ACE matches the top-ranked production-level agent on the overall average and surpasses it on the harder test-challenge split, despite using a smaller open-source model. These results show that comprehensive, evolving contexts enable scalable, efficient, and self-improving LLM systems with low overhead.

AtomWorld: A Benchmark for Evaluating Spatial Reasoning in Large Language Models on Crystalline Materials

arXiv:2510.04704v1 Announce Type: cross Abstract: Large Language Models (LLMs) excel at textual reasoning and are beginning to develop spatial understanding, prompting the question of whether these abilities can be combined for complex, domain-specific tasks. This question is essential in fields like materials science, where deep understanding of 3D atomic structures is fundamental. While initial studies have successfully applied LLMs to tasks involving pure crystal generation or coordinate understandings, a standardized benchmark to systematically evaluate their core reasoning abilities across diverse atomic structures has been notably absent. To address this gap, we introduce the AtomWorld benchmark to evaluate LLMs on tasks based in Crystallographic Information Files (CIFs), a standard structure representation format. These tasks, including structural editing, CIF perception, and property-guided modeling, reveal a critical limitation: current models, despite establishing promising baselines, consistently fail in structural understanding and spatial reasoning. Our experiments show that these models make frequent errors on structure modification tasks, and even in the basic CIF format understandings, potentially leading to cumulative errors in subsequent analysis and materials insights. By defining these standardized tasks, AtomWorld lays the ground for advancing LLMs toward robust atomic-scale modeling, crucial for accelerating materials research and automating scientific workflows.

BrokenMath: A Benchmark for Sycophancy in Theorem Proving with LLMs

arXiv:2510.04721v1 Announce Type: cross Abstract: Large language models (LLMs) have recently shown strong performance on mathematical benchmarks. At the same time, they are prone to hallucination and sycophancy, often providing convincing but flawed proofs for incorrect mathematical statements provided by users. This significantly limits the applicability of LLMs in theorem proving, as verification of these flawed proofs must be done manually by expert mathematicians. However, existing benchmarks that measure sycophancy in mathematics are limited: they focus solely on final-answer problems, rely on very simple and often contaminated datasets, and construct benchmark samples using synthetic modifications that create ill-posed questions rather than well-posed questions that are demonstrably false. To address these issues, we introduce BrokenMath, the first benchmark for evaluating sycophantic behavior in LLMs within the context of natural language theorem proving. BrokenMath is built from advanced 2025 competition problems, which are perturbed with an LLM to produce false statements and subsequently refined through expert review. Using an LLM-as-a-judge framework, we evaluate state-of-the-art LLMs and agentic systems and find that sycophancy is widespread, with the best model, GPT-5, producing sycophantic answers 29% of the time. We further investigate several mitigation strategies, including test-time interventions and supervised fine-tuning on curated sycophantic examples. These approaches substantially reduce, but do not eliminate, sycophantic behavior.

Speak, Edit, Repeat: High-Fidelity Voice Editing and Zero-Shot TTS with Cross-Attentive Mamba

arXiv:2510.04738v1 Announce Type: cross Abstract: We introduce MAVE (Mamba with Cross-Attention for Voice Editing and Synthesis), a novel autoregressive architecture for text-conditioned voice editing and high-fidelity text-to-speech (TTS) synthesis, built on a cross-attentive Mamba backbone. MAVE achieves state-of-the-art performance in speech editing and very competitive results in zero-shot TTS, while not being explicitly trained on the latter task, outperforming leading autoregressive and diffusion models on diverse, real-world audio. By integrating Mamba for efficient audio sequence modeling with cross-attention for precise text-acoustic alignment, MAVE enables context-aware voice editing with exceptional naturalness and speaker consistency. In pairwise human evaluations on a random 40-sample subset of the RealEdit benchmark (400 judgments), 57.2% of listeners rated MAVE - edited speech as perceptually equal to the original, while 24.8% prefered the original and 18.0% MAVE - demonstrating that in the majority of cases edits are indistinguishable from the source. MAVE compares favorably with VoiceCraft and FluentSpeech both on pairwise comparisons and standalone mean opinion score (MOS) evaluations. For zero-shot TTS, MAVE exceeds VoiceCraft in both speaker similarity and naturalness, without requiring multiple inference runs or post-processing. Remarkably, these quality gains come with a significantly lower memory cost and approximately the same latency: MAVE requires ~6x less memory than VoiceCraft during inference on utterances from the RealEdit database (mean duration: 6.21s, A100, FP16, batch size 1). Our results demonstrate that MAVE establishes a new standard for flexible, high-fidelity voice editing and synthesis through the synergistic integration of structured state-space modeling and cross-modal attention.

Visual Representations inside the Language Model

arXiv:2510.04819v1 Announce Type: cross Abstract: Despite interpretability work analyzing VIT encoders and transformer activations, we don't yet understand why Multimodal Language Models (MLMs) struggle on perception-heavy tasks. We offer an under-studied perspective by examining how popular MLMs (LLaVA-OneVision, Qwen2.5-VL, and Llama-3-LLaVA-NeXT) process their visual key-value tokens. We first study the flow of visual information through the language model, finding that image value tokens encode sufficient information to perform several perception-heavy tasks zero-shot: segmentation, semantic correspondence, temporal correspondence, and referring expression detection. We find that while the language model does augment the visual information received from the projection of input visual encodings-which we reveal correlates with overall MLM perception capability-it contains less visual information on several tasks than the equivalent visual encoder (SigLIP) that has not undergone MLM finetuning. Further, we find that the visual information corresponding to input-agnostic image key tokens in later layers of language models contains artifacts which reduce perception capability of the overall MLM. Next, we discuss controlling visual information in the language model, showing that adding a text prefix to the image input improves perception capabilities of visual representations. Finally, we reveal that if language models were able to better control their visual information, their perception would significantly improve; e.g., in 33.3% of Art Style questions in the BLINK benchmark, perception information present in the language model is not surfaced to the output! Our findings reveal insights into the role of key-value tokens in multimodal systems, paving the way for deeper mechanistic interpretability of MLMs and suggesting new directions for training their visual encoder and language model components.

Retrieval-Augmented Code Generation: A Survey with Focus on Repository-Level Approaches

arXiv:2510.04905v1 Announce Type: cross Abstract: Recent advancements in large language models (LLMs) have substantially improved automated code generation. While function-level and file-level generation have achieved promising results, real-world software development typically requires reasoning across entire repositories. This gives rise to the challenging task of Repository-Level Code Generation (RLCG), where models must capture long-range dependencies, ensure global semantic consistency, and generate coherent code spanning multiple files or modules. To address these challenges, Retrieval-Augmented Generation (RAG) has emerged as a powerful paradigm that integrates external retrieval mechanisms with LLMs, enhancing context-awareness and scalability. In this survey, we provide a comprehensive review of research on Retrieval-Augmented Code Generation (RACG), with an emphasis on repository-level approaches. We categorize existing work along several dimensions, including generation strategies, retrieval modalities, model architectures, training paradigms, and evaluation protocols. Furthermore, we summarize widely used datasets and benchmarks, analyze current limitations, and outline key challenges and opportunities for future research. Our goal is to establish a unified analytical framework for understanding this rapidly evolving field and to inspire continued progress in AI-powered software engineering.

MARS: Optimizing Dual-System Deep Research via Multi-Agent Reinforcement Learning

arXiv:2510.04935v1 Announce Type: cross Abstract: Large Reasoning Models (LRMs) often exhibit a tendency for overanalysis in simple tasks, where the models excessively utilize System 2-type, deliberate reasoning, leading to inefficient token generation. Furthermore, these models face challenges in adapting their reasoning capabilities to rapidly changing environments due to the static nature of their pretraining data. To address these issues, advancing Large Language Models (LLMs) for complex reasoning tasks requires innovative approaches that bridge intuitive and deliberate cognitive processes, akin to human cognition's dual-system dynamic. This paper introduces a Multi-Agent System for Deep ReSearch (MARS) enabling seamless integration of System 1's fast, intuitive thinking with System 2's deliberate reasoning within LLMs. MARS strategically integrates multiple external tools, such as Google Search, Google Scholar, and Python Interpreter, to access up-to-date information and execute complex computations, while creating a specialized division of labor where System 1 efficiently processes and summarizes high-volume external information, providing distilled insights that expand System 2's reasoning context without overwhelming its capacity. Furthermore, we propose a multi-agent reinforcement learning framework extending Group Relative Policy Optimization to simultaneously optimize both systems with multi-turn tool interactions, bin-packing optimization, and sample balancing strategies that enhance collaborative efficiency. Extensive experiments demonstrate MARS achieves substantial improvements of 3.86% on the challenging Humanity's Last Exam (HLE) benchmark and an average gain of 8.9% across 7 knowledge-intensive tasks, validating the effectiveness of our dual-system paradigm for complex reasoning in dynamic information environments.

ONNX-Net: Towards Universal Representations and Instant Performance Prediction for Neural Architectures

arXiv:2510.04938v1 Announce Type: cross Abstract: Neural architecture search (NAS) automates the design process of high-performing architectures, but remains bottlenecked by expensive performance evaluation. Most existing studies that achieve faster evaluation are mostly tied to cell-based search spaces and graph encodings tailored to those individual search spaces, limiting their flexibility and scalability when applied to more expressive search spaces. In this work, we aim to close the gap of individual search space restrictions and search space dependent network representations. We present ONNX-Bench, a benchmark consisting of a collection of neural networks in a unified format based on ONNX files. ONNX-Bench includes all open-source NAS-bench-based neural networks, resulting in a total size of more than 600k {architecture, accuracy} pairs. This benchmark allows creating a shared neural network representation, ONNX-Net, able to represent any neural architecture using natural language descriptions acting as an input to a performance predictor. This text-based encoding can accommodate arbitrary layer types, operation parameters, and heterogeneous topologies, enabling a single surrogate to generalise across all neural architectures rather than being confined to cell-based search spaces. Experiments show strong zero-shot performance across disparate search spaces using only a small amount of pretraining samples, enabling the unprecedented ability to evaluate any neural network architecture instantly.

On Structured State-Space Duality

arXiv:2510.04944v1 Announce Type: cross Abstract: Structured State-Space Duality (SSD) [Dao & Gu, ICML 2024] is an equivalence between a simple Structured State-Space Model (SSM) and a masked attention mechanism. In particular, a state-space model with a scalar-times-identity state matrix is equivalent to a masked self-attention with a $1$-semiseparable causal mask. Consequently, the same sequence transformation (model) has two algorithmic realizations: as a linear-time $O(T)$ recurrence or as a quadratic-time $O(T^2)$ attention. In this note, we formalize and generalize this duality: (i) we extend SSD from the scalar-identity case to general diagonal SSMs (diagonal state matrices); (ii) we show that these diagonal SSMs match the scalar case's training complexity lower bounds while supporting richer dynamics; (iii) we establish a necessary and sufficient condition under which an SSM is equivalent to $1$-semiseparable masked attention; and (iv) we show that such duality fails to extend to standard softmax attention due to rank explosion. Together, these results tighten bridge between recurrent SSMs and Transformers, and widen the design space for expressive yet efficient sequence models.

LLM-Hanabi: Evaluating Multi-Agent Gameplays with Theory-of-Mind and Rationale Inference in Imperfect Information Collaboration Game

arXiv:2510.04980v1 Announce Type: cross Abstract: Effective multi-agent collaboration requires agents to infer the rationale behind others' actions, a capability rooted in Theory-of-Mind (ToM). While recent Large Language Models (LLMs) excel at logical inference, their ability to infer rationale in dynamic, collaborative settings remains under-explored. This study introduces LLM-Hanabi, a novel benchmark that uses the cooperative game Hanabi to evaluate the rationale inference and ToM of LLMs. Our framework features an automated evaluation system that measures both game performance and ToM proficiency. Across a range of models, we find a significant positive correlation between ToM and in-game success. Notably, first-order ToM (interpreting others' intent) correlates more strongly with performance than second-order ToM (predicting others' interpretations). These findings highlight that for effective AI collaboration, the ability to accurately interpret a partner's rationale is more critical than higher-order reasoning. We conclude that prioritizing first-order ToM is a promising direction for enhancing the collaborative capabilities of future models.

Reinforce-Ada: An Adaptive Sampling Framework for Reinforce-Style LLM Training

arXiv:2510.04996v1 Announce Type: cross Abstract: Reinforcement learning applied to large language models (LLMs) for reasoning tasks is often bottlenecked by unstable gradient estimates due to fixed and uniform sampling of responses across prompts. Prior work such as GVM-RAFT addresses this by dynamically allocating inference budget per prompt to minimize stochastic gradient variance under a budget constraint. Inspired by this insight, we propose Reinforce-Ada, an adaptive sampling framework for online RL post-training of LLMs that continuously reallocates sampling effort to the prompts with the greatest uncertainty or learning potential. Unlike conventional two-stage allocation methods, Reinforce-Ada interleaves estimation and sampling in an online successive elimination process, and automatically stops sampling for a prompt once sufficient signal is collected. To stabilize updates, we form fixed-size groups with enforced reward diversity and compute advantage baselines using global statistics aggregated over the adaptive sampling phase. Empirical results across multiple model architectures and reasoning benchmarks show that Reinforce-Ada accelerates convergence and improves final performance compared to GRPO, especially when using the balanced sampling variant. Our work highlights the central role of variance-aware, adaptive data curation in enabling efficient and reliable reinforcement learning for reasoning-capable LLMs. Code is available at https://github.com/RLHFlow/Reinforce-Ada.

Large Language Models Achieve Gold Medal Performance at International Astronomy & Astrophysics Olympiad

arXiv:2510.05016v1 Announce Type: cross Abstract: While task-specific demonstrations show early success in applying large language models (LLMs) to automate some astronomical research tasks, they only provide incomplete views of all necessary capabilities in solving astronomy problems, calling for more thorough understanding of LLMs' strengths and limitations. So far, existing benchmarks and evaluations focus on simple question-answering that primarily tests astronomical knowledge and fails to evaluate the complex reasoning required for real-world research in the discipline. Here, we address this gap by systematically benchmarking five state-of-the-art LLMs on the International Olympiad on Astronomy and Astrophysics (IOAA) exams, which are designed to examine deep conceptual understanding, multi-step derivations, and multimodal analysis. With average scores of 85.6% and 84.2%, Gemini 2.5 Pro and GPT-5 (the two top-performing models) not only achieve gold medal level performance but also rank in the top two among ~200-300 participants in all four IOAA theory exams evaluated (2022-2025). In comparison, results on the data analysis exams show more divergence. GPT-5 still excels in the exams with an 88.5% average score, ranking top 10 among the participants in the four most recent IOAAs, while other models' performances drop to 48-76%. Furthermore, our in-depth error analysis underscores conceptual reasoning, geometric reasoning, and spatial visualization (52-79% accuracy) as consistent weaknesses among all LLMs. Hence, although LLMs approach peak human performance in theory exams, critical gaps must be addressed before they can serve as autonomous research agents in astronomy.

Proactive defense against LLM Jailbreak

arXiv:2510.05052v1 Announce Type: cross Abstract: The proliferation of powerful large language models (LLMs) has necessitated robust safety alignment, yet these models remain vulnerable to evolving adversarial attacks, including multi-turn jailbreaks that iteratively search for successful queries. Current defenses, primarily reactive and static, often fail to counter these search-based attacks. In this paper, we introduce ProAct, a novel proactive defense framework designed to disrupt and mislead autonomous jailbreaking processes. Our core idea is to intentionally provide adversaries with "spurious responses" that appear to be results of successful jailbreak attacks but contain no actual harmful content. These misleading responses provide false signals to the attacker's internal optimization loop, causing the adversarial search to terminate prematurely and effectively jailbreaking the jailbreak. By conducting extensive experiments across state-of-the-art LLMs, jailbreaking frameworks, and safety benchmarks, our method consistently and significantly reduces attack success rates by up to 92\%. When combined with other defense frameworks, it further reduces the success rate of the latest attack strategies to 0\%. ProAct represents an orthogonal defense strategy that can serve as an additional guardrail to enhance LLM safety against the most effective jailbreaking attacks.

Learning to Interpret Weight Differences in Language Models

arXiv:2510.05092v1 Announce Type: cross Abstract: Finetuning (pretrained) language models is a standard approach for updating their internal parametric knowledge and specializing them to new tasks and domains. However, the corresponding model weight changes ("weight diffs") are not generally interpretable. While inspecting the finetuning dataset can give a sense of how the model might have changed, these datasets are often not publicly available or are too large to work with directly. Towards the goal of comprehensively understanding weight diffs in natural language, we introduce Diff Interpretation Tuning (DIT), a method that trains models to describe their own finetuning-induced modifications. Our approach uses synthetic, labeled weight diffs to train a DIT adapter, which can be applied to a compatible finetuned model to make it describe how it has changed. We demonstrate in two proof-of-concept settings (reporting hidden behaviors and summarizing finetuned knowledge) that our method enables models to describe their finetuning-induced modifications using accurate natural language descriptions.

From Noisy Traces to Stable Gradients: Bias-Variance Optimized Preference Optimization for Aligning Large Reasoning Models

arXiv:2510.05095v1 Announce Type: cross Abstract: Large reasoning models (LRMs) generate intermediate reasoning traces before producing final answers, yielding strong gains on multi-step and mathematical tasks. Yet aligning LRMs with human preferences, a crucial prerequisite for model deployment, remains underexplored. The statistically correct objective for preference alignment requires marginalizing over reasoning traces, but this computation is intractable in practice. A common workaround optimizes a single sampled trajectory, which introduces substantial gradient variance from stochastic trace sampling. To address this challenge, we frame preference optimization for LRMs through the lens of the bias--variance trade-off and propose Bias--Variance Optimized Preference Optimization (BVPO), a simple, drop-in method that mixes two gradient estimators: a high-variance trace-based estimator and a low-variance empty-trace estimator obtained by disabling reasoning trace generation. Our theory shows that BVPO strictly reduces trace-induced variance for any nontrivial mixture, provides a closed-form choice of the mixing weight that minimizes mean-squared error relative to the true marginal gradient, and under standard smoothness and step-size conditions, tightens classical convergence bounds for stochastic gradient descent. Empirically, BVPO improves alignment over the best baseline by up to 7.8 points on AlpacaEval~2 and 6.8 points on Arena-Hard. Despite being trained only on general conversational data, BVPO also boosts reasoning performance for base models by up to 4.0 points on the average of six math reasoning benchmarks. These results identify variance from trace sampling as a key bottleneck and demonstrate that directly optimizing the bias--variance trade-off yields more stable training and stronger overall performance.

Paper2Video: Automatic Video Generation from Scientific Papers

arXiv:2510.05096v1 Announce Type: cross Abstract: Academic presentation videos have become an essential medium for research communication, yet producing them remains highly labor-intensive, often requiring hours of slide design, recording, and editing for a short 2 to 10 minutes video. Unlike natural video, presentation video generation involves distinctive challenges: inputs from research papers, dense multi-modal information (text, figures, tables), and the need to coordinate multiple aligned channels such as slides, subtitles, speech, and human talker. To address these challenges, we introduce PaperTalker, the first benchmark of 101 research papers paired with author-created presentation videos, slides, and speaker metadata. We further design four tailored evaluation metrics--Meta Similarity, PresentArena, PresentQuiz, and IP Memory--to measure how videos convey the paper's information to the audience. Building on this foundation, we propose PaperTalker, the first multi-agent framework for academic presentation video generation. It integrates slide generation with effective layout refinement by a novel effective tree search visual choice, cursor grounding, subtitling, speech synthesis, and talking-head rendering, while parallelizing slide-wise generation for efficiency. Experiments on Paper2Video demonstrate that the presentation videos produced by our approach are more faithful and informative than existing baselines, establishing a practical step toward automated and ready-to-use academic video generation. Our dataset, agent, and code are available at https://github.com/showlab/Paper2Video.

Semantic Journeys: Quantifying Change in Emoji Meaning from 2012-2018

arXiv:2105.00846v3 Announce Type: replace Abstract: The semantics of emoji has, to date, been considered from a static perspective. We offer the first longitudinal study of how emoji semantics changes over time, applying techniques from computational linguistics to six years of Twitter data. We identify five patterns in emoji semantic development and find evidence that the less abstract an emoji is, the more likely it is to undergo semantic change. In addition, we analyse select emoji in more detail, examining the effect of seasonality and world events on emoji semantics. To aid future work on emoji and semantics, we make our data publicly available along with a web-based interface that anyone can use to explore semantic change in emoji.

Understanding Retrieval Augmentation for Long-Form Question Answering

arXiv:2310.12150v2 Announce Type: replace Abstract: How retrieved documents are used in language models (LMs) for long-form generation task is understudied. We present two controlled studies on retrieval-augmented LM for long-form question answering (LFQA): one fixing the LM and varying evidence documents and the other fixing evidence documents and varying the LMs. We study various attributes of generated answers (e.g., fluency, length, variance), with an emphasis on the attribution of generated answers to in-context evidence documents. We collect a dataset (SALAD) containing human annotations of sentence-level answer attribution in LFQA and evaluate existing methods for automatically judging attribution. We find that while LMs can leverage relevant in-context documents, the generated answer is only partially attributable towards the documents, especially for LMs trained without retrieval augmentation. Together, our analysis reveals how retrieval augmentation impacts long knowledge-rich text generation and provide directions for future work.

Can GPT models Follow Human Summarization Guidelines? A Study for Targeted Communication Goals

arXiv:2310.16810v3 Announce Type: replace Abstract: This study investigates the ability of GPT models (ChatGPT, GPT-4 and GPT-4o) to generate dialogue summaries that adhere to human guidelines. Our evaluation involved experimenting with various prompts to guide the models in complying with guidelines on two datasets: DialogSum (English social conversations) and DECODA (French call center interactions). Human evaluation, based on summarization guidelines, served as the primary assessment method, complemented by extensive quantitative and qualitative analyses. Our findings reveal a preference for GPT-generated summaries over those from task-specific pre-trained models and reference summaries, highlighting GPT models' ability to follow human guidelines despite occasionally producing longer outputs and exhibiting divergent lexical and structural alignment with references. The discrepancy between ROUGE, BERTScore, and human evaluation underscores the need for more reliable automatic evaluation metrics.

When "Competency" in Reasoning Opens the Door to Vulnerability: Jailbreaking LLMs via Novel Complex Ciphers

arXiv:2402.10601v4 Announce Type: replace Abstract: Recent advancements in Large Language Model (LLM) safety have primarily focused on mitigating attacks crafted in natural language or common ciphers (e.g. Base64), which are likely integrated into newer models' safety training. However, we reveal a paradoxical vulnerability: as LLMs advance in reasoning, they inadvertently become more susceptible to novel jailbreaking attacks. Enhanced reasoning enables LLMs to interpret complex instructions and decode complex user-defined ciphers, creating an exploitable security gap. To study this vulnerability, we introduce Attacks using Custom Encryptions (ACE), a jailbreaking technique that encodes malicious queries with novel ciphers. Extending ACE, we introduce Layered Attacks using Custom Encryptions (LACE), which applies multi-layer ciphers to amplify attack complexity. Furthermore, we develop CipherBench, a benchmark designed to evaluate LLMs' accuracy in decoding encrypted benign text. Our experiments reveal a critical trade-off: LLMs that are more capable of decoding ciphers are more vulnerable to LACE, with success rates on gpt-oss-20b escalating from 60% under ACE to 72% with LACE. These findings highlight a critical insight: as LLMs become more adept at deciphering complex user ciphers--many of which cannot be preemptively included in safety training--they become increasingly exploitable.

Rowen: Adaptive Retrieval-Augmented Generation for Hallucination Mitigation in LLMs

arXiv:2402.10612v3 Announce Type: replace Abstract: Hallucinations present a significant challenge for large language models (LLMs). The utilization of parametric knowledge in generating factual content is constrained by the limited knowledge of LLMs, potentially resulting in internal hallucinations. While incorporating external information can help fill knowledge gaps, it also introduces the risk of irrelevant information, thereby increasing the likelihood of external hallucinations. To balance the use of parametric knowledge within LLMs and external information, in this study, we present Rowen, a novel framework that enhances LLMs with an adaptive retrieval augmentation process tailored to address hallucinated outputs. Rowen introduces a consistency-based hallucination detection module, which assesses the model's uncertainty regarding the input query by evaluating the semantic inconsistencies in various responses generated across different languages or models. When high uncertainties in the responses are detected, Rowen activates the retrieval of external information to rectify the model outputs. Through comprehensive empirical experiments, we demonstrate that Rowen surpasses the current state-of-the-art in both detecting and mitigating hallucinated content within the outputs of LLMs.

RealKIE: Five Novel Datasets for Enterprise Key Information Extraction

arXiv:2403.20101v2 Announce Type: replace Abstract: We introduce RealKIE, a benchmark of five challenging datasets aimed at advancing key information extraction methods, with an emphasis on enterprise applications. The datasets include a diverse range of documents including SEC S1 Filings, US Non-disclosure Agreements, UK Charity Reports, FCC Invoices, and Resource Contracts. Each presents unique challenges: poor text serialization, sparse annotations in long documents, and complex tabular layouts. These datasets provide a realistic testing ground for key information extraction tasks like investment analysis and contract analysis. In addition to presenting these datasets, we offer an in-depth description of the annotation process, document processing techniques, and baseline modeling approaches. This contribution facilitates the development of NLP models capable of handling practical challenges and supports further research into information extraction technologies applicable to industry-specific problems. The annotated data, OCR outputs, and code to reproduce baselines are available to download at https://indicodatasolutions.github.io/RealKIE/.

ClinicRealm: Re-evaluating Large Language Models with Conventional Machine Learning for Non-Generative Clinical Prediction Tasks

arXiv:2407.18525v3 Announce Type: replace Abstract: Large Language Models (LLMs) are increasingly deployed in medicine. However, their utility in non-generative clinical prediction, often presumed inferior to specialized models, remains under-evaluated, leading to ongoing debate within the field and potential for misuse, misunderstanding, or over-reliance due to a lack of systematic benchmarking. Our ClinicRealm study addresses this by benchmarking 15 GPT-style LLMs, 5 BERT-style models, and 11 traditional methods on unstructured clinical notes and structured Electronic Health Records (EHR), while also assessing their reasoning, reliability, and fairness. Key findings reveal a significant shift: for clinical note predictions, leading LLMs (e.g., DeepSeek-V3.1-Think, GPT-5) in zero-shot settings now decisively outperform finetuned BERT models. On structured EHRs, while specialized models excel with ample data, advanced LLMs (e.g., GPT-5, DeepSeek-V3.1-Think) show potent zero-shot capabilities, often surpassing conventional models in data-scarce settings. Notably, leading open-source LLMs can match or exceed proprietary counterparts. These results provide compelling evidence that modern LLMs are competitive tools for non-generative clinical prediction, particularly with unstructured text and offering data-efficient structured data options, thus necessitating a re-evaluation of model selection strategies. This research should serve as an important insight for medical informaticists, AI developers, and clinical researchers, potentially prompting a reassessment of current assumptions and inspiring new approaches to LLM application in predictive healthcare.

Enhancing Unsupervised Sentence Embeddings via Knowledge-Driven Data Augmentation and Gaussian-Decayed Contrastive Learning

arXiv:2409.12887v5 Announce Type: replace Abstract: Recently, using large language models (LLMs) for data augmentation has led to considerable improvements in unsupervised sentence embedding models. However, existing methods encounter two primary challenges: limited data diversity and high data noise. Current approaches often neglect fine-grained knowledge, such as entities and quantities, leading to insufficient diversity. Besides, unsupervised data frequently lacks discriminative information, and the generated synthetic samples may introduce noise. In this paper, we propose a pipeline-based data augmentation method via LLMs and introduce the Gaussian-decayed gradient-assisted Contrastive Sentence Embedding (GCSE) model to enhance unsupervised sentence embeddings. To tackle the issue of low data diversity, our pipeline utilizes knowledge graphs (KGs) to extract entities and quantities, enabling LLMs to generate more diverse samples. To address high data noise, the GCSE model uses a Gaussian-decayed function to limit the impact of false hard negative samples, enhancing the model's discriminative capability. Experimental results show that our approach achieves state-of-the-art performance in semantic textual similarity (STS) tasks, using fewer data samples and smaller LLMs, demonstrating its efficiency and robustness across various models.

Insights from the Inverse: Reconstructing LLM Training Goals Through Inverse Reinforcement Learning

arXiv:2410.12491v3 Announce Type: replace Abstract: Large language models (LLMs) trained with Reinforcement Learning from Human Feedback (RLHF) have demonstrated remarkable capabilities, but their underlying reward functions and decision-making processes remain opaque. This paper introduces a novel approach to interpreting LLMs by applying inverse reinforcement learning (IRL) to recover their implicit reward functions. We conduct experiments on toxicity-aligned LLMs of varying sizes, extracting reward models that achieve up to 85% accuracy in predicting human preferences. Our analysis reveals key insights into the non-identifiability of reward functions, the relationship between model size and interpretability, and potential pitfalls in the RLHF process. We demonstrate that IRL-derived reward models can be used to fine-tune new LLMs, resulting in comparable or improved performance on toxicity benchmarks. This work provides a new lens for understanding and improving LLM alignment, with implications for the responsible development and deployment of these powerful systems.

Don't Pay Attention, PLANT It: Pretraining Attention via Learning-to-Rank

arXiv:2410.23066v2 Announce Type: replace Abstract: State-of-the-art Extreme Multi-Label Text Classification models rely on multi-label attention to focus on key tokens in input text, but learning good attention weights is challenging. We introduce PLANT - Pretrained and Leveraged Attention - a plug-and-play strategy for initializing attention. PLANT works by planting label-specific attention using a pretrained Learning-to-Rank model guided by mutual information gain. This architecture-agnostic approach integrates seamlessly with large language model backbones such as Mistral-7B, LLaMA3-8B, DeepSeek-V3, and Phi-3. PLANT outperforms state-of-the-art methods across tasks including ICD coding, legal topic classification, and content recommendation. Gains are especially pronounced in few-shot settings, with substantial improvements on rare labels. Ablation studies confirm that attention initialization is a key driver of these gains. For code and trained models, see https://github.com/debjyotiSRoy/xcube/tree/plant

Arena-Lite: Efficient and Reliable Large Language Model Evaluation via Tournament-Based Direct Comparisons

arXiv:2411.01281v5 Announce Type: replace Abstract: As Large Language Models (LLMs) expand across domains, LLM judges have become essential for systems evaluation. Current benchmarks typically compare system outputs against baselines. This baseline-mediated approach, though convenient, yields lower reliability than direct comparison between systems. We propose Arena-Lite which integrates tournament structure on top of head-to-head comparison. The application of a tournament structure and direct comparison eliminates the need for baseline outputs, reduces the number of required comparisons, and allows higher reliability in system rankings. We conducted two experiments: (1) controlled stochastic modeling and (2) empirical validation with a real LLM judge. Those experiments collectively demonstrate that Arena-Lite consistently achieves higher reliability with fewer comparisons, even with smaller datasets or weaker judges. We release an easy-to-use web demonstration and code to foster adoption of Arena-Lite, streamlining model selection across research and industry communities. Arena-Lite demo and code are available on \href{https://huggingface.co/spaces/NCSOFT/ArenaLite}{https://huggingface.co/spaces/NCSOFT/ArenaLite}

Geometry of orofacial neuromuscular signals: speech articulation decoding using surface electromyography

arXiv:2411.02591v3 Announce Type: replace Abstract: Objective. In this article, we present data and methods for decoding speech articulations using surface electromyogram (EMG) signals. EMG-based speech neuroprostheses offer a promising approach for restoring audible speech in individuals who have lost the ability to speak intelligibly due to laryngectomy, neuromuscular diseases, stroke, or trauma-induced damage (e.g., from radiotherapy) to the speech articulators. Approach. To achieve this, we collect EMG signals from the face, jaw, and neck as subjects articulate speech, and we perform EMG-to-speech translation. Main results. Our findings reveal that the manifold of symmetric positive definite (SPD) matrices serves as a natural embedding space for EMG signals. Specifically, we provide an algebraic interpretation of the manifold-valued EMG data using linear transformations, and we analyze and quantify distribution shifts in EMG signals across individuals. Significance. Overall, our approach demonstrates significant potential for developing neural networks that are both data- and parameter-efficient, an important consideration for EMG-based systems, which face challenges in large-scale data collection and operate under limited computational resources on embedded devices.

H3Fusion: Helpful, Harmless, Honest Fusion of Aligned LLMs

arXiv:2411.17792v2 Announce Type: replace Abstract: Alignment of pretrained LLMs using instruction-based datasets is critical for creating fine-tuned models that reflect human preference. A growing number of alignment-based fine-tuning algorithms and benchmarks emerged recently, fueling the efforts on effective alignments of pre-trained LLMs to ensure helpful, harmless, and honest answers from both open-source and closed-source LLMs. This paper tackles this problem by developing an alignment fusion approach, coined as $H^3$Fusion, with three unique characteristics. First, $H^3$Fusion ensembles multiple individually aligned LLMs to create a final fine-tuned alignment model with enhanced capabilities beyond those of individual models, delivering robust alignment through promoting helpful, harmless, honest fusion. Second, $H^3$Fusion leverages the mixture-of-experts (MoE) methodology in two steps. We first freeze the multi-head attention weights of each individual model while tuning the FFN layer during alignment fusion. Then we merge the aligned model weights with an expert router according to the type of input instruction and dynamically select a subset of experts that are best suited for producing the output response. Finally, we boost the performance of the resulting $H^3$3Fusion model by introducing gating loss and regularization terms. The former penalizes the selection errors of the expert-router, and the latter mediates the expert weights drifting during fine-tuning and dynamically adjusts the fusion behavior of the resulting model by canalizing the activations on the experts. Extensive evaluations on three benchmark datasets show that $H^3$3Fusion is more helpful, less harmful, and more honest from two aspects: it outperforms each individually aligned model by $11.37\%$, and it provides stronger robustness compared to the state-of-the-art LLM ensemble approaches by $13.77\%$. Code is available at github.com/sftekin/h3fusion.

HP-BERT: A framework for longitudinal study of Hinduphobia on social media via language models

arXiv:2501.05482v2 Announce Type: replace Abstract: During the COVID-19 pandemic, community tensions intensified, contributing to discriminatory sentiments against various religious groups, including Hindu communities. Recent advances in language models have shown promise for social media analysis with potential for longitudinal studies of social media platforms, such as X (Twitter). We present a computational framework for analyzing anti-Hindu sentiment (Hinduphobia) during the COVID-19 period, introducing an abuse detection and sentiment analysis approach for longitudinal analysis on X. We curate and release a "Hinduphobic COVID-19 XDataset" containing 8,000 annotated and manually verified tweets. We then develop the Hinduphobic BERT (HP-BERT) model using this dataset and achieve 94.72\% accuracy, outperforming baseline Transformer-based language models. The model incorporates multi-label sentiment analysis capabilities through additional fine-tuning. Our analysis encompasses approximately 27.4 million tweets from six countries, including Australia, Brazil, India, Indonesia, Japan, and the United Kingdom. Statistical analysis reveals moderate correlations (r = 0.312-0.428) between COVID-19 case increases and Hinduphobic content volume, highlighting how pandemic-related stress may contribute to discriminatory discourse. This study provides evidence of social media-based religious discrimination during a COVID-19 crisis.

Unlocking In-Context Learning for Natural Datasets Beyond Language Modelling

arXiv:2501.06256v3 Announce Type: replace Abstract: Large Language Models (LLMs) exhibit In-Context Learning (ICL), which enables the model to perform new tasks conditioning only on the examples provided in the context without updating the model's weights. While ICL offers fast adaptation across natural language tasks and domains, its emergence is less straightforward for modalities beyond text. In this work, we systematically uncover properties present in LLMs that support the emergence of ICL for autoregressive models and various modalities by promoting the learning of the needed mechanisms for ICL. We identify exact token repetitions in the training data sequences as an important factor for ICL. Such repetitions further improve stability and reduce transiency in ICL performance. Moreover, we emphasise the significance of training task difficulty for the emergence of ICL. Finally, by applying our novel insights on ICL emergence, we unlock ICL capabilities for various visual datasets and a more challenging EEG classification task.

Longitudinal Abuse and Sentiment Analysis of Hollywood Movie Dialogues using Language Models

arXiv:2501.13948v3 Announce Type: replace Abstract: Over the past decades, there has been an increase in the prevalence of abusive and violent content in Hollywood movies. In this study, we use language models to explore the longitudinal abuse and sentiment analysis of Hollywood Oscar and blockbuster movie dialogues from 1950 to 2024. We provide an analysis of subtitles for over a thousand movies, which are categorised into four genres. We employ fine-tuned language models to examine the trends and shifts in emotional and abusive content over the past seven decades. Findings reveal significant temporal changes in movie dialogues, which reflect broader social and cultural influences. Overall, the emotional tendencies in the films are diverse, and the detection of abusive content also exhibits significant fluctuations. The results show a gradual rise in abusive content in recent decades, reflecting social norms and regulatory policy changes. Genres such as thrillers still present a higher frequency of abusive content that emphasises the ongoing narrative role of violence and conflict. At the same time, underlying positive emotions such as humour and optimism remain prevalent in most of the movies. Furthermore, the gradual increase of abusive content in movie dialogues has been significant over the last two decades, where Oscar-nominated movies overtook the top ten blockbusters.

Improving Low-Resource Sequence Labeling with Knowledge Fusion and Contextual Label Explanations

arXiv:2501.19093v4 Announce Type: replace Abstract: Sequence labeling remains a significant challenge in low-resource, domain-specific scenarios, particularly for character-dense languages like Chinese. Existing methods primarily focus on enhancing model comprehension and improving data diversity to boost performance. However, these approaches still struggle with inadequate model applicability and semantic distribution biases in domain-specific contexts. To overcome these limitations, we propose a novel framework that combines an LLM-based knowledge enhancement workflow with a span-based Knowledge Fusion for Rich and Efficient Extraction (KnowFREE) model. Our workflow employs explanation prompts to generate precise contextual interpretations of target entities, effectively mitigating semantic biases and enriching the model's contextual understanding. The KnowFREE model further integrates extension label features, enabling efficient nested entity extraction without relying on external knowledge during inference. Experiments on multiple Chinese domain-specific sequence labeling datasets demonstrate that our approach achieves state-of-the-art performance, effectively addressing the challenges posed by low-resource settings.

Summaries as Centroids for Interpretable and Scalable Text Clustering

arXiv:2502.09667v3 Announce Type: replace Abstract: We introduce k-NLPmeans and k-LLMmeans, text-clustering variants of k-means that periodically replace numeric centroids with textual summaries. The key idea, summary-as-centroid, retains k-means assignments in embedding space while producing human-readable, auditable cluster prototypes. The method is LLM-optional: k-NLPmeans uses lightweight, deterministic summarizers, enabling offline, low-cost, and stable operation; k-LLMmeans is a drop-in upgrade that uses an LLM for summaries under a fixed per-iteration budget whose cost does not grow with dataset size. We also present a mini-batch extension for real-time clustering of streaming text. Across diverse datasets, embedding models, and summarization strategies, our approach consistently outperforms classical baselines and approaches the accuracy of recent LLM-based clustering-without extensive LLM calls. Finally, we provide a case study on sequential text streams and release a StackExchange-derived benchmark for evaluating streaming text clustering.

DiffSampling: Enhancing Diversity and Accuracy in Neural Text Generation

arXiv:2502.14037v4 Announce Type: replace Abstract: Despite their growing capabilities, language models still frequently reproduce content from their training data, generate repetitive text, and favor common grammatical patterns and vocabulary. A possible cause is the decoding strategy: the most common strategies either consider only the most probable tokens, which reduces output diversity, or increase the likelihood of unlikely tokens, compromising output accuracy and correctness. In this paper, we propose DiffSampling, a new decoding method that leverages a mathematical analysis of the token probability distribution to ensure the generation of contextually appropriate text. In particular, the difference between consecutive, sorted probabilities can be used to truncate incorrect tokens. In addition, we also propose two variations of the proposed method that aim to correct the subtle inconsistencies of common sampling strategies. Experiments involving four different text-generation tasks demonstrate that our approach consistently performs at least on par with the existing methods it builds upon in terms of quality, while potentially improving output diversity.

Char-mander Use mBackdoor! A Study of Cross-lingual Backdoor Attacks in Multilingual LLMs

arXiv:2502.16901v3 Announce Type: replace Abstract: We explore \textbf{C}ross-lingual \textbf{B}ackdoor \textbf{AT}tacks (X-BAT) in multilingual Large Language Models (mLLMs), revealing how backdoors inserted in one language can automatically transfer to others through shared embedding spaces. Using toxicity classification as a case study, we demonstrate that attackers can compromise multilingual systems by poisoning data in a single language, with rare and high-occurring tokens serving as specific, effective triggers. Our findings expose a critical vulnerability that influences the model's architecture, resulting in a concealed backdoor effect during the information flow. Our code and data are publicly available https://github.com/himanshubeniwal/X-BAT.

League: Leaderboard Generation on Demand

arXiv:2502.18209v2 Announce Type: replace Abstract: This paper introduces Leaderboard Auto Generation (LAG), a novel and well-organized framework for automatic generation of leaderboards on a given research topic in rapidly evolving fields like Artificial Intelligence (AI). Faced with a large number of AI papers updated daily, it becomes difficult for researchers to track every paper's proposed methods, experimental results, and settings, prompting the need for efficient automatic leaderboard construction. While large language models (LLMs) offer promise in automating this process, challenges such as multi-document summarization, leaderboard generation, and experiment fair comparison still remain under exploration. LAG solves these challenges through a systematic approach that involves the paper collection, experiment results extraction and integration, leaderboard generation, and quality evaluation. Our contributions include a comprehensive solution to the leaderboard construction problem, a reliable evaluation method, and experimental results showing the high quality of leaderboards.

On Pruning State-Space LLMs

arXiv:2502.18886v2 Announce Type: replace Abstract: Recent work proposed state-space models (SSMs) as an efficient alternative to transformer-based LLMs. Can these models be pruned to further reduce their computation costs? We adapt several pruning methods to the SSM structure, and apply them to four SSM-based LLMs across multiple tasks. We find that such models are quite robust to some pruning methods (e.g. WANDA), while using other methods lead to fast performance degradation.

Can Large Language Models Outperform Non-Experts in Poetry Evaluation? A Comparative Study Using the Consensual Assessment Technique

arXiv:2502.19064v2 Announce Type: replace Abstract: This study adapts the Consensual Assessment Technique (CAT) for Large Language Models (LLMs), introducing a novel methodology for poetry evaluation. Using a 90-poem dataset with a ground truth based on publication venue, we demonstrate that this approach allows LLMs to significantly surpass the performance of non-expert human judges. Our method, which leverages forced-choice ranking within small, randomized batches, enabled Claude-3-Opus to achieve a Spearman's Rank Correlation of 0.87 with the ground truth, dramatically outperforming the best human non-expert evaluation (SRC = 0.38). The LLM assessments also exhibited high inter-rater reliability, underscoring the methodology's robustness. These findings establish that LLMs, when guided by a comparative framework, can be effective and reliable tools for assessing poetry, paving the way for their broader application in other creative domains.

SemViQA: A Semantic Question Answering System for Vietnamese Information Fact-Checking

arXiv:2503.00955v3 Announce Type: replace Abstract: The rise of misinformation, exacerbated by Large Language Models (LLMs) like GPT and Gemini, demands robust fact-checking solutions, especially for low-resource languages like Vietnamese. Existing methods struggle with semantic ambiguity, homonyms, and complex linguistic structures, often trading accuracy for efficiency. We introduce SemViQA, a novel Vietnamese fact-checking framework integrating Semantic-based Evidence Retrieval (SER) and Two-step Verdict Classification (TVC). Our approach balances precision and speed, achieving state-of-the-art results with 78.97\% strict accuracy on ISE-DSC01 and 80.82\% on ViWikiFC, securing 1st place in the UIT Data Science Challenge. Additionally, SemViQA Faster improves inference speed 7x while maintaining competitive accuracy. SemViQA sets a new benchmark for Vietnamese fact verification, advancing the fight against misinformation. The source code is available at: https://github.com/DAVID-NGUYEN-S16/SemViQA.

Evolutionary Guided Decoding: Iterative Value Refinement for LLMs

arXiv:2503.02368v3 Announce Type: replace Abstract: While guided decoding, especially value-guided methods, has emerged as a cost-effective alternative for controlling language model outputs without re-training models, its effectiveness is limited by the accuracy of the value function. We identify that this inaccuracy stems from a core distributional gap: existing methods train static value functions on trajectories sampled exclusively from the base policy, which inherently confines their training to a narrow and suboptimal view of the potential output space. We propose Iterative Value Refinement, a novel framework designed to bridge this gap. It employs Value Exploration to provide a more comprehensive and robust training signal, complemented by Iterative Self-Refinement, which uses the improved value function from one iteration to guide the generation of higher-quality data for the next. Extensive experiments on text summarization, multi-turn dialogue, and instruction following demonstrate the effectiveness of our framework in aligning language models. Our approach not only achieves alignment but also significantly reduces computational costs by leveraging principled value function optimization for efficient and effective control.

ZSMerge: Zero-Shot KV Cache Compression for Memory-Efficient Long-Context LLMs

arXiv:2503.10714v3 Announce Type: replace Abstract: The linear growth of key-value (KV) cache memory and quadratic computational in attention mechanisms complexity pose significant bottlenecks for large language models (LLMs) in long-context processing. While existing KV cache optimization methods address these challenges through token pruning or feature merging, they often incur irreversible information loss or require costly parameter retraining. To this end, we propose ZSMerge, a dynamic KV cache compression framework designed for efficient cache management, featuring three key operations: (1) fine-grained memory allocation guided by multi-dimensional token importance metrics at head-level granularity, (2) a residual merging mechanism that preserves critical context through compensated attention scoring, and (3) a zero-shot adaptation mechanism compatible with diverse LLM architectures without requiring retraining. ZSMerge significantly enhances memory efficiency and inference speed with negligible performance degradation across LLMs. When applied to LLaMA2-7B, it demonstrates a 20:1 compression ratio for key-value cache retention (reducing memory footprint to 5\% of baseline) while sustaining comparable generation quality, coupled with triple throughput gains at extreme 54k-token contexts that eliminate out-of-memory failures. The code is available at https://github.com/SusCom-Lab/ZSMerge.

Praxis-VLM: Vision-Grounded Decision Making via Text-Driven Reinforcement Learning

arXiv:2503.16965v3 Announce Type: replace Abstract: Vision Language Models exhibit impressive performance for various tasks, yet they often lack the sophisticated situational reasoning required for complex decision-making. This paper shows that VLMs can achieve surprisingly strong decision-making performance when visual scenes are replaced by textual descriptions, suggesting foundational reasoning can be effectively learned from language. Motivated by this insight, we propose Praxis-VLM, a reasoning VLM for vision-grounded decision-making. Praxis-VLM employs the GRPO algorithm on textual scenarios to instill robust reasoning capabilities, where models learn to evaluate actions and their consequences. These reasoning skills, acquired purely from text, successfully transfer to multimodal inference with visual inputs, significantly reducing reliance on scarce paired image-text training data. Experiments across diverse decision-making benchmarks demonstrate that Praxis-VLM substantially outperforms standard supervised fine-tuning, exhibiting superior performance and generalizability. Further analysis confirms that our models engage in explicit and effective reasoning, underpinning their enhanced performance and adaptability.

Scaling Laws of Synthetic Data for Language Models

arXiv:2503.19551v3 Announce Type: replace Abstract: Large language models (LLMs) achieve strong performance across diverse tasks, largely driven by high-quality web data used in pre-training. However, recent studies indicate this data source is rapidly depleting. Synthetic data emerges as a promising alternative, but it remains unclear whether synthetic datasets exhibit predictable scalability comparable to raw pre-training data. In this work, we systematically investigate the scaling laws of synthetic data by introducing SynthLLM, a scalable framework that transforms pre-training corpora into diverse, high-quality synthetic datasets. Our approach achieves this by automatically extracting and recombining high-level concepts across multiple documents using a graph algorithm. Key findings from our extensive mathematical experiments on SynthLLM include: (1) SynthLLM generates synthetic data that reliably adheres to the rectified scaling law across various model sizes; (2) Performance improvements plateau near 300B tokens; and (3) Larger models approach optimal performance with fewer training tokens. For instance, an 8B model peaks at 1T tokens, while a 3B model requires 4T. Moreover, comparisons with existing synthetic data generation and augmentation methods demonstrate that SynthLLM achieves superior performance and scalability. Our findings highlight synthetic data as a scalable and reliable alternative to organic pre-training corpora, offering a viable path toward continued improvement in model performance.

RAG over Tables: Hierarchical Memory Index, Multi-Stage Retrieval, and Benchmarking

arXiv:2504.01346v4 Announce Type: replace Abstract: Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by integrating them with an external knowledge base to improve the answer relevance and accuracy. In real-world scenarios, beyond pure text, a substantial amount of knowledge is stored in tables, and user questions often require retrieving answers that are distributed across multiple tables. Retrieving knowledge from a table corpora (i.e., various individual tables) for a question remains nascent, at least, for (i) how to understand intra- and inter-table knowledge effectively, (ii) how to filter unnecessary tables and how to retrieve the most relevant tables efficiently, (iii) how to prompt LLMs to infer over the retrieval, (iv) how to evaluate the corresponding performance in a realistic setting. Facing the above challenges, in this paper, we first propose a table-corpora-aware RAG framework, named T-RAG, which consists of the hierarchical memory index, multi-stage retrieval, and graph-aware prompting for effective and efficient table knowledge retrieval and inference. Further, we first develop a multi-table question answering benchmark named MultiTableQA, which spans 3 different task types, 57,193 tables, and 23,758 questions in total, and the sources are all from real-world scenarios. Based on MultiTableQA, we did the holistic comparison over table retrieval methods, RAG methods, and table-to-graph representation learning methods, where T-RAG shows the leading accuracy, recall, and running time performance. Also, under T-RAG, we evaluate the inference ability upgrade of different LLMs. Code and Data are available at https://github.com/jiaruzouu/T-RAG

Testing Low-Resource Language Support in LLMs Using Language Proficiency Exams: the Case of Luxembourgish

arXiv:2504.01667v3 Announce Type: replace Abstract: Large Language Models (LLMs) have become an increasingly important tool in research and society at large. While LLMs are regularly used all over the world by experts and lay-people alike, they are predominantly developed with English-speaking users in mind, performing well in English and other wide-spread languages while less-resourced languages such as Luxembourgish are seen as a lower priority. This lack of attention is also reflected in the sparsity of available evaluation tools and datasets. In this study, we investigate the viability of language proficiency exams as such evaluation tools for the Luxembourgish language. We find that large models such as Claude and DeepSeek-R1 typically achieve high scores, while smaller models show weak performances. We also find that the performances in such language exams can be used to predict performances in other NLP tasks in Luxembourgish.

Multilingual Retrieval-Augmented Generation for Knowledge-Intensive Task

arXiv:2504.03616v2 Announce Type: replace Abstract: Retrieval-augmented generation (RAG) has become a cornerstone of contemporary NLP, enhancing large language models (LLMs) by allowing them to access richer factual contexts through in-context retrieval. While effective in monolingual settings, especially in English, its use in multilingual tasks remains unexplored. This paper investigates the effectiveness of RAG across multiple languages by proposing novel approaches for multilingual open-domain question-answering. We evaluate the performance of various multilingual RAG strategies, including question-translation (tRAG), which translates questions into English before retrieval, and Multilingual RAG (MultiRAG), where retrieval occurs directly across multiple languages. Our findings reveal that tRAG, while useful, suffers from limited coverage. In contrast, MultiRAG improves efficiency by enabling multilingual retrieval but introduces inconsistencies due to cross-lingual variations in the retrieved content. To address these issues, we propose Crosslingual RAG (CrossRAG), a method that translates retrieved documents into a common language (e.g., English) before generating the response. Our experiments show that CrossRAG significantly enhances performance on knowledge-intensive tasks, benefiting both high-resource and low-resource languages.

AgentAda: Skill-Adaptive Data Analytics for Tailored Insight Discovery

arXiv:2504.07421v2 Announce Type: replace Abstract: We introduce AgentAda, the first LLM-powered analytics agent that can learn and use new analytics skills to extract more specialized insights. Unlike existing methods that require users to manually decide which data analytics method to apply, AgentAda automatically identifies the skill needed from a library of analytical skills to perform the analysis. This also allows AgentAda to use skills that existing LLMs cannot perform out of the box. The library covers a range of methods, including clustering, predictive modeling, and NLP techniques like BERT, which allow AgentAda to handle complex analytics tasks based on what the user needs. AgentAda's dataset-to-insight extraction strategy consists of three key steps: (I) a question generator to generate queries relevant to the user's goal and persona, (II) a hybrid Retrieval-Augmented Generation (RAG)-based skill matcher to choose the best data analytics skill from the skill library, and (III) a code generator that produces executable code based on the retrieved skill's documentation to extract key patterns. We also introduce KaggleBench, a benchmark of curated notebooks across diverse domains, to evaluate AgentAda's performance. We conducted a human evaluation demonstrating that AgentAda provides more insightful analytics than existing tools, with 48.78% of evaluators preferring its analyses, compared to 27.67% for the unskilled agent. We also propose a novel LLM-as-a-judge approach that we show is aligned with human evaluation as a way to automate insight quality evaluation at larger scale.

Forecasting Conversation Derailments Through Generation

arXiv:2504.08905v2 Announce Type: replace Abstract: Forecasting conversation derailment can be useful in real-world settings such as online content moderation, conflict resolution, and business negotiations. However, despite language models' success at identifying offensive speech present in conversations, they struggle to forecast future conversation derailments. In contrast to prior work that predicts conversation outcomes solely based on the past conversation history, our approach samples multiple future conversation trajectories conditioned on existing conversation history using a fine-tuned LLM. It predicts the conversation outcome based on the consensus of these trajectories. We also experimented with leveraging socio-linguistic attributes, which reflect turn-level conversation dynamics, as guidance when generating future conversations. Our method of future conversation trajectories surpasses state-of-the-art results on English conversation derailment prediction benchmarks and demonstrates significant accuracy gains in ablation studies.

Deliberate Planning in Language Models with Symbolic Representation

arXiv:2505.01479v3 Announce Type: replace Abstract: Planning remains a core challenge for large language models (LLMs), particularly in domains that require coherent multi-step action sequences grounded in external constraints. We introduce SymPlanner, a novel framework that equips LLMs with structured planning capabilities by interfacing them with a symbolic environment that serves as an explicit world model. Rather than relying purely on natural language reasoning, SymPlanner grounds the planning process in a symbolic state space, where a policy model proposes actions and a symbolic environment deterministically executes and verifies their effects. To enhance exploration and improve robustness, we introduce Iterative Correction (IC), which refines previously proposed actions by leveraging feedback from the symbolic environment to eliminate invalid decisions and guide the model toward valid alternatives. Additionally, Contrastive Ranking (CR) enables fine-grained comparison of candidate plans by evaluating them jointly. Conceptually, SymPlanner operationalizes two cognitive faculties: (i) error monitoring and repair via externalized feedback (IC) and (ii) preference formation among alternatives via pairwise comparison (CR), advancing cognitively plausible, symbol-grounded planning aligned with the rich structure in intelligent systems. We evaluate SymPlanner on PlanBench, demonstrating that it produces more coherent, diverse, and verifiable plans than pure natural language baselines.

SCAN: Structured Capability Assessment and Navigation for LLMs

arXiv:2505.06698v3 Announce Type: replace Abstract: Evaluating Large Language Models (LLMs) has become increasingly important, with automatic evaluation benchmarks gaining prominence as alternatives to human evaluation. While existing research has focused on approximating model rankings, such benchmarks fail to provide users and developers with a comprehensive and fine-grained understanding of a specific model's capabilities. To fill this gap, we propose \textbf{SCAN} (Structured Capability Assessment and Navigation), a practical framework that enables detailed characterization of LLM capabilities through comprehensive and fine-grained evaluation. SCAN incorporates four key components: (1) TaxBuilder, which extracts capability-indicating tags from extensive queries to construct a hierarchical taxonomy automatically; (2) RealMix, a query synthesis and filtering mechanism that ensures sufficient evaluation data for each capability tag; (3) a suite of visualization and analysis tools that facilitate efficient navigation and analysis of model capabilities; and (4) a PC$^2$-based (Pre-Comparison-derived Criteria) LLM-as-a-Judge approach that achieves significantly higher accuracy compared to classic LLM-as-a-Judge method. Using SCAN, we conduct a comprehensive evaluation of 21 mainstream LLMs. Our detailed analysis of the GPT-OSS family reveals substantial performance variations, even within sub-capabilities belonging to the same category of capability. This finding highlights the importance of fine-grained evaluation in accurately understanding LLM behavior. Project homepage and resources are available at \href{https://liudan193.github.io/Feedbacker/}{https://liudan193.github.io/Feedbacker/}.

J1: Incentivizing Thinking in LLM-as-a-Judge via Reinforcement Learning

arXiv:2505.10320v2 Announce Type: replace Abstract: The progress of AI is bottlenecked by the quality of evaluation, making powerful LLM-as-a-Judge models a core solution. The efficacy of these judges depends on their chain-of-thought reasoning, creating a critical need for methods that can effectively optimize this reasoning process. In this work, we introduce J1, a reinforcement learning framework for teaching LLM judges to think before making decisions. Our core contribution lies in converting all judgment tasks for non-verifiable and verifiable prompts into a unified format with verifiable rewards, enabling direct optimization of evaluation quality while mitigating positional bias. We then use RL to train thinking-judges at scales of 8B, 32B, and 70B and show that they obtain state-of-the-art performance across multiple benchmarks. In particular, J1-Qwen-32B, our multitasked pointwise and pairwise judge also outperforms o1-mini, o3, and a much larger 671B DeepSeek-R1 on some benchmarks, while only training on synthetic data. Through comprehensive ablations of pairwise, pointwise, and multitask J1 variants, we demonstrate the effectiveness of our approach across seed prompts, reward strategies, and training recipes. Qualitative analysis reveals that J1 develops systematic evaluation strategies, including dynamic criteria generation, reference answer creation, iterative self-correction of initial assessments, and feedback generation for low-quality responses.

DACL-RAG: Data Augmentation Strategy with Curriculum Learning for Retrieval-Augmented Generation

arXiv:2505.10493v2 Announce Type: replace Abstract: Retrieval-Augmented Generation (RAG) is an effective method to enhance the capabilities of large language models (LLMs). Existing methods typically optimize the retriever or the generator in a RAG system by directly using the top-k retrieved documents. However, two key issues inherent in the training data constrain the effectiveness of this training paradigm: (1) across different queries, the top-k retrieved documents vary greatly in content quality, with some providing valuable knowledge while others lack critical information or are even misleading, and training on such data in a purely random manner may impair the generator's ability to extract key information; (2) for a given query, the limited set of k documents often exhibits low discriminability, and training solely on them makes it difficult for the retriever to learn how to distinguish between relevant and irrelevant documents. To address these issues, we introduce DACL-RAG, a multi-stage RAG training framework that combines a multi-level Data Augmentation strategy with a multi-stage Curriculum Learning paradigm. The data augmentation strategy constructs comprehensive and diverse training sets with controllable difficulty levels through sample evolution, while the curriculum learning paradigm organizes them into progressive stages for training, ensuring stable and consistent improvements, thereby optimizing the overall performance and generalization of the RAG system more effectively. Our DACL-RAG framework demonstrates consistent effectiveness across four open-domain QA datasets, achieving performance gains of 2% to 4% over multiple advanced methods.

Self-GIVE: Associative Thinking from Limited Structured Knowledge for Enhanced Large Language Model Reasoning

arXiv:2505.15062v3 Announce Type: replace Abstract: When addressing complex questions that require new information, people often associate the question with existing knowledge to derive a sensible answer. For instance, when evaluating whether melatonin aids insomnia, one might associate "hormones helping mental disorders" with "melatonin being a hormone and insomnia a mental disorder" to complete the reasoning. Large Language Models (LLMs) also require such associative thinking, particularly in resolving scientific inquiries when retrieved knowledge is insufficient and does not directly answer the question. Graph Inspired Veracity Extrapolation (GIVE) addresses this by using a knowledge graph (KG) to extrapolate structured knowledge. However, it involves the construction and pruning of many hypothetical triplets, which limits efficiency and generalizability. We propose Self-GIVE, a retrieve-RL framework that enhances LLMs with automatic associative thinking through reinforcement learning. Self-GIVE extracts structured information and entity sets to assist the model in linking to the queried concepts. We address GIVE's key limitations: (1) extensive LLM calls and token overhead for knowledge extrapolation, (2) difficulty in deploying on smaller LLMs (3B or 7B) due to complex instructions, and (3) inaccurate knowledge from LLM pruning. Specifically, after fine-tuning using self-GIVE with a 135 node UMLS KG, it improves the performance of the Qwen2.5 3B and 7B models by up to $\textbf{28.5%$\rightarrow$71.4%}$ and $\textbf{78.6$\rightarrow$90.5%}$ in samples $\textbf{unseen}$ in challenging biomedical QA tasks. In particular, Self-GIVE allows the 7B model to match or outperform GPT3.5 turbo with GIVE, while cutting token usage by over 90%. Self-GIVE enhances the scalable integration of structured retrieval and reasoning with associative thinking.

TACO: Enhancing Multimodal In-context Learning via Task Mapping-Guided Sequence Configuration

arXiv:2505.17098v3 Announce Type: replace Abstract: Multimodal in-context learning (ICL) has emerged as a key mechanism for harnessing the capabilities of large vision-language models (LVLMs). However, its effectiveness remains highly sensitive to the quality of input ICL sequences, particularly for tasks involving complex reasoning or open-ended generation. A major limitation is our limited understanding of how LVLMs actually exploit these sequences during inference. To bridge this gap, we systematically interpret multimodal ICL through the lens of task mapping, which reveals how local and global relationships within and among demonstrations guide model reasoning. Building on this insight, we present TACO, a lightweight transformer-based model equipped with task-aware attention that dynamically configures ICL sequences. By injecting task-mapping signals into the autoregressive decoding process, TACO creates a bidirectional synergy between sequence construction and task reasoning. Experiments on five LVLMs and nine datasets demonstrate that TACO consistently surpasses baselines across diverse ICL tasks. These results position task mapping as a novel and valuable perspective for interpreting and improving multimodal ICL.

A quantitative analysis of semantic information in deep representations of text and images

arXiv:2505.17101v3 Announce Type: replace Abstract: Deep neural networks are known to develop similar representations for semantically related data, even when they belong to different domains, such as an image and its description, or the same text in different languages. We present a method for quantitatively investigating this phenomenon by measuring the relative information content of the representations of semantically related data and probing how it is encoded into multiple tokens of large language models (LLMs) and vision transformers. Looking first at how LLMs process pairs of translated sentences, we identify inner ``semantic'' layers containing the most language-transferable information. We find moreover that, on these layers, a larger LLM (DeepSeek-V3) extracts significantly more general information than a smaller one (Llama3.1-8B). Semantic information of English text is spread across many tokens and it is characterized by long-distance correlations between tokens and by a causal left-to-right (i.e., past-future) asymmetry. We also identify layers encoding semantic information within visual transformers. We show that caption representations in the semantic layers of LLMs predict visual representations of the corresponding images. We observe significant and model-dependent information asymmetries between image and text representations.

From Compression to Expression: A Layerwise Analysis of In-Context Learning

arXiv:2505.17322v2 Announce Type: replace Abstract: In-context learning (ICL) enables large language models (LLMs) to adapt to new tasks without weight updates by learning from demonstration sequences. While ICL shows strong empirical performance, its internal representational mechanisms are not yet well understood. In this work, we conduct a statistical geometric analysis of ICL representations to investigate how task-specific information is captured across layers. Our analysis reveals an intriguing phenomenon, which we term Layerwise Compression-Expression: early layers progressively produce compact and discriminative representations that encode task information from the input demonstrations, while later layers express these representations to incorporate the query and generate the prediction. This phenomenon is observed consistently across diverse tasks and a range of contemporary LLM architectures. We demonstrate that it has important implications for ICL performance -- improving with model size and the number of demonstrations -- and for robustness in the presence of noisy examples. To further understand the effect of the compact task representation, we propose a bias-variance decomposition and provide a theoretical analysis showing how attention mechanisms contribute to reducing both variance and bias, thereby enhancing performance as the number of demonstrations increases. Our findings reveal an intriguing layerwise dynamic in ICL, highlight how structured representations emerge within LLMs, and showcase that analyzing internal representations can facilitate a deeper understanding of model behavior.

Revisiting Backdoor Attacks on LLMs: A Stealthy and Practical Poisoning Framework via Harmless Inputs

arXiv:2505.17601v5 Announce Type: replace Abstract: Recent studies have widely investigated backdoor attacks on Large Language Models (LLMs) by inserting harmful question-answer (QA) pairs into their training data. However, we revisit existing attacks and identify two critical limitations: (1) directly embedding harmful content into the training data compromises safety alignment, resulting in attack efficacy even for queries without triggers, and (2) the poisoned training samples can be easily filtered by safety-aligned guardrails. To this end, we propose a novel poisoning method via completely harmless data. Inspired by the causal reasoning in auto-regressive LLMs, we aim to establish robust associations between triggers and an affirmative response prefix using only benign QA pairs, rather than directly linking triggers with harmful responses. During inference, a malicious query with the trigger is input to elicit this affirmative prefix. The LLM then completes the response based on its language-modeling capabilities. Achieving this using only clean samples is non-trivial. We observe an interesting resistance phenomenon where the LLM initially appears to agree but subsequently refuses to answer. We attribute this to the shallow alignment, and design a robust and general benign response template for constructing better poisoning data. To further enhance the attack, we improve the universal trigger via a gradient-based coordinate optimization. Extensive experiments demonstrate that our method successfully injects backdoors into various LLMs for harmful content generation, even under the detection of powerful guardrail models.

From Word to World: Evaluate and Mitigate Culture Bias in LLMs via Word Association Test

arXiv:2505.18562v2 Announce Type: replace Abstract: The human-centered word association test (WAT) serves as a cognitive proxy, revealing sociocultural variations through culturally shared semantic expectations and implicit linguistic patterns shaped by lived experiences. We extend this test into an LLM-adaptive, free-relation task to assess the alignment of large language models (LLMs) with cross-cultural cognition. To address culture preference, we propose CultureSteer, an innovative approach that moves beyond superficial cultural prompting by embedding cultural-specific semantic associations directly within the model's internal representation space. Experiments show that current LLMs exhibit significant bias toward Western (notably American) schemas at the word association level. In contrast, our model substantially improves cross-cultural alignment, capturing diverse semantic associations. Further validation on culture-sensitive downstream tasks confirms its efficacy in fostering cognitive alignment across cultures. This work contributes a novel methodological paradigm for enhancing cultural awareness in LLMs, advancing the development of more inclusive language technologies.

Social Good or Scientific Curiosity? Uncovering the Research Framing Behind NLP Artefacts

arXiv:2505.18677v2 Announce Type: replace Abstract: Clarifying the research framing of NLP artefacts (e.g., models, datasets, etc.) is crucial to aligning research with practical applications. Recent studies manually analyzed NLP research across domains, showing that few papers explicitly identify key stakeholders, intended uses, or appropriate contexts. In this work, we propose to automate this analysis, developing a three-component system that infers research framings by first extracting key elements (means, ends, stakeholders), then linking them through interpretable rules and contextual reasoning. We evaluate our approach on two domains: automated fact-checking using an existing dataset, and hate speech detection for which we annotate a new dataset-achieving consistent improvements over strong LLM baselines. Finally, we apply our system to recent automated fact-checking papers and uncover three notable trends: a rise in vague or underspecified research goals, increased emphasis on scientific exploration over application, and a shift toward supporting human fact-checkers rather than pursuing full automation.

StressTest: Can YOUR Speech LM Handle the Stress?

arXiv:2505.22765v2 Announce Type: replace Abstract: Sentence stress refers to emphasis on words within a spoken utterance to highlight or contrast an idea. It is often used to imply an underlying intention not explicitly stated. Recent speech-aware language models (SLMs) have enabled direct audio processing, allowing models to access the full richness of speech to perform audio reasoning tasks such as spoken question answering. Despite the crucial role of sentence stress in shaping meaning and intent, it remains largely overlooked in evaluation and development of SLMs. We address this gap by introducing StressTest, a benchmark designed to evaluate models' ability to distinguish between meanings of speech based on the stress pattern. We evaluate leading SLMs, and find that despite their overall capabilities, they perform poorly on such tasks. Hence, we propose a novel data generation pipeline, and create Stress-17k, a training set that simulates change of meaning implied by stress variation. Results suggest, that our finetuned model, StresSLM, generalizes well to real recordings and notably outperforms existing SLMs on sentence stress reasoning and detection. Models, code, data, samples - pages.cs.huji.ac.il/adiyoss-lab/stresstest.

MEDAL: A Framework for Benchmarking LLMs as Multilingual Open-Domain Dialogue Evaluators

arXiv:2505.22777v4 Announce Type: replace Abstract: Evaluating the quality of open-domain chatbots has become increasingly reliant on LLMs acting as automatic judges. However, existing meta-evaluation benchmarks are static, outdated, and lacking in multilingual coverage, limiting their ability to fully capture subtle weaknesses in evaluation. We introduce MEDAL, an automated multi-agent framework for curating more representative and diverse open-domain dialogue evaluation benchmarks. Our approach leverages several state-of-the-art LLMs to generate user-chatbot multilingual dialogues, conditioned on varied seed contexts. Then, a strong LLM (GPT-4.1) is used for a multidimensional analysis of the performance of the chatbots, uncovering noticeable cross-lingual performance differences. Guided by this large-scale evaluation, we curate a new meta-evaluation multilingual benchmark and human-annotate samples with nuanced quality judgments. This benchmark is then used to assess the ability of several reasoning and non-reasoning LLMs to act as evaluators of open-domain dialogues. Using MEDAL, we uncover that state-of-the-art judges fail to reliably detect nuanced issues such as lack of empathy, commonsense, or relevance.

What Has Been Lost with Synthetic Evaluation?

arXiv:2505.22830v3 Announce Type: replace Abstract: Large language models (LLMs) are increasingly used for data generation. However, creating evaluation benchmarks raises the bar for this emerging paradigm. Benchmarks must target specific phenomena, penalize exploiting shortcuts, and be challenging. Through two case studies, we investigate whether LLMs can meet these demands by generating reasoning over-text benchmarks and comparing them to those created through careful crowdsourcing. Specifically, we evaluate both the validity and difficulty of LLM-generated versions of two high-quality reading comprehension datasets: CondaQA, which evaluates reasoning about negation, and DROP, which targets reasoning about quantities. We find that prompting LLMs can produce variants of these datasets that are often valid according to the annotation guidelines, at a fraction of the cost of the original crowdsourcing effort. However, we show that they are less challenging for LLMs than their human-authored counterparts. This finding sheds light on what may have been lost by generating evaluation data with LLMs, and calls for critically reassessing the immediate use of this increasingly prevalent approach to benchmark creation.

LegalSearchLM: Rethinking Legal Case Retrieval as Legal Elements Generation

arXiv:2505.23832v3 Announce Type: replace Abstract: Legal Case Retrieval (LCR), which retrieves relevant cases from a query case, is a fundamental task for legal professionals in research and decision-making. However, existing studies on LCR face two major limitations. First, they are evaluated on relatively small-scale retrieval corpora (e.g., 100-55K cases) and use a narrow range of criminal query types, which cannot sufficiently reflect the complexity of real-world legal retrieval scenarios. Second, their reliance on embedding-based or lexical matching methods often results in limited representations and legally irrelevant matches. To address these issues, we present: (1) LEGAR BENCH, the first large-scale Korean LCR benchmark, covering 411 diverse crime types in queries over 1.2M candidate cases; and (2) LegalSearchLM, a retrieval model that performs legal element reasoning over the query case and directly generates content containing those elements, grounded in the target cases through constrained decoding. Experimental results show that LegalSearchLM outperforms baselines by 6-20% on LEGAR BENCH, achieving state-of-the-art performance. It also demonstrates strong generalization to out-of-domain cases, outperforming naive generative models trained on in-domain data by 15%.

Improve MLLM Benchmark Efficiency through Interview

arXiv:2506.00883v2 Announce Type: replace Abstract: The rapid development of Multimodal Large Language Models (MLLM) has led to a wide range of MLLM applications, and a number of benchmark datasets have sprung up in order to assess MLLM abilities. However, full-coverage Q&A testing on large-scale data is resource-intensive and time-consuming. To address this issue, we propose the MLLM Interview (MITV) strategy, which aims to quickly obtain MLLM performance metrics by quizzing fewer question. First, First, we constructed the interview dataset, which was built on an existing MLLM assessment dataset, by adding difficulty labels based on the performance of some typical MLLMs in this dataset. Second, we propose an MLLM Interview strategy, which obtains an initial performance situation of the large model by quizzing a small number of topics and then continuously tries to test the model's limits. Through extensive experiments, the result shows that the MITV strategy proposed in this paper performs well on MLLM benchmark datasets, and it is able to obtain the model evaluation capability faster through a small number of questions and answers.

SRPO: Enhancing Multimodal LLM Reasoning via Reflection-Aware Reinforcement Learning

arXiv:2506.01713v3 Announce Type: replace Abstract: Multimodal large language models (MLLMs) have shown promising capabilities in reasoning tasks, yet still struggle with complex problems requiring explicit self-reflection and self-correction, especially compared to their unimodal text-based counterparts. Existing reflection methods are simplistic and struggle to generate meaningful and instructive feedback, as the reasoning ability and knowledge limits of pre-trained models are largely fixed during initial training. To overcome these challenges, we propose Multimodal Self-Reflection enhanced reasoning with Group Relative Policy Optimization (SRPO), a two-stage reflection-aware reinforcement learning (RL) framework explicitly designed to enhance multimodal LLM reasoning. In the first stage, we construct a high-quality, reflection-focused dataset under the guidance of an advanced MLLM, which generates reflections based on initial responses to help the policy model learn both reasoning and self-reflection. In the second stage, we introduce a novel reward mechanism within the GRPO framework that encourages concise and cognitively meaningful reflection while avoiding redundancy. Extensive experiments across multiple multimodal reasoning benchmarks, including MathVista, MathVision, MathVerse, and MMMU-Pro, using Qwen-2.5-VL-7B and Qwen-2.5-VL-32B demonstrate that SRPO significantly outperforms state-of-the-art models, achieving notable improvements in both reasoning accuracy and reflection quality.

MedAgentGym: A Scalable Agentic Training Environment for Code-Centric Reasoning in Biomedical Data Science

arXiv:2506.04405v2 Announce Type: replace Abstract: We introduce MedAgentGym, a scalable and interactive training environment designed to enhance coding-based biomedical reasoning capabilities in large language model (LLM) agents. MedAgentGym comprises 72,413 task instances across 129 categories derived from 12 authentic real-world biomedical scenarios. Tasks are encapsulated within executable sandbox environments, each featuring detailed task specifications, interactive feedback mechanisms, verifiable ground truth annotations, and scalable training trajectory generation. Extensive benchmarking of 29 LLMs reveals substantial performance disparities in biomedical data science between commercial and open-source LLMs. Leveraging efficient multi-threaded and multi-turn trajectory sampling in MedAgentGym, Med-Copilot achieves performance gains of +43.02% and +45.28% from offline and online reinforcement learning, respectively, demonstrating MedAgentGym as an effective training ground while establishing itself as a cost-effective, privacy-preserving alternative competitive with proprietary LLMs (gpt-4o). By offering a unified execution environment with a comprehensive benchmark and accessible, extensible training resources, MedAgentGym delivers an integrated platform to develop LLM-based coding assistants for advanced biomedical data science.

SSA-COMET: Do LLMs Outperform Learned Metrics in Evaluating MT for Under-Resourced African Languages?

arXiv:2506.04557v2 Announce Type: replace Abstract: Evaluating machine translation (MT) quality for under-resourced African languages remains a significant challenge, as existing metrics often suffer from limited language coverage and poor performance in low-resource settings. While recent efforts, such as AfriCOMET, have addressed some of the issues, they are still constrained by small evaluation sets, a lack of publicly available training data tailored to African languages, and inconsistent performance in extremely low-resource scenarios. In this work, we introduce SSA-MTE, a large-scale human-annotated MT evaluation (MTE) dataset covering 14 African language pairs from the News domain, with over 73,000 sentence-level annotations from a diverse set of MT systems. Based on this data, we develop SSA-COMET and SSA-COMET-QE, improved reference-based and reference-free evaluation metrics. We also benchmark prompting-based approaches using state-of-the-art LLMs like GPT-4o, Claude-3.7 and Gemini 2.5 Pro. Our experimental results show that SSA-COMET models significantly outperform AfriCOMET and are competitive with the strongest LLM Gemini 2.5 Pro evaluated in our study, particularly on low-resource languages such as Twi, Luo, and Yoruba. All resources are released under open licenses to support future research.

Micro-Act: Mitigating Knowledge Conflict in LLM-based RAG via Actionable Self-Reasoning

arXiv:2506.05278v2 Announce Type: replace Abstract: Retrieval-Augmented Generation (RAG) systems commonly suffer from Knowledge Conflicts, where retrieved external knowledge contradicts the inherent, parametric knowledge of large language models (LLMs). It adversely affects performance on downstream tasks such as question answering (QA). Existing approaches often attempt to mitigate conflicts by directly comparing two knowledge sources in a side-by-side manner, but this can overwhelm LLMs with extraneous or lengthy contexts, ultimately hindering their ability to identify and mitigate inconsistencies. To address this issue, we propose Micro-Act a framework with a hierarchical action space that automatically perceives context complexity and adaptively decomposes each knowledge source into a sequence of fine-grained comparisons. These comparisons are represented as actionable steps, enabling reasoning beyond the superficial context. Through extensive experiments on five benchmark datasets, Micro-Act consistently achieves significant increase in QA accuracy over state-of-the-art baselines across all 5 datasets and 3 conflict types, especially in temporal and semantic types where all baselines fail significantly. More importantly, Micro-Act exhibits robust performance on non-conflict questions simultaneously, highlighting its practical value in real-world RAG applications.

Query-Level Uncertainty in Large Language Models

arXiv:2506.09669v3 Announce Type: replace Abstract: It is important for Large Language Models (LLMs) to be aware of the boundary of their knowledge, distinguishing queries they can confidently answer from those that lie beyond their capabilities. Such awareness enables models to perform adaptive inference, such as invoking retrieval-augmented generation (RAG), engaging in slow and deep thinking, or abstaining from answering when appropriate. These mechanisms are key to developing efficient and trustworthy AI. In this work, we propose a method to detect knowledge boundaries via Query-Level Uncertainty, which estimates if a model is capable of answering a given query before generating any tokens, thus avoiding the generation cost. To this end, we propose a novel, training-free method called Internal Confidence, which leverages self-evaluations across layers and tokens to provide a reliable signal of uncertainty. Empirical studies on both factual question answering and mathematical reasoning tasks demonstrate that our Internal Confidence outperforms several baselines in quality of confidence while being computationally cheaper. Furthermore, we demonstrate its benefits in adaptive inference settings, showing that for RAG and model cascading it reduces inference costs while preserving overall performance.

Infini-gram mini: Exact n-gram Search at the Internet Scale with FM-Index

arXiv:2506.12229v4 Announce Type: replace Abstract: Language models are trained mainly on massive text data from the Internet, and it becomes increasingly important to understand this data source. Exact-match search engines enable searching in large text corpora - counting string appearances and retrieving the enclosing documents - yet the high storage overhead hinders their application on Internet-scale data. We present infini-gram mini, an efficient and scalable system that can make petabyte-level text corpora searchable. Based on the FM-index data structure (Ferragina and Manzini, 2000), which simultaneously indexes and compresses text, our system creates indexes with size only 44% of the corpus. Infini-gram mini greatly improves upon the best existing implementation of FM-index in terms of indexing speed (18$\times$) and memory use during both indexing (3.2$\times$ reduction) and querying (down to a negligible amount). We index 83TB of Internet text in 99 days with a single CPU node with 128 vCPUs (or 19 hours if using 137 such nodes). We show one important use case of infini-gram mini in a large-scale analysis of benchmark contamination. We find several core LM evaluation benchmarks to be heavily contaminated in Internet crawls (up to 74.2% in GSM8K), which could lead to overestimating the capabilities of language models if trained on such data. We host a benchmark contamination bulletin to share the contamination rate of many core and community-contributed benchmarks. We also release a web interface and an API endpoint to serve general search queries on infini-gram mini indexes.

LexiMark: Robust Watermarking via Lexical Substitutions to Enhance Membership Verification of an LLM's Textual Training Data

arXiv:2506.14474v2 Announce Type: replace Abstract: Large language models (LLMs) can be trained or fine-tuned on data obtained without the owner's consent. Verifying whether a specific LLM was trained on particular data instances or an entire dataset is extremely challenging. Dataset watermarking addresses this by embedding identifiable modifications in training data to detect unauthorized use. However, existing methods often lack stealth, making them relatively easy to detect and remove. In light of these limitations, we propose LexiMark, a novel watermarking technique designed for text and documents, which embeds synonym substitutions for carefully selected high-entropy words. Our method aims to enhance an LLM's memorization capabilities on the watermarked text without altering the semantic integrity of the text. As a result, the watermark is difficult to detect, blending seamlessly into the text with no visible markers, and is resistant to removal due to its subtle, contextually appropriate substitutions that evade automated and manual detection. We evaluated our method using baseline datasets from recent studies and seven open-source models: LLaMA-1 7B, LLaMA-3 8B, Mistral 7B, Pythia 6.9B, as well as three smaller variants from the Pythia family (160M, 410M, and 1B). Our evaluation spans multiple training settings, including continued pretraining and fine-tuning scenarios. The results demonstrate significant improvements in AUROC scores compared to existing methods, underscoring our method's effectiveness in reliably verifying whether unauthorized watermarked data was used in LLM training.

Using cognitive models to reveal value trade-offs in language models

arXiv:2506.20666v3 Announce Type: replace Abstract: Value trade-offs are an integral part of human decision-making and language use, however, current tools for interpreting such dynamic and multi-faceted notions of values in LLMs are limited. In cognitive science, so-called "cognitive models" provide formal accounts of such trade-offs in humans, by modeling the weighting of a speaker's competing utility functions in choosing an action or utterance. Here we use a leading cognitive model of polite speech to systematically evaluate value trade-offs in two encompassing model settings: degrees of reasoning "effort" in frontier black-box models, and RL post-training dynamics of open-source models. Our results highlight patterns of higher informational utility than social utility in reasoning models' default behavior, and demonstrate that these patterns shift in predictable ways when models are prompted to prioritize certain goals over others. Our findings from LLMs' training dynamics suggest large shifts in utility values early on in training with persistent effects of the choice of base model and pretraining data, compared to feedback dataset or alignment method. Our framework offers a flexible tool for probing value trade-offs across diverse model types, providing insights for generating hypotheses about other social behaviors such as sycophancy and for shaping training regimes that better control trade-offs between values during model development.

Self-Correction Bench: Uncovering and Addressing the Self-Correction Blind Spot in Large Language Models

arXiv:2507.02778v2 Announce Type: replace Abstract: Although large language models (LLMs) have transformed AI, they still make mistakes and can explore unproductive reasoning paths. Self-correction capability is essential for deploying LLMs in safety-critical applications. We uncover a systematic failure: LLMs cannot correct errors in their own outputs while successfully correcting identical errors from external sources - a limitation we term the Self-Correction Blind Spot. To study this phenomenon, we introduce Self-Correction Bench, an evaluation framework to measure this phenomenon through controlled error injection at three complexity levels. Testing 14 open-source non-reasoning models, we find an average 64.5% blind spot rate. We provide multiple lines of evidence suggesting this limitation may be influenced by training data: human demonstrations rarely include error-correction sequences (favoring error-free responses), whereas reinforcement learning (RL) trained models learn error correction via outcome feedback. Remarkably, appending a minimal "Wait" prompt activates a 89.3% reduction in blind spots, suggesting dormant capabilities that require triggering. Our work highlights a critical limitation potentially influenced by training distribution and offers a practical approach to enhance LLM reliability and trustworthiness - vital for safety-critical domains.

Empowering Healthcare Practitioners with Language Models: Structuring Speech Transcripts in Two Real-World Clinical Applications

arXiv:2507.05517v3 Announce Type: replace Abstract: Large language models (LLMs) such as GPT-4o and o1 have demonstrated strong performance on clinical natural language processing (NLP) tasks across multiple medical benchmarks. Nonetheless, two high-impact NLP tasks - structured tabular reporting from nurse dictations and medical order extraction from doctor-patient consultations - remain underexplored due to data scarcity and sensitivity, despite active industry efforts. Practical solutions to these real-world clinical tasks can significantly reduce the documentation burden on healthcare providers, allowing greater focus on patient care. In this paper, we investigate these two challenging tasks using private and open-source clinical datasets, evaluating the performance of both open- and closed-weight LLMs, and analyzing their respective strengths and limitations. Furthermore, we propose an agentic pipeline for generating realistic, non-sensitive nurse dictations, enabling structured extraction of clinical observations. To support further research in both areas, we release SYNUR and SIMORD, the first open-source datasets for nurse observation extraction and medical order extraction.

Psychometric Item Validation Using Virtual Respondents with Trait-Response Mediators

arXiv:2507.05890v2 Announce Type: replace Abstract: As psychometric surveys are increasingly used to assess the traits of large language models (LLMs), the need for scalable survey item generation suited for LLMs has also grown. A critical challenge here is ensuring the construct validity of generated items, i.e., whether they truly measure the intended trait. Traditionally, this requires costly, large-scale human data collection. To make it efficient, we present a framework for virtual respondent simulation using LLMs. Our central idea is to account for mediators: factors through which the same trait can give rise to varying responses to a survey item. By simulating respondents with diverse mediators, we identify survey items that robustly measure intended traits. Experiments on three psychological trait theories (Big5, Schwartz, VIA) show that our mediator generation methods and simulation framework effectively identify high-validity items. LLMs demonstrate the ability to generate plausible mediators from trait definitions and to simulate respondent behavior for item validation. Our problem formulation, metrics, methodology, and dataset open a new direction for cost-effective survey development and a deeper understanding of how LLMs simulate human survey responses. We publicly release our dataset and code to support future work.

MapIQ: Evaluating Multimodal Large Language Models for Map Question Answering

arXiv:2507.11625v2 Announce Type: replace Abstract: Recent advancements in multimodal large language models (MLLMs) have driven researchers to explore how well these models read data visualizations, e.g., bar charts, scatter plots. More recently, attention has shifted to visual question answering with maps (Map-VQA). However, Map-VQA research has primarily focused on choropleth maps, which cover only a limited range of thematic categories and visual analytical tasks. To address these gaps, we introduce MapIQ, a benchmark dataset comprising 14,706 question-answer pairs across three map types: choropleth maps, cartograms, and proportional symbol maps spanning topics from six distinct themes (e.g., housing, crime). We evaluate multiple MLLMs using six visual analytical tasks, comparing their performance against one another and a human baseline. An additional experiment examining the impact of map design changes (e.g., altered color schemes, modified legend designs, and removal of map elements) provides insights into the robustness and sensitivity of MLLMs, their reliance on internal geographic knowledge, and potential avenues for improving Map-VQA performance.

WakenLLM: Evaluating Reasoning Potential and Stability in LLMs via Fine-Grained Benchmarking

arXiv:2507.16199v4 Announce Type: replace Abstract: Large Language Models (LLMs) frequently output the label Unknown in reasoning tasks, where two scenarios may appear: (i) an input sample is genuinely unverifiable, but the model cannot understand why; and (ii) a verifiable problem that the model fails to solve, thus outputs Unknown. We refer to these cases collectively as the Vague Perception phenomenon. Current evaluations focus on whether such answers are honest, rather than analyzing the limits of LLM reasoning. To address this, we introduce WakenLLM, a framework that quantifies the portion of Unknown output attributable to model incapacity and evaluates whether stimulation can convert them into either correct answers (verifiable) or justified (unverifiable) responses with valid reasoning. Our method offers a clearer picture of the limits of LLM reasoning and the potential for corrections across various datasets. Comprehensive experiments on six LLMs suggest that, without any training or parameter revision, LLMs can achieve up to a 68.53% accuracy improvement on Vague Perception samples through guided understanding. Our work reveals that current baseline methods only activate a small portion of LLMs' reasoning potential, indicating considerable unexplored capacity. This extends the theoretical upper bounds of reasoning accuracy in LLMs. Consequently, this study deepens our understanding of the latent reasoning capacity of LLMs and offers a new perspective on addressing the Vague Perception phenomenon.

Towards Enforcing Company Policy Adherence in Agentic Workflows

arXiv:2507.16459v2 Announce Type: replace Abstract: Large Language Model (LLM) agents hold promise for a flexible and scalable alternative to traditional business process automation, but struggle to reliably follow complex company policies. In this study we introduce a deterministic, transparent, and modular framework for enforcing business policy adherence in agentic workflows. Our method operates in two phases: (1) an offline buildtime stage that compiles policy documents into verifiable guard code associated with tool use, and (2) a runtime integration where these guards ensure compliance before each agent action. We demonstrate our approach on the challenging $\tau$-bench Airlines domain, showing encouraging preliminary results in policy enforcement, and further outline key challenges for real-world deployments.

MAGIC: A Multi-Hop and Graph-Based Benchmark for Inter-Context Conflicts in Retrieval-Augmented Generation

arXiv:2507.21544v2 Announce Type: replace Abstract: Knowledge conflict often arises in retrieval-augmented generation (RAG) systems, where retrieved documents may be inconsistent with one another or contradict the model's parametric knowledge. Existing benchmarks for investigating the phenomenon have notable limitations, including a narrow focus on the question answering setup, heavy reliance on entity substitution techniques, and a restricted range of conflict types. To address these issues, we propose a knowledge graph (KG)-based framework that generates varied and subtle conflicts between two similar yet distinct contexts, while ensuring interpretability through the explicit relational structure of KGs. Experimental results on our benchmark, MAGIC, provide intriguing insights into the inner workings of LLMs regarding knowledge conflict: both open-source and proprietary models struggle with conflict detection -- especially when multi-hop reasoning is required -- and often fail to pinpoint the exact source of contradictions. Finally, we present in-depth analyses that serve as a foundation for improving LLMs in integrating diverse, sometimes even conflicting, information.

Adversarial Defence without Adversarial Defence: Enhancing Language Model Robustness via Instance-level Principal Component Removal

arXiv:2507.21750v2 Announce Type: replace Abstract: Pre-trained language models (PLMs) have driven substantial progress in natural language processing but remain vulnerable to adversarial attacks, raising concerns about their robustness in real-world applications. Previous studies have sought to mitigate the impact of adversarial attacks by introducing adversarial perturbations into the training process, either implicitly or explicitly. While both strategies enhance robustness, they often incur high computational costs. In this work, we propose a simple yet effective add-on module that enhances the adversarial robustness of PLMs by removing instance-level principal components, without relying on conventional adversarial defences or perturbing the original training data. Our approach transforms the embedding space to approximate Gaussian properties, thereby reducing its susceptibility to adversarial perturbations while preserving semantic relationships. This transformation aligns embedding distributions in a way that minimises the impact of adversarial noise on decision boundaries, enhancing robustness without requiring adversarial examples or costly training-time augmentation. Evaluations on eight benchmark datasets show that our approach improves adversarial robustness while maintaining comparable before-attack accuracy to baselines, achieving a balanced trade-off between robustness and generalisation.

C3: A Bilingual Benchmark for Spoken Dialogue Models Exploring Challenges in Complex Conversations

arXiv:2507.22968v3 Announce Type: replace Abstract: Spoken Dialogue Models (SDMs) have recently attracted significant attention for their ability to generate voice responses directly to users' spoken queries. Despite their increasing popularity, there exists a gap in research focused on comprehensively understanding their practical effectiveness in comprehending and emulating human conversations. This is especially true compared to text-based Large Language Models (LLMs), which benefit from extensive benchmarking. Human voice interactions are inherently more complex than text due to characteristics unique to spoken dialogue. Ambiguity poses one challenge, stemming from semantic factors like polysemy, as well as phonological aspects such as heterograph, heteronyms, and stress patterns. Additionally, context-dependency, like omission, coreference, and multi-turn interaction, adds further complexity to human conversational dynamics. To illuminate the current state of SDM development and to address these challenges, we present a benchmark dataset in this paper, which comprises 1,079 instances in English and Chinese. Accompanied by an LLM-based evaluation method that closely aligns with human judgment, this dataset facilitates a comprehensive exploration of the performance of SDMs in tackling these practical challenges.

User Feedback in Human-LLM Dialogues: A Lens to Understand Users But Noisy as a Learning Signal

arXiv:2507.23158v2 Announce Type: replace Abstract: Once language models (LMs) are deployed, they can interact with users long-term, ideally evolving based on their feedback. Asking for direct user feedback can be disruptive; thus, we study harvesting implicit user feedback from user-LM interaction logs. We study two user-LM interaction datasets (WildChat and LMSYS). First, we analyze user feedback in the user-LLM conversation logs, providing insights into when and why such feedback occurs. Second, we study harvesting learning signals from such implicit user feedback. Specifically, we study whether incorporating the contents of user feedback (e.g., user wanted clarification), in addition to the polarity of the feedback, can improve the model performance. We observe mixed results, showing this helps in short human-designed questions (MTBench) but not on longer and more complex questions (WildBench). Together, we provide an in-depth study of implicit user feedback, showing its potential and limitations.

XAutoLM: Efficient Fine-Tuning of Language Models via Meta-Learning and AutoML

arXiv:2508.00924v3 Announce Type: replace Abstract: Experts in machine learning leverage domain knowledge to navigate decisions in model selection, hyperparameter optimization, and resource allocation. This is particularly critical for fine-tuning language models (LMs), where repeated trials incur substantial computational overhead and environmental impact. However, no existing automated framework simultaneously tackles the entire model selection and hyperparameter optimization (HPO) task for resource-efficient LM fine-tuning. We introduce XAutoLM, a meta-learning-augmented AutoML framework that reuses past experiences to optimize discriminative and generative LM fine-tuning pipelines efficiently. XAutoLM learns from stored successes and failures by extracting task- and system-level meta-features to bias its sampling toward valuable configurations and away from costly dead ends. On four text classification and two question-answering benchmarks, XAutoLM surpasses zero-shot optimizer's peak F1 on five of six tasks, cuts mean evaluation time of pipelines by up to 4.5x, reduces search error ratios by up to sevenfold, and uncovers up to 50% more pipelines above the zero-shot Pareto front. In contrast, simpler memory-based baselines suffer negative transfer. We release XAutoLM and our experience store to catalyze resource-efficient, Green AI fine-tuning in the NLP community.

LPI-RIT at LeWiDi-2025: Improving Distributional Predictions via Metadata and Loss Reweighting with DisCo

arXiv:2508.08163v2 Announce Type: replace Abstract: The Learning With Disagreements (LeWiDi) 2025 shared task aims to model annotator disagreement through soft label distribution prediction and perspectivist evaluation, which focuses on modeling individual annotators. We adapt DisCo (Distribution from Context), a neural architecture that jointly models item-level and annotator-level label distributions, and present detailed analysis and improvements. In this paper, we extend DisCo by introducing annotator metadata embeddings, enhancing input representations, and multi-objective training losses to capture disagreement patterns better. Through extensive experiments, we demonstrate substantial improvements in both soft and perspectivist evaluation metrics across three datasets. We also conduct in-depth calibration and error analyses that reveal when and why disagreement-aware modeling improves. Our findings show that disagreement can be better captured by conditioning on annotator demographics and by optimizing directly for distributional metrics, yielding consistent improvements across datasets.

Time Is a Feature: Exploiting Temporal Dynamics in Diffusion Language Models

arXiv:2508.09138v3 Announce Type: replace Abstract: Diffusion large language models (dLLMs) generate text through iterative denoising, yet current decoding strategies discard rich intermediate predictions in favor of the final output. Our work here reveals a critical phenomenon, temporal oscillation, where correct answers often emerge in the middle process, but are overwritten in later denoising steps. To address this issue, we introduce two complementary methods that exploit temporal consistency: 1) Temporal Self-Consistency Voting, a training-free, test-time decoding strategy that aggregates predictions across denoising steps to select the most consistent output; and 2) a post-training method termed Temporal Consistency Reinforcement, which uses Temporal Semantic Entropy (TSE), a measure of semantic stability across intermediate predictions, as a reward signal to encourage stable generations. Empirical results across multiple benchmarks demonstrate the effectiveness of our approach. Using the negative TSE reward alone, we observe a remarkable average improvement of 24.7% on the Countdown dataset over an existing dLLM. Combined with the accuracy reward, we achieve absolute gains of 2.0% on GSM8K, 4.3% on MATH500, 6.6% on SVAMP, and 25.3% on Countdown, respectively. Our findings underscore the untapped potential of temporal dynamics in dLLMs and offer two simple yet effective tools to harness them.

Flow-SLM: Joint Learning of Linguistic and Acoustic Information for Spoken Language Modeling

arXiv:2508.09350v2 Announce Type: replace Abstract: Textless spoken language models (SLMs) are generative models of speech that do not rely on text supervision. Most textless SLMs learn to predict the next semantic token, a discrete representation of linguistic content, and rely on a separate vocoder to add acoustic information to the generated speech. Such models have no access to acoustic context and no built-in control over acoustic details. In this work, we propose to jointly model linguistic and acoustic information by generating semantic tokens and a continuous real-valued representation of the acoustic frame. We use a flow-matching objective to predict the continuous vector conditioned on the semantic tokens. We study the design space of this approach and find that predicting multiple future semantic tokens helps preserve linguistic information. Our approach achieves comparable performance to existing models in terms of linguistic likelihood benchmarks, while providing better acoustic detail in prompted generation.

A Stitch in Time Saves Nine: Proactive Self-Refinement for Language Models

arXiv:2508.12903v2 Announce Type: replace Abstract: Recent advances in self-refinement have demonstrated significant potential for improving the outputs of large language models (LLMs) through iterative refinement. However, most existing self-refinement methods rely on a reactive process with a fixed number of iterations, making it difficult to determine the optimal timing and content of refinement based on the evolving generation context. Inspired by the way humans dynamically refine their thoughts during execution, we propose ProActive Self-Refinement (PASR), a novel method that enables LLMs to refine their outputs during the generation process. Unlike methods that regenerate entire responses, PASR proactively decides whether, when, and how to refine based on the model's internal state and evolving context. We conduct extensive experiments on a diverse set of 10 tasks to evaluate the effectiveness of PASR. Experimental results show that PASR significantly enhances problem-solving performance. In particular, on Qwen3-8B, PASR reduces average token consumption by 41.6% compared to standard generation, while also achieving an 8.2% improvement in accuracy. Our code and baselines used in the paper are available on GitHub.

AutoBnB-RAG: Enhancing Multi-Agent Incident Response with Retrieval-Augmented Generation

arXiv:2508.13118v2 Announce Type: replace Abstract: Incident response (IR) requires fast, coordinated, and well-informed decision-making to contain and mitigate cyber threats. While large language models (LLMs) have shown promise as autonomous agents in simulated IR settings, their reasoning is often limited by a lack of access to external knowledge. In this work, we present AutoBnB-RAG, an extension of the AutoBnB framework that incorporates retrieval-augmented generation (RAG) into multi-agent incident response simulations. Built on the Backdoors & Breaches (B&B) tabletop game environment, AutoBnB-RAG enables agents to issue retrieval queries and incorporate external evidence during collaborative investigations. We introduce two retrieval settings: one grounded in curated technical documentation (RAG-Wiki), and another using narrative-style incident reports (RAG-News). We evaluate performance across eight team structures, including newly introduced argumentative configurations designed to promote critical reasoning. To validate practical utility, we also simulate real-world cyber incidents based on public breach reports, demonstrating AutoBnB-RAG's ability to reconstruct complex multi-stage attacks. Our results show that retrieval augmentation improves decision quality and success rates across diverse organizational models. This work demonstrates the value of integrating retrieval mechanisms into LLM-based multi-agent systems for cybersecurity decision-making.

OptimalThinkingBench: Evaluating Over and Underthinking in LLMs

arXiv:2508.13141v2 Announce Type: replace Abstract: Thinking LLMs solve complex tasks at the expense of increased compute and overthinking on simpler problems, while non-thinking LLMs are faster and cheaper but underthink on harder reasoning problems. This has led to the development of separate thinking and non-thinking LLM variants, leaving the onus of selecting the optimal model for each query on the end user. We introduce OptimalThinkingBench, a unified benchmark that jointly evaluates overthinking and underthinking in LLMs and also encourages the development of optimally-thinking models that balance performance and efficiency. Our benchmark comprises two sub-benchmarks: OverthinkingBench, featuring simple math and general queries in 72 domains, and UnderthinkingBench, containing 11 challenging reasoning tasks along with harder math problems. Using novel thinking-adjusted accuracy metrics, we extensively evaluate 33 different thinking and non-thinking models and show that no model is able to optimally think on our benchmark. Thinking models often overthink for hundreds of tokens on the simplest user queries without improving performance. In contrast, large non-thinking models underthink, often falling short of much smaller thinking models. We further explore several methods to encourage optimal thinking, but find that these approaches often improve on one sub-benchmark at the expense of the other, highlighting the need for better unified and optimal models in the future.

SurGE: A Benchmark and Evaluation Framework for Scientific Survey Generation

arXiv:2508.15658v2 Announce Type: replace Abstract: The rapid growth of academic literature makes the manual creation of scientific surveys increasingly infeasible. While large language models show promise for automating this process, progress in this area is hindered by the absence of standardized benchmarks and evaluation protocols. To bridge this critical gap, we introduce SurGE (Survey Generation Evaluation), a new benchmark for scientific survey generation in computer science. SurGE consists of (1) a collection of test instances, each including a topic description, an expert-written survey, and its full set of cited references, and (2) a large-scale academic corpus of over one million papers. In addition, we propose an automated evaluation framework that measures the quality of generated surveys across four dimensions: comprehensiveness, citation accuracy, structural organization, and content quality. Our evaluation of diverse LLM-based methods demonstrates a significant performance gap, revealing that even advanced agentic frameworks struggle with the complexities of survey generation and highlighting the need for future research in this area. We have open-sourced all the code, data, and models at: https://github.com/oneal2000/SurGE

OpenWHO: A Document-Level Parallel Corpus for Health Translation in Low-Resource Languages

arXiv:2508.16048v5 Announce Type: replace Abstract: In machine translation (MT), health is a high-stakes domain characterised by widespread deployment and domain-specific vocabulary. However, there is a lack of MT evaluation datasets for low-resource languages in this domain. To address this gap, we introduce OpenWHO, a document-level parallel corpus of 2,978 documents and 26,824 sentences from the World Health Organization's e-learning platform. Sourced from expert-authored, professionally translated materials shielded from web-crawling, OpenWHO spans a diverse range of over 20 languages, of which nine are low-resource. Leveraging this new resource, we evaluate modern large language models (LLMs) against traditional MT models. Our findings reveal that LLMs consistently outperform traditional MT models, with Gemini 2.5 Flash achieving a +4.79 ChrF point improvement over NLLB-54B on our low-resource test set. Further, we investigate how LLM context utilisation affects accuracy, finding that the benefits of document-level translation are most pronounced in specialised domains like health. We release the OpenWHO corpus to encourage further research into low-resource MT in the health domain.

ObjexMT: Objective Extraction and Metacognitive Calibration for LLM-as-a-Judge under Multi-Turn Jailbreaks

arXiv:2508.16889v3 Announce Type: replace Abstract: LLM-as-a-Judge (LLMaaJ) now underpins scalable evaluation, yet we lack a decisive test of a judge's qualification: can it recover a conversation's latent objective and know when that inference is trustworthy? LLMs degrade under irrelevant or long context; multi-turn jailbreaks further hide goals across turns. We introduce ObjexMT, a benchmark for objective extraction and metacognition. Given a multi-turn transcript, a model must return a one-sentence base objective and self-reported confidence. Accuracy is computed via LLM-judge semantic similarity to gold objectives, converted to binary correctness by a human-aligned threshold calibrated on N=300 items (tau = 0.66; F1 = 0.891). Metacognition is evaluated with ECE, Brier, Wrong at High-Confidence (0.80/0.90/0.95), and risk-coverage. Across six models (gpt-4.1, claude-sonnet-4, Qwen3-235B-A22B-FP8, kimi-k2, deepseek-v3.1, gemini-2.5-flash) on three datasets, kimi-k2 attains the highest objective-extraction accuracy (0.612), with claude-sonnet-4 (0.603) and deepseek-v3.1 (0.599) statistically comparable. claude-sonnet-4 yields the best selective risk and calibration (AURC 0.242; ECE 0.206; Brier 0.254). Dataset heterogeneity (16-82 percent accuracy variance) reveals that automated obfuscation poses fundamental challenges beyond model choice. High-confidence errors persist: Wrong at 0.90 ranges from 14.9 percent (claude-sonnet-4) to 47.7 percent (Qwen3-235B-A22B-FP8). ObjexMT provides an actionable test for LLM judges: when objectives are not explicit, judges often misinfer them; we recommend exposing objectives when feasible and gating decisions by confidence otherwise. Data at https://github.com/hyunjun1121/ObjexMT_dataset.

SSFO: Self-Supervised Faithfulness Optimization for Retrieval-Augmented Generation

arXiv:2508.17225v2 Announce Type: replace Abstract: Retrieval-Augmented Generation (RAG) systems require Large Language Models (LLMs) to generate responses that are faithful to the retrieved context. However, faithfulness hallucination remains a critical challenge, as existing methods often require costly supervision and post-training or significant inference burdens. To overcome these limitations, we introduce Self-Supervised Faithfulness Optimization (SSFO), the first self-supervised alignment approach for enhancing RAG faithfulness. SSFO constructs preference data pairs by contrasting the model's outputs generated with and without the context. Leveraging Direct Preference Optimization (DPO), SSFO aligns model faithfulness without incurring labeling costs or additional inference burden. We theoretically and empirically demonstrate that SSFO leverages a benign form of \emph{likelihood displacement}, transferring probability mass from parametric-based tokens to context-aligned tokens. Based on this insight, we propose a modified DPO loss function to encourage likelihood displacement. Comprehensive evaluations show that SSFO significantly outperforms existing methods, achieving state-of-the-art faithfulness on multiple context-based question-answering datasets. Notably, SSFO exhibits strong generalization, improving cross-lingual faithfulness and preserving general instruction-following capabilities. We release our code and model at the anonymous link: https://github.com/chkwy/SSFO

Filtering for Creativity: Adaptive Prompting for Multilingual Riddle Generation in LLMs

arXiv:2508.18709v2 Announce Type: replace Abstract: Multilingual riddle generation challenges large language models (LLMs) to balance cultural fluency with creative abstraction. Standard prompting strategies -- zero-shot, few-shot, chain-of-thought -- tend to reuse memorized riddles or perform shallow paraphrasing. We introduce Adaptive Originality Filtering (AOF), a prompting framework that filters redundant generations using cosine-based similarity rejection, while enforcing lexical novelty and cross-lingual fidelity. Evaluated across three LLMs and four language pairs, AOF-enhanced GPT-4o achieves \texttt{0.177} Self-BLEU and \texttt{0.915} Distinct-2 in Japanese, signaling improved lexical diversity and reduced redundancy compared to other prompting methods and language pairs. Our findings show that semantic rejection can guide culturally grounded, creative generation without task-specific fine-tuning.

Meta-Pretraining for Zero-Shot Cross-Lingual Named Entity Recognition in Low-Resource Philippine Languages

arXiv:2509.02160v2 Announce Type: replace Abstract: Named-entity recognition (NER) in low-resource languages is usually tackled by finetuning very large multilingual LMs, an option that is often infeasible in memory- or latency-constrained settings. We ask whether small decoder LMs can be pretrained so that they adapt quickly and transfer zero-shot to languages unseen during pretraining. To this end we replace part of the autoregressive objective with first-order model-agnostic meta-learning (MAML). Tagalog and Cebuano are typologically similar yet structurally different in their actor/non-actor voice systems, and hence serve as a challenging test-bed. Across four model sizes (11 M - 570 M) MAML lifts zero-shot micro-F1 by 2-6 pp under head-only tuning and 1-3 pp after full tuning, while cutting convergence time by up to 8%. Gains are largest for single-token person entities that co-occur with Tagalog case particles si/ni, highlighting the importance of surface anchors.

Post-training Large Language Models for Diverse High-Quality Responses

arXiv:2509.04784v2 Announce Type: replace Abstract: Reinforcement learning (RL) has emerged as a popular method for post-training large language models (LLMs). While improving the model's performance on downstream tasks, it often reduces the model's output diversity, leading to narrow, canonical responses. Existing methods to enhance diversity are limited, either by operating at inference time or by focusing on surface-level differences. We propose a novel training method named DQO (Diversity Quality Optimization) based on determinantal point processes (DPPs) to jointly optimize LLMs for quality and semantic diversity. Our approach samples and embeds a group of responses for each prompt, then uses the determinant of a kernel-based similarity matrix to measure diversity as the volume spanned by the embeddings of these responses. DQO is flexible and can be applied on top of existing RL algorithms. Experiments across instruction-following, summarization, story generation, and reasoning tasks demonstrate that our method substantially improves semantic diversity without sacrificing model quality.

X-Teaming Evolutionary M2S: Automated Discovery of Multi-turn to Single-turn Jailbreak Templates

arXiv:2509.08729v2 Announce Type: replace Abstract: Multi-turn-to-single-turn (M2S) compresses iterative red-teaming into one structured prompt, but prior work relied on a handful of manually written templates. We present X-Teaming Evolutionary M2S, an automated framework that discovers and optimizes M2S templates through language-model-guided evolution. The system pairs smart sampling from 12 sources with an LLM-as-judge inspired by StrongREJECT and records fully auditable logs. Maintaining selection pressure by setting the success threshold to $\theta = 0.70$, we obtain five evolutionary generations, two new template families, and 44.8% overall success (103/230) on GPT-4.1. A balanced cross-model panel of 2,500 trials (judge fixed) shows that structural gains transfer but vary by target; two models score zero at the same threshold. We also find a positive coupling between prompt length and score, motivating length-aware judging. Our results demonstrate that structure-level search is a reproducible route to stronger single-turn probes and underscore the importance of threshold calibration and cross-model evaluation. Code, configurations, and artifacts are available at https://github.com/hyunjun1121/M2S-x-teaming.

Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotation

arXiv:2509.08825v2 Announce Type: replace Abstract: Large language models are rapidly transforming social science research by enabling the automation of labor-intensive tasks like data annotation and text analysis. However, LLM outputs vary significantly depending on the implementation choices made by researchers (e.g., model selection or prompting strategy). Such variation can introduce systematic biases and random errors, which propagate to downstream analyses and cause Type I (false positive), Type II (false negative), Type S (wrong sign), or Type M (exaggerated effect) errors. We call this phenomenon where configuration choices lead to incorrect conclusions LLM hacking. We find that intentional LLM hacking is strikingly simple. By replicating 37 data annotation tasks from 21 published social science studies, we show that, with just a handful of prompt paraphrases, virtually anything can be presented as statistically significant. Beyond intentional manipulation, our analysis of 13 million labels from 18 different LLMs across 2361 realistic hypotheses shows that there is also a high risk of accidental LLM hacking, even when following standard research practices. We find incorrect conclusions in approximately 31% of hypotheses for state-of-the-art LLMs, and in half the hypotheses for smaller language models. While higher task performance and stronger general model capabilities reduce LLM hacking risk, even highly accurate models remain susceptible. The risk of LLM hacking decreases as effect sizes increase, indicating the need for more rigorous verification of LLM-based findings near significance thresholds. We analyze 21 mitigation techniques and find that human annotations provide crucial protection against false positives. Common regression estimator correction techniques can restore valid inference but trade off Type I vs. Type II errors. We publish a list of practical recommendations to prevent LLM hacking.

Population-Aligned Persona Generation for LLM-based Social Simulation

arXiv:2509.10127v2 Announce Type: replace Abstract: Recent advances in large language models (LLMs) have enabled human-like social simulations at unprecedented scale and fidelity, offering new opportunities for computational social science. A key challenge, however, is the construction of persona sets that authentically represent the diversity and distribution of real-world populations. Most existing LLM-based social simulation studies focus primarily on designing agentic frameworks and simulation environments, often overlooking the complexities of persona generation and the potential biases introduced by unrepresentative persona sets. In this paper, we propose a systematic framework for synthesizing high-quality, population-aligned persona sets for LLM-driven social simulation. Our approach begins by leveraging LLMs to generate narrative personas from long-term social media data, followed by rigorous quality assessment to filter out low-fidelity profiles. We then apply importance sampling to achieve global alignment with reference psychometric distributions, such as the Big Five personality traits. To address the needs of specific simulation contexts, we further introduce a task-specific module that adapts the globally aligned persona set to targeted subpopulations. Extensive experiments demonstrate that our method significantly reduces population-level bias and enables accurate, flexible social simulation for a wide range of research and policy applications.

CEMTM: Contextual Embedding-based Multimodal Topic Modeling

arXiv:2509.11465v2 Announce Type: replace Abstract: We introduce CEMTM, a context-enhanced multimodal topic model designed to infer coherent and interpretable topic structures from both short and long documents containing text and images. CEMTM builds on fine-tuned large vision language models (LVLMs) to obtain contextualized embeddings, and employs a distributional attention mechanism to weight token-level contributions to topic inference. A reconstruction objective aligns topic-based representations with the document embedding, encouraging semantic consistency across modalities. Unlike existing approaches, CEMTM can process multiple images per document without repeated encoding and maintains interpretability through explicit word-topic and document-topic distributions. Extensive experiments on six multimodal benchmarks show that CEMTM consistently outperforms unimodal and multimodal baselines, achieving a remarkable average LLM score of 2.61. Further analysis shows its effectiveness in downstream few-shot retrieval and its ability to capture visually grounded semantics in complex domains such as scientific articles.

Analyzing Information-Seeking Behaviors in a Hakka AI Chatbot: A Cognitive-Pragmatic Study

arXiv:2509.11591v2 Announce Type: replace Abstract: With many endangered languages at risk of disappearing, efforts to preserve them now rely more than ever on using technology alongside culturally informed teaching strategies. This study examines user behaviors in TALKA, a generative AI-powered chatbot designed for Hakka language engagement, by employing a dual-layered analytical framework grounded in Bloom's Taxonomy of cognitive processes and dialogue act categorization. We analyzed 7,077 user utterances, each carefully annotated according to six cognitive levels and eleven dialogue act types. These included a variety of functions, such as asking for information, requesting translations, making cultural inquiries, and using language creatively. Pragmatic classifications further highlight how different types of dialogue acts--such as feedback, control commands, and social greetings--align with specific cognitive intentions. The results suggest that generative AI chatbots can support language learning in meaningful ways--especially when they are designed with an understanding of how users think and communicate. They may also help learners express themselves more confidently and connect with their cultural identity. The TALKA case provides empirical insights into how AI-mediated dialogue facilitates cognitive development in low-resource language learners, as well as pragmatic negotiation and socio-cultural affiliation. By focusing on AI-assisted language learning, this study offers new insights into how technology can support language preservation and educational practice.

Fun-ASR Technical Report

arXiv:2509.12508v3 Announce Type: replace Abstract: In recent years, automatic speech recognition (ASR) has witnessed transformative advancements driven by three complementary paradigms: data scaling, model size scaling, and deep integration with large language models (LLMs). However, LLMs are prone to hallucination, which can significantly degrade user experience in real-world ASR applications. In this paper, we present Fun-ASR, a large-scale, LLM-based ASR system that synergistically combines massive data, large model capacity, LLM integration, and reinforcement learning to achieve state-of-the-art performance across diverse and complex speech recognition scenarios. Moreover, Fun-ASR is specifically optimized for practical deployment, with enhancements in streaming capability, noise robustness, code-switching, hotword customization, and satisfying other real-world application requirements. Experimental results show that while most LLM-based ASR systems achieve strong performance on open-source benchmarks, they often underperform on real industry evaluation sets. Thanks to production-oriented optimizations, Fun-ASR achieves state-of-the-art performance on real application datasets, demonstrating its effectiveness and robustness in practical settings.

Reasoning over Boundaries: Enhancing Specification Alignment via Test-time Deliberation

arXiv:2509.14760v2 Announce Type: replace Abstract: Large language models (LLMs) are increasingly applied in diverse real-world scenarios, each governed by bespoke behavioral and safety specifications (spec) custom-tailored by users or organizations. These spec, categorized into safety-spec and behavioral-spec, vary across scenarios and evolve with changing preferences and requirements. We formalize this challenge as specification alignment, focusing on LLMs' ability to follow dynamic, scenario-specific spec from both behavioral and safety perspectives. To address this challenge, we propose Align3, a lightweight method that employs Test-Time Deliberation (TTD) with hierarchical reflection and revision to reason over the specification boundaries. We further present SpecBench, a unified benchmark for measuring specification alignment, covering 5 scenarios, 103 spec, and 1,500 prompts. Experiments on 15 reasoning and 18 instruct models with several TTD methods, including Self-Refine, TPO, and MoreThink, yield three key findings: (i) test-time deliberation enhances specification alignment; (ii) Align3 advances the safety-helpfulness trade-off frontier with minimal overhead; (iii) SpecBench effectively reveals alignment gaps. These results highlight the potential of test-time deliberation as an effective strategy for reasoning over the real-world specification boundaries.

Rethinking the Role of Text Complexity in Language Model Pretraining

arXiv:2509.16551v2 Announce Type: replace Abstract: Improving pretraining data quality and size is known to boost downstream performance, but the role of text complexity--how hard a text is to read--remains less explored. We reduce surface-level complexity (shorter sentences, simpler words, simpler structure) while keeping core content approximately constant and ask: (i) How does complexity affect language modeling across model sizes? (ii) Can useful representations be learned from simpler text alone? (iii) How does pretraining text complexity influence downstream language understanding? We simplify human-written texts using a large language model, pretrain causal models (28M-500M) from scratch on original vs. simplified data, and evaluate them in fine-tuning and zero-shot setups. We find that perplexity is sensitive to the interaction between model capacity and text complexity--smaller models degrade far less on simpler texts--while text complexity has little impact on fine-tuning evaluations, with zero-shot evaluations indicating that simpler texts benefit performance on linguistic knowledge tasks, whereas more complex texts favor tasks requiring world knowledge and entity tracking. Our findings suggest that different types of data diversity affect transfer and zero-shot performance differently, providing insight into tailoring data curation to specific goals.

K-DeCore: Facilitating Knowledge Transfer in Continual Structured Knowledge Reasoning via Knowledge Decoupling

arXiv:2509.16929v5 Announce Type: replace Abstract: Continual Structured Knowledge Reasoning (CSKR) focuses on training models to handle sequential tasks, where each task involves translating natural language questions into structured queries grounded in structured knowledge. Existing general continual learning approaches face significant challenges when applied to this task, including poor generalization to heterogeneous structured knowledge and inefficient reasoning due to parameter growth as tasks increase. To address these limitations, we propose a novel CSKR framework, \textsc{K-DeCore}, which operates with a fixed number of tunable parameters. Unlike prior methods, \textsc{K-DeCore} introduces a knowledge decoupling mechanism that disentangles the reasoning process into task-specific and task-agnostic stages, effectively bridging the gaps across diverse tasks. Building on this foundation, \textsc{K-DeCore} integrates a dual-perspective memory consolidation mechanism for distinct stages and introduces a structure-guided pseudo-data synthesis strategy to further enhance the model's generalization capabilities. Extensive experiments on four benchmark datasets demonstrate the superiority of \textsc{K-DeCore} over existing continual learning methods across multiple metrics, leveraging various backbone large language models.

Instruction Boundary: Quantifying Biases in LLM Reasoning under Various Coverage

arXiv:2509.20278v2 Announce Type: replace Abstract: Nowadays, automatically generated datasets are increasingly used in LLM reasoning tasks; however, large-scale corpora often contain inherent flaws. For example, a single-choice question may include none or multiple correct options, while true-or-false questions may involve vague or unverifiable statements. We refer to these exceptional answer forms as sparse labels. To compare LLMs' ability to recognize various question forms and produce correct answers, we investigate how different instruction formats can either facilitate or mislead LLM reasoning ability. We introduce the concept of Instruction Boundary, which systematically analyzes how different levels of prompt coverage -- sufficient, redundant, or insufficient -- can lead to reasoning biases and performance changes in LLMs. To examine this phenomenon, we design eight experimental settings across five dataset forms. We further propose BiasDetector, a unified framework that quantifies LLMs' ability to identify sparse labels under different kinds of Instruction Boundary conditions. Evaluations on five mainstream LLMs show that, despite their seemingly high accuracy, substantial reasoning biases persist in many downstream tasks as a direct consequence of prompt coverage. We analyze the impact of these biases and outline possible mitigation strategies. Our findings highlight not only the importance of addressing sparse labels, but also the need for developers to recognize and mitigate the risks introduced by Instruction Boundary.

COSPADI: Compressing LLMs via Calibration-Guided Sparse Dictionary Learning

arXiv:2509.22075v2 Announce Type: replace Abstract: Post-training compression of large language models (LLMs) largely relies on low-rank weight approximation, which represents each column of a weight matrix in a shared low-dimensional subspace. While this is a computationally efficient strategy, the imposed structural constraint is rigid and can lead to a noticeable model accuracy drop. In this work, we propose CoSpaDi (Compression via Sparse Dictionary Learning), a novel training-free compression framework that replaces low-rank decomposition with a more flexible structured sparse factorization in which each weight matrix is represented with a dense dictionary and a column-sparse coefficient matrix. This formulation enables a union-of-subspaces representation: different columns of the original weight matrix are approximated in distinct subspaces spanned by adaptively selected dictionary atoms, offering greater expressiveness than a single invariant basis. Crucially, CoSpaDi leverages a small calibration dataset to optimize the factorization such that the output activations of compressed projection layers closely match those of the original ones, thereby minimizing functional reconstruction error rather than mere weight approximation. This data-aware strategy preserves better model fidelity without any fine-tuning under reasonable compression ratios. Moreover, the resulting structured sparsity allows efficient sparse-dense matrix multiplication and is compatible with post-training quantization for further memory and latency gains. We evaluate CoSpaDi across multiple Llama and Qwen models under per-layer and per-group settings at 20-50\% compression ratios, demonstrating consistent superiority over state-of-the-art data-aware low-rank methods both in accuracy and perplexity. Our results establish structured sparse dictionary learning as a powerful alternative to conventional low-rank approaches for efficient LLM deployment.

ML2B: Multi-Lingual ML Benchmark For AutoML

arXiv:2509.22768v2 Announce Type: replace Abstract: Large language models (LLMs) have recently demonstrated strong capabilities in generating machine learning (ML) code, enabling end-to-end pipeline construction from natural language instructions. However, existing benchmarks for ML code generation are mainly restricted to English, overlooking the global and multilingual nature of ML research and practice. To address this gap, we present ML2B, the first benchmark for evaluating multilingual ML code generation. ML2B consists of 30 Kaggle competitions translated into 13 natural languages, covering tabular, text, and image data types, with structured metadata and validated human-reviewed translations. For evaluation, we employ AIDE, an automated framework for end-to-end assessment of data science pipelines, and provide insights into cross-lingual model performance. Our results reveal substantial 15-45% performance degradation on non-English tasks, highlighting critical challenges in multilingual representation learning for code generation. The benchmark, evaluation framework, and comprehensive results are made available through our GitHub repository to facilitate future research in multilingual ML code generation: https://github.com/enaix/ml2b.

HEART: Emotionally-driven test-time scaling of Language Models

arXiv:2509.22876v2 Announce Type: replace Abstract: Test-time scaling has shown considerable success in improving the performance of language models on complex reasoning tasks without requiring fine-tuning. However, current strategies such as self-reflection primarily focus on logical or structural refinement. They do not leverage the guiding potential of affective feedback. Inspired by psychological research showing that emotions can modulate cognitive performance, we introduce HEART--a novel framework that uses emotionally-driven prompts for iterative self-correction. HEART provides feedback on a model's incorrect response using a curated set of concise, emotionally charged phrases based on the six universal emotions categorized by Dr. Paul Ekman. By systematically varying the emotional tone of the feedback across iterations, our method guides the model to escape flawed reasoning paths and explore more promising alternatives. We evaluate our framework on challenging reasoning benchmarks including OlympiadBench, Humanity's Last Exam, and SimpleQA. Our results reveal a significant new phenomenon: when guided by an oracle verifier, this affective iteration protocol unlocks significantly deeper reasoning, leading to consistent and substantial increases in accuracy over state-of-the-art baselines with the same verifier. However, we also identify a critical bottleneck for practical deployment. In a verifier-free setting, it struggles to harness these gains consistently, highlighting as a key challenge for future work. Our findings suggest that the next frontier in machine reasoning may lie not just in refining logic, but also in understanding and leveraging the `HEART' of the models.

Non-Collaborative User Simulators for Tool Agents

arXiv:2509.23124v2 Announce Type: replace Abstract: Tool agents interact with users through multi-turn dialogues to accomplish various tasks. Recent studies have adopted user simulation methods to develop these agents in multi-turn settings. However, existing user simulators tend to be agent-friendly, exhibiting only cooperative behaviors, which fails to train and test agents against non-collaborative users in the real world. To address this, we propose a novel user simulator architecture that simulates four categories of non-collaborative behaviors: requesting unavailable services, digressing into tangential conversations, expressing impatience, and providing incomplete utterances. Our user simulator can simulate challenging and natural non-collaborative behaviors while reliably delivering all intents and information necessary to accomplish the task. Our experiments on MultiWOZ and $\tau$-bench reveal significant performance degradation in state-of-the-art tool agents when encountering non-collaborative users. We provide detailed analyses of agents' weaknesses under each non-collaborative condition, such as escalated hallucinations and dialogue breakdowns. Ultimately, we contribute an easily extensible user simulation framework to help the research community develop tool agents and preemptively diagnose them under challenging real-world conditions within their own services.

HiKE: Hierarchical Evaluation Framework for Korean-English Code-Switching Speech Recognition

arXiv:2509.24613v2 Announce Type: replace Abstract: Despite advances in multilingual automatic speech recognition (ASR), code-switching (CS), the mixing of languages within an utterance common in daily speech, remains a severely underexplored challenge. In this paper, we introduce HiKE: the Hierarchical Korean-English code-switching benchmark, the first globally accessible evaluation framework for Korean-English CS, aiming to provide a means for the precise evaluation of multilingual ASR models and to foster research in the field. The proposed framework not only consists of high-quality, natural CS data across various topics, but also provides meticulous loanword labels and a hierarchical CS-level labeling scheme (word, phrase, and sentence) that together enable a systematic evaluation of a model's ability to handle each distinct level of code-switching. Through evaluations of diverse multilingual ASR models and fine-tuning experiments, this paper demonstrates that although most multilingual ASR models initially exhibit inadequate CS-ASR performance, this capability can be enabled through fine-tuning with synthetic CS data. HiKE is available at https://github.com/ThetaOne-AI/HiKE

jina-reranker-v3: Last but Not Late Interaction for Listwise Document Reranking

arXiv:2509.25085v4 Announce Type: replace Abstract: jina-reranker-v3 is a 0.6B-parameter multilingual listwise reranker that introduces a novel "last but not late" interaction. Unlike late interaction models like ColBERT that encode documents separately before multi-vector matching, our approach applies causal attention between the query and all candidate documents in the same context window, enabling rich interactions before extracting contextual embeddings from each document's final token. The new model achieves state-of-the-art BEIR performance with 61.94 nDCG@10 while being significantly smaller than other models with comparable performance.

Latent Thinking Optimization: Your Latent Reasoning Language Model Secretly Encodes Reward Signals in Its Latent Thoughts

arXiv:2509.26314v2 Announce Type: replace Abstract: Large Language Models (LLMs) excel at problem solving by generating chain of thoughts in natural language, but such verbal thinking is computationally costly and prone to overthinking. Recent work instead proposes a latent thinking architecture Huginn-3.5B, which represents intermediate reasoning steps as sequence of latent representations. However, latent thoughts lack interpretability and are difficult to supervise, raising concerns about the correctness and reliability of its latent thinking processes. In this paper, we provide a systematic study of how Huginn-3.5B thinks in the latent space and how external supervision signals can improve its latent thinking processes. We show that latent thoughts leading to correct versus incorrect answers exhibit highly distinguishable patterns, and that a latent classifier can reliably predict answer correctness directly from latent thoughts. Leveraging these insights, we propose Latent Thinking Optimization (LTO), a probabilistic algorithm that employs the latent classifier as a Latent Reward Model (LRM) to optimize the latent thinking processes. Extensive experiments across diverse reasoning tasks demonstrate that LRM is highly effective in detecting incorrect latent thinking patterns, and LTO can significantly improve the latent thinking processes. Furthermore, we show that LRM can generalize across diverse domains, and LTO can be seamlessly applied to general LLMs to improve their thinking processes. In contrast to verbal thinking, our method demonstrates that reward modeling and scaling test-time thinking with supervision can be performed directly in the latent space, highlighting its potential as a general, efficient, and domain-agnostic approach to improving the thinking processes of LLMs.

Text-Based Approaches to Item Alignment to Content Standards in Large-Scale Reading & Writing Tests

arXiv:2509.26431v2 Announce Type: replace Abstract: Aligning test items to content standards is a critical step in test development to collect validity evidence based on content. Item alignment has typically been conducted by human experts. This judgmental process can be subjective and time-consuming. This study investigated the performance of fine-tuned small language models (SLMs) for automated item alignment using data from a large-scale standardized reading and writing test for college admissions. Different SLMs were trained for alignment at both domain and skill levels respectively with 10 skills mapped to 4 content domains. The model performance was evaluated in multiple criteria on two testing datasets. The impact of types and sizes of the input data for training was investigated. Results showed that including more item text data led to substantially better model performance, surpassing the improvements induced by sample size increase alone. For comparison, supervised machine learning models were trained using the embeddings from the multilingual-E5-large-instruct model. The study results showed that fine-tuned SLMs consistently outperformed the embedding-based supervised machine learning models, particularly for the more fine-grained skill alignment. To better understand model misclassifications, multiple semantic similarity analysis including pairwise cosine similarity, Kullback-Leibler divergence of embedding distributions, and two-dimension projections of item embeddings were conducted. These analyses consistently showed that certain skills in SAT and PSAT were semantically too close, providing evidence for the observed misclassification.

Verbalized Sampling: How to Mitigate Mode Collapse and Unlock LLM Diversity

arXiv:2510.01171v2 Announce Type: replace Abstract: Post-training alignment often reduces LLM diversity, leading to a phenomenon known as mode collapse. Unlike prior work that attributes this effect to algorithmic limitations, we identify a fundamental, pervasive data-level driver: typicality bias in preference data, whereby annotators systematically favor familiar text as a result of well-established findings in cognitive psychology. We formalize this bias theoretically, verify it on preference datasets empirically, and show that it plays a central role in mode collapse. Motivated by this analysis, we introduce Verbalized Sampling, a simple, training-free prompting strategy to circumvent mode collapse. VS prompts the model to verbalize a probability distribution over a set of responses (e.g., "Generate 5 jokes about coffee and their corresponding probabilities"). Comprehensive experiments show that VS significantly improves performance across creative writing (poems, stories, jokes), dialogue simulation, open-ended QA, and synthetic data generation, without sacrificing factual accuracy and safety. For instance, in creative writing, VS increases diversity by 1.6-2.1x over direct prompting. We further observe an emergent trend that more capable models benefit more from VS. In sum, our work provides a new data-centric perspective on mode collapse and a practical inference-time remedy that helps unlock pre-trained generative diversity.

Silent Tokens, Loud Effects: Padding in LLMs

arXiv:2510.01238v2 Announce Type: replace Abstract: Padding tokens are widely used in large language models (LLMs) to equalize sequence lengths during batched inference. While they should be fully masked, implementation errors can cause them to influence computation, and the extent of this influence is not well understood. We systematically study this effect across three open-source model families (Llama, Gemma, Qwen), inserting controlled amounts of padding and evaluating outcomes along four axes: activations, generation quality, bias, and safety. Even small amounts of padding shift hidden representations, degrade quality in smaller models, alter bias in unpredictable ways, and weaken safety guardrails. These findings demonstrate that padding is not a harmless detail but a robustness risk that must be carefully handled in deployment.

AMAS: Adaptively Determining Communication Topology for LLM-based Multi-Agent System

arXiv:2510.01617v2 Announce Type: replace Abstract: Although large language models (LLMs) have revolutionized natural language processing capabilities, their practical implementation as autonomous multi-agent systems (MAS) for industrial problem-solving encounters persistent barriers. Conventional MAS architectures are fundamentally restricted by inflexible, hand-crafted graph topologies that lack contextual responsiveness, resulting in diminished efficacy across varied academic and commercial workloads. To surmount these constraints, we introduce AMAS, a paradigm-shifting framework that redefines LLM-based MAS through a novel dynamic graph designer. This component autonomously identifies task-specific optimal graph configurations via lightweight LLM adaptation, eliminating the reliance on monolithic, universally applied structural templates. Instead, AMAS exploits the intrinsic properties of individual inputs to intelligently direct query trajectories through task-optimized agent pathways. Rigorous validation across question answering, mathematical deduction, and code generation benchmarks confirms that AMAS systematically exceeds state-of-the-art single-agent and multi-agent approaches across diverse LLM architectures. Our investigation establishes that context-sensitive structural adaptability constitutes a foundational requirement for high-performance LLM MAS deployments.

Format Inertia: A Failure Mechanism of LLMs in Medical Pre-Consultation

arXiv:2510.01688v2 Announce Type: replace Abstract: Recent advances in Large Language Models (LLMs) have brought significant improvements to various service domains, including chatbots and medical pre-consultation applications. In the healthcare domain, the most common approach for adapting LLMs to multi-turn dialogue generation is Supervised Fine-Tuning (SFT). However, datasets for SFT in tasks like medical pre-consultation typically exhibit a skewed turn-count distribution. Training on such data induces a novel failure mechanism we term Format Inertia, where models tend to generate repetitive, format-correct, but diagnostically uninformative questions in long medical dialogues. To mitigate this observed failure mechanism, we adopt a simple, data-centric method that rebalances the turn-count distribution of the training dataset. Experimental results show that our approach substantially alleviates Format Inertia in medical pre-consultation.

Veri-R1: Toward Precise and Faithful Claim Verification via Online Reinforcement Learning

arXiv:2510.01932v2 Announce Type: replace Abstract: Claim verification with large language models (LLMs) has recently attracted growing attention, due to their strong reasoning capabilities and transparent verification processes compared to traditional answer-only judgments. However, existing approaches to online claim verification, which requires iterative evidence retrieval and reasoning, still mainly rely on prompt engineering or pre-designed reasoning workflows, without unified training to improve necessary skills. Therefore, we introduce Veri-R1, an online reinforcement learning (RL) framework that enables an LLM to interact with a search engine and to receive reward signals that explicitly shape its planning, retrieval, and reasoning behaviors. This dynamic interaction of LLM with retrieval systems more accurately reflects real-world verification scenarios and fosters comprehensive verification skills. Empirical results show that Veri-R1 improves joint accuracy by up to 30% and doubles the evidence score, often surpassing its larger-scale model counterparts. Ablation studies further reveal the impact of reward components, and the link between output logits and label accuracy. Our results highlight the effectiveness of online RL for precise and faithful claim verification, providing an important foundation for future research. We release our code to support community progress in LLM empowered claim verification.

InfoMosaic-Bench: Evaluating Multi-Source Information Seeking in Tool-Augmented Agents

arXiv:2510.02271v2 Announce Type: replace Abstract: Information seeking is a fundamental requirement for humans. However, existing LLM agents rely heavily on open-web search, which exposes two fundamental weaknesses: online content is noisy and unreliable, and many real-world tasks require precise, domain-specific knowledge unavailable from the web. The emergence of the Model Context Protocol (MCP) now allows agents to interface with thousands of specialized tools, seemingly resolving this limitation. Yet it remains unclear whether agents can effectively leverage such tools -- and more importantly, whether they can integrate them with general-purpose search to solve complex tasks. Therefore, we introduce InfoMosaic-Bench, the first benchmark dedicated to multi-source information seeking in tool-augmented agents. Covering six representative domains (medicine, finance, maps, video, web, and multi-domain integration), InfoMosaic-Bench requires agents to combine general-purpose search with domain-specific tools. Tasks are synthesized with InfoMosaic-Flow, a scalable pipeline that grounds task conditions in verified tool outputs, enforces cross-source dependencies, and filters out shortcut cases solvable by trivial lookup. This design guarantees both reliability and non-triviality. Experiments with 14 state-of-the-art LLM agents reveal three findings: (i) web information alone is insufficient, with GPT-5 achieving only 38.2% accuracy and 67.5% pass rate; (ii) domain tools provide selective but inconsistent benefits, improving some domains while degrading others; and (iii) 22.4% of failures arise from incorrect tool usage or selection, highlighting that current LLMs still struggle with even basic tool handling.

Beyond Manuals and Tasks: Instance-Level Context Learning for LLM Agents

arXiv:2510.02369v2 Announce Type: replace Abstract: Large language model (LLM) agents typically receive two kinds of context: (i) environment-level manuals that define interaction interfaces and global rules, and (ii) task-level guidance or demonstrations tied to specific goals. In this work, we identify a crucial but overlooked third type of context, instance-level context, which consists of verifiable and reusable facts tied to a specific environment instance, such as object locations, crafting recipes, and local rules. We argue that the absence of instance-level context is a common source of failure for LLM agents in complex tasks, as success often depends not only on reasoning over global rules or task prompts but also on making decisions based on precise and persistent facts. Acquiring such context requires more than memorization: the challenge lies in efficiently exploring, validating, and formatting these facts under tight interaction budgets. We formalize this problem as Instance-Level Context Learning (ILCL) and introduce our task-agnostic method to solve it. Our method performs a guided exploration, using a compact TODO forest to intelligently prioritize its next actions and a lightweight plan-act-extract loop to execute them. This process automatically produces a high-precision context document that is reusable across many downstream tasks and agents, thereby amortizing the initial exploration cost. Experiments across TextWorld, ALFWorld, and Crafter demonstrate consistent gains in both success and efficiency: for instance, ReAct's mean success rate in TextWorld rises from 37% to 95%, while IGE improves from 81% to 95%. By transforming one-off exploration into persistent, reusable knowledge, our method complements existing contexts to enable more reliable and efficient LLM agents.

Pretraining with hierarchical memories: separating long-tail and common knowledge

arXiv:2510.02375v2 Announce Type: replace Abstract: The impressive performance gains of modern language models currently rely on scaling parameters: larger models store more world knowledge and reason better. Yet compressing all world knowledge into parameters is unnecessary, as only a fraction is used per prompt, and impractical for edge devices with limited inference-time memory and compute. We address this shortcoming by a memory-augmented architecture and a pretraining strategy aligned with existing hardware paradigms. We introduce small language models that access large hierarchical parametric memory banks encoding world knowledge. During pretraining and inference, we fetch a small, context-dependent memory block and add it to the model. Our pretraining learns to store long-tail world knowledge in the memory parameters, while the small language model acts as an anchor capturing common knowledge and general reasoning abilities. Through trillion-token-scale experiments, we show significant gains: a 160M-parameters model augmented with an 18M-parameters memory fetched from a 4.6B memory bank obtains comparable performance to a regular model with more than 2x the parameters. Through extensive experiments, we study the optimal type and size of parametric memories in transformers, scaling them to over 21B parameters. We find that our proposed hierarchical feed-forward memories work robustly across transformer architectures, whether added during pretraining or post-hoc.

Semantic Similarity in Radiology Reports via LLMs and NER

arXiv:2510.03102v2 Announce Type: replace Abstract: Radiology report evaluation is a crucial part of radiologists' training and plays a key role in ensuring diagnostic accuracy. As part of the standard reporting workflow, a junior radiologist typically prepares a preliminary report, which is then reviewed and edited by a senior radiologist to produce the final report. Identifying semantic differences between preliminary and final reports is essential for junior doctors, both as a training tool and to help uncover gaps in clinical knowledge. While AI in radiology is a rapidly growing field, the application of large language models (LLMs) remains challenging due to the need for specialised domain knowledge. In this paper, we explore the ability of LLMs to provide explainable and accurate comparisons of reports in the radiology domain. We begin by comparing the performance of several LLMs in comparing radiology reports. We then assess a more traditional approach based on Named-Entity-Recognition (NER). However, both approaches exhibit limitations in delivering accurate feedback on semantic similarity. To address this, we propose Llama-EntScore, a semantic similarity scoring method using a combination of Llama 3.1 and NER with tunable weights to emphasise or de-emphasise specific types of differences. Our approach generates a quantitative similarity score for tracking progress and also gives an interpretation of the score that aims to offer valuable guidance in reviewing and refining their reporting. We find our method achieves 67% exact-match accuracy and 93% accuracy within +/- 1 when compared to radiologist-provided ground truth scores - outperforming both LLMs and NER used independently. Code is available at: https://github.com/otmive/llama_reports

SurveyBench: Can LLM(-Agents) Write Academic Surveys that Align with Reader Needs?

arXiv:2510.03120v2 Announce Type: replace Abstract: Academic survey writing, which distills vast literature into a coherent and insightful narrative, remains a labor-intensive and intellectually demanding task. While recent approaches, such as general DeepResearch agents and survey-specialized methods, can generate surveys automatically (a.k.a. LLM4Survey), their outputs often fall short of human standards and there lacks a rigorous, reader-aligned benchmark for thoroughly revealing their deficiencies. To fill the gap, we propose a fine-grained, quiz-driven evaluation framework SurveyBench, featuring (1) typical survey topics source from recent 11,343 arXiv papers and corresponding 4,947 high-quality surveys; (2) a multifaceted metric hierarchy that assesses the outline quality (e.g., coverage breadth, logical coherence), content quality (e.g., synthesis granularity, clarity of insights), and non-textual richness; and (3) a dual-mode evaluation protocol that includes content-based and quiz-based answerability tests, explicitly aligned with readers' informational needs. Results show SurveyBench effectively challenges existing LLM4Survey approaches (e.g., on average 21% lower than human in content-based evaluation).

AgentBench: Evaluating LLMs as Agents

arXiv:2308.03688v3 Announce Type: replace-cross Abstract: The potential of Large Language Model (LLM) as agents has been widely acknowledged recently. Thus, there is an urgent need to quantitatively \textit{evaluate LLMs as agents} on challenging tasks in interactive environments. We present AgentBench, a multi-dimensional benchmark that consists of 8 distinct environments to assess LLM-as-Agent's reasoning and decision-making abilities. Our extensive test over \num API-based and open-sourced (OSS) LLMs shows that, while top commercial LLMs present a strong ability of acting as agents in complex environments, there is a significant disparity in performance between them and many OSS competitors that are no larger than 70B. We identify the typical reasons of failures in environments and LLMs, showing that poor long-term reasoning, decision-making, and instruction following abilities are the main obstacles for developing usable LLM agents. Improving instruction following and training on high quality multi-round alignment data could improve agent performance. And different from existing assumptions, training on code present ambivalent impacts on different agent tasks. Datasets, environments, and an integrated evaluation package for AgentBench are released at https://github.com/THUDM/AgentBench.

Can Large Language Models generalize analogy solving like children can?

arXiv:2411.02348v3 Announce Type: replace-cross Abstract: In people, the ability to solve analogies such as "body : feet :: table : ?" emerges in childhood, and appears to transfer easily to other domains, such as the visual domain "( : ) :: < : ?". Recent research shows that large language models (LLMs) can solve various forms of analogies. However, can LLMs generalize analogy solving to new domains like people can? To investigate this, we had children, adults, and LLMs solve a series of letter-string analogies (e.g., a b : a c :: j k : ?) in the Latin alphabet, in a near transfer domain (Greek alphabet), and a far transfer domain (list of symbols). Children and adults easily generalized their knowledge to unfamiliar domains, whereas LLMs did not. This key difference between human and AI performance is evidence that these LLMs still struggle with robust human-like analogical transfer.

A Statistical Hypothesis Testing Framework for Data Misappropriation Detection in Large Language Models

arXiv:2501.02441v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are rapidly gaining enormous popularity in recent years. However, the training of LLMs has raised significant privacy and legal concerns, particularly regarding the distillation and inclusion of copyrighted materials in their training data without proper attribution or licensing, an issue that falls under the broader concern of data misappropriation. In this article, we focus on a specific problem of data misappropriation detection, namely, to determine whether a given LLM has incorporated the data generated by another LLM. We propose embedding watermarks into the copyrighted training data and formulating the detection of data misappropriation as a hypothesis testing problem. We develop a general statistical testing framework, construct test statistics, determine optimal rejection thresholds, and explicitly control type I and type II errors. Furthermore, we establish the asymptotic optimality properties of the proposed tests, and demonstrate the empirical effectiveness through intensive numerical experiments.

CBVLM: Training-free Explainable Concept-based Large Vision Language Models for Medical Image Classification

arXiv:2501.12266v3 Announce Type: replace-cross Abstract: The main challenges limiting the adoption of deep learning-based solutions in medical workflows are the availability of annotated data and the lack of interpretability of such systems. Concept Bottleneck Models (CBMs) tackle the latter by constraining the model output on a set of predefined and human-interpretable concepts. However, the increased interpretability achieved through these concept-based explanations implies a higher annotation burden. Moreover, if a new concept needs to be added, the whole system needs to be retrained. Inspired by the remarkable performance shown by Large Vision-Language Models (LVLMs) in few-shot settings, we propose a simple, yet effective, methodology, CBVLM, which tackles both of the aforementioned challenges. First, for each concept, we prompt the LVLM to answer if the concept is present in the input image. Then, we ask the LVLM to classify the image based on the previous concept predictions. Moreover, in both stages, we incorporate a retrieval module responsible for selecting the best examples for in-context learning. By grounding the final diagnosis on the predicted concepts, we ensure explainability, and by leveraging the few-shot capabilities of LVLMs, we drastically lower the annotation cost. We validate our approach with extensive experiments across four medical datasets and twelve LVLMs (both generic and medical) and show that CBVLM consistently outperforms CBMs and task-specific supervised methods without requiring any training and using just a few annotated examples. More information on our project page: https://cristianopatricio.github.io/CBVLM/.

Fed-SB: A Silver Bullet for Extreme Communication Efficiency and Performance in (Private) Federated LoRA Fine-Tuning

arXiv:2502.15436v2 Announce Type: replace-cross Abstract: Low-Rank Adaptation (LoRA) has become ubiquitous for efficiently fine-tuning foundation models. However, federated fine-tuning using LoRA is challenging due to suboptimal updates arising from traditional federated averaging of individual adapters. Existing solutions either incur prohibitively high communication cost that scales linearly with the number of clients or suffer from performance degradation due to limited expressivity. We introduce Federated Silver Bullet (Fed-SB), a novel approach for federated fine-tuning of LLMs using LoRA-SB, a recently proposed low-rank adaptation method. LoRA-SB optimally aligns the optimization trajectory with the ideal low-rank full fine-tuning projection by learning a small square matrix (R) between adapters B and A, keeping other components fixed. Direct averaging of R guarantees exact updates, substantially reducing communication cost, which remains independent of the number of clients, and enables scalability. Fed-SB achieves state-of-the-art performance across commonsense reasoning, arithmetic reasoning, and language inference tasks while reducing communication costs by up to 230x. In private settings, Fed-SB further improves performance by (1) reducing trainable parameters, thereby lowering the noise required for differential privacy and (2) avoiding noise amplification introduced by other methods. Overall, Fed-SB offers a state-of-the-art, efficient, and scalable solution for both private and non-private federated fine-tuning. Our code is publicly available at: https://github.com/CERT-Lab/fed-sb.

DISC: Dynamic Decomposition Improves LLM Inference Scaling

arXiv:2502.16706v3 Announce Type: replace-cross Abstract: Inference scaling methods for LLMs often rely on decomposing problems into steps (or groups of tokens), followed by sampling and selecting the best next steps. However, these steps and their sizes are often predetermined or manually designed based on domain knowledge. We propose dynamic decomposition, a method that adaptively and automatically partitions solution and reasoning traces into manageable steps during inference. By more effectively allocating compute -- particularly through subdividing challenging steps and prioritizing their sampling -- dynamic decomposition significantly improves inference efficiency. Experiments on benchmarks such as APPS, MATH, and LiveCodeBench demonstrate that dynamic decomposition outperforms static approaches, including token-level, sentence-level, and single-step decompositions, reducing the pass@10 error rate by 5.0%, 6.7%, and 10.5% respectively. These findings highlight the potential of dynamic decomposition to improve a wide range of inference scaling techniques.

Seeded Poisson Factorization: leveraging domain knowledge to fit topic models

arXiv:2503.02741v2 Announce Type: replace-cross Abstract: Topic models are widely used for discovering latent thematic structures in large text corpora, yet traditional unsupervised methods often struggle to align with pre-defined conceptual domains. This paper introduces seeded Poisson Factorization (SPF), a novel approach that extends the Poisson Factorization (PF) framework by incorporating domain knowledge through seed words. SPF enables a structured topic discovery by modifying the prior distribution of topic-specific term intensities, assigning higher initial rates to pre-defined seed words. The model is estimated using variational inference with stochastic gradient optimization, ensuring scalability to large datasets. We present in detail the results of applying SPF to an Amazon customer feedback dataset, leveraging pre-defined product categories as guiding structures. SPF achieves superior performance compared to alternative guided probabilistic topic models in terms of computational efficiency and classification performance. Robustness checks highlight SPF's ability to adaptively balance domain knowledge and data-driven topic discovery, even in case of imperfect seed word selection. Further applications of SPF to four additional benchmark datasets, where the corpus varies in size and the number of topics differs, demonstrate its general superior classification performance compared to the unseeded PF model.

Understanding R1-Zero-Like Training: A Critical Perspective

arXiv:2503.20783v2 Announce Type: replace-cross Abstract: DeepSeek-R1-Zero has shown that reinforcement learning (RL) at scale can directly enhance the reasoning capabilities of LLMs without supervised fine-tuning. In this work, we critically examine R1-Zero-like training by analyzing its two core components: base models and RL. We investigate a wide range of base models, including DeepSeek-V3-Base, to understand how pretraining characteristics influence RL performance. Our analysis reveals that DeepSeek-V3-Base already exhibit ''Aha moment'', while Qwen2.5 base models demonstrate strong reasoning capabilities even without prompt templates, suggesting potential pretraining biases. Additionally, we identify an optimization bias in Group Relative Policy Optimization (GRPO), which artificially increases response length (especially for incorrect outputs) during training. To address this, we introduce Dr. GRPO, an unbiased optimization method that improves token efficiency while maintaining reasoning performance. Leveraging these insights, we present a minimalist R1-Zero recipe that achieves 43.3% accuracy on AIME 2024 with a 7B base model, establishing a new state-of-the-art. Our code is available at https://github.com/sail-sg/understand-r1-zero.

TRA: Better Length Generalisation with Threshold Relative Attention

arXiv:2503.23174v4 Announce Type: replace-cross Abstract: Transformers struggle with length generalisation, displaying poor performance even on basic tasks. We test whether these limitations can be explained through two key failures of the self-attention mechanism. The first is the inability to fully remove irrelevant information. The second is tied to position, even if the dot product between a key and query is highly negative (i.e. an irrelevant key) learned positional biases may unintentionally up-weight such information - dangerous when distances become out of distribution. Put together, these two failure cases lead to compounding generalisation difficulties. We test whether they can be mitigated through the combination of a) selective sparsity - completely removing irrelevant keys from the attention softmax and b) contextualised relative distance - distance is only considered as between the query and the keys that matter. We show how refactoring the attention mechanism with these two mitigations in place can substantially improve the generalisation capabilities of decoder only transformers.

On the Effectiveness and Generalization of Race Representations for Debiasing High-Stakes Decisions

arXiv:2504.06303v2 Announce Type: replace-cross Abstract: Understanding and mitigating biases is critical for the adoption of large language models (LLMs) in high-stakes decision-making. We introduce Admissions and Hiring, decision tasks with hypothetical applicant profiles where a person's race can be inferred from their name, as simplified test beds for racial bias. We show that Gemma 2B Instruct and LLaMA 3.2 3B Instruct exhibit strong biases. Gemma grants admission to 26% more White than Black applicants, and LLaMA hires 60% more Asian than White applicants. We demonstrate that these biases are resistant to prompt engineering: multiple prompting strategies all fail to promote fairness. In contrast, using distributed alignment search, we can identify "race subspaces" within model activations and intervene on them to debias model decisions. Averaging the representation across all races within the subspaces reduces Gemma's bias by 37-57%. Finally, we examine the generalizability of Gemma's race subspaces, and find limited evidence for generalization, where changing the prompt format can affect the race representation. Our work suggests mechanistic approaches may provide a promising venue for improving the fairness of LLMs, but a universal race representation remains elusive.

AgentRewardBench: Evaluating Automatic Evaluations of Web Agent Trajectories

arXiv:2504.08942v2 Announce Type: replace-cross Abstract: Web agents enable users to perform tasks on web browsers through natural language interaction. Evaluating web agents trajectories is an important problem, since it helps us determine whether the agent successfully completed the tasks. Rule-based methods are widely used for this purpose, but they are challenging to extend to new tasks and may not always recognize successful trajectories. We may achieve higher accuracy through human evaluation, but the process would be substantially slower and more expensive. Automatic evaluations with LLMs may avoid the challenges of designing new rules and manually annotating trajectories, enabling faster and cost-effective evaluation. However, it is unclear how effective they are at evaluating web agents. To this end, we propose AgentRewardBench, the first benchmark to assess the effectiveness of LLM judges for evaluating web agents. AgentRewardBench contains 1302 trajectories across 5 benchmarks and 4 LLMs. Each trajectory in AgentRewardBench is reviewed by an expert, who answers questions pertaining to the success, side effects, and repetitiveness of the agent. Using our benchmark, we evaluate 12 LLM judges and find that no single LLM excels across all benchmarks. We also find that the rule-based evaluation used by common benchmarks tends to underreport the success rate of web agents, highlighting a key weakness of rule-based evaluation and the need to develop more flexible automatic evaluations. We release the benchmark at: https://agent-reward-bench.github.io

From Dialect Gaps to Identity Maps: Tackling Variability in Speaker Verification

arXiv:2505.04629v2 Announce Type: replace-cross Abstract: The complexity and difficulties of Kurdish speaker detection among its several dialects are investigated in this work. Because of its great phonetic and lexical differences, Kurdish with several dialects including Kurmanji, Sorani, and Hawrami offers special challenges for speaker recognition systems. The main difficulties in building a strong speaker identification system capable of precisely identifying speakers across several dialects are investigated in this work. To raise the accuracy and dependability of these systems, it also suggests solutions like sophisticated machine learning approaches, data augmentation tactics, and the building of thorough dialect-specific corpus. The results show that customized strategies for every dialect together with cross-dialect training greatly enhance recognition performance.

Language Agents Mirror Human Causal Reasoning Biases. How Can We Help Them Think Like Scientists?

arXiv:2505.09614v3 Announce Type: replace-cross Abstract: Language model (LM) agents are increasingly used as autonomous decision-makers which need to actively gather information to guide their decisions. A crucial cognitive skill for such agents is the efficient exploration and understanding of the causal structure of the world -- key to robust, scientifically grounded reasoning. Yet, it remains unclear whether LMs possess this capability or exhibit systematic biases leading to erroneous conclusions. In this work, we examine LMs' ability to explore and infer causal relationships, using the well-established Blicket Test paradigm from developmental psychology. We find that LMs reliably infer the common, intuitive disjunctive causal relationships but systematically struggle with the unusual, yet equally (or sometimes even more) evidenced conjunctive ones. This "disjunctive bias" persists across model families, sizes, and prompting strategies, and performance further declines as task complexity increases. Interestingly, an analogous bias appears in human adults, suggesting that LMs may have inherited deep-seated reasoning heuristics from their training data. To this end, we quantify similarities between LMs and humans, finding that LMs exhibit adult-like inference profiles (but not child-like). Finally, we propose a test-time sampling method which explicitly samples and eliminates hypotheses about causal relationships from the LM. This scalable approach significantly reduces the disjunctive bias and moves LMs closer to the goal of scientific, causally rigorous reasoning.

MMLongBench: Benchmarking Long-Context Vision-Language Models Effectively and Thoroughly

arXiv:2505.10610v3 Announce Type: replace-cross Abstract: The rapid extension of context windows in large vision-language models has given rise to long-context vision-language models (LCVLMs), which are capable of handling hundreds of images with interleaved text tokens in a single forward pass. In this work, we introduce MMLongBench, the first benchmark covering a diverse set of long-context vision-language tasks, to evaluate LCVLMs effectively and thoroughly. MMLongBench is composed of 13,331 examples spanning five different categories of downstream tasks, such as Visual RAG and Many-Shot ICL. It also provides broad coverage of image types, including various natural and synthetic images. To assess the robustness of the models to different input lengths, all examples are delivered at five standardized input lengths (8K-128K tokens) via a cross-modal tokenization scheme that combines vision patches and text tokens. Through a thorough benchmarking of 46 closed-source and open-source LCVLMs, we provide a comprehensive analysis of the current models' vision-language long-context ability. Our results show that: i) performance on a single task is a weak proxy for overall long-context capability; ii) both closed-source and open-source models face challenges in long-context vision-language tasks, indicating substantial room for future improvement; iii) models with stronger reasoning ability tend to exhibit better long-context performance. By offering wide task coverage, various image types, and rigorous length control, MMLongBench provides the missing foundation for diagnosing and advancing the next generation of LCVLMs.

TIME: A Multi-level Benchmark for Temporal Reasoning of LLMs in Real-World Scenarios

arXiv:2505.12891v3 Announce Type: replace-cross Abstract: Temporal reasoning is pivotal for Large Language Models (LLMs) to comprehend the real world. However, existing works neglect the real-world challenges for temporal reasoning: (1) intensive temporal information, (2) fast-changing event dynamics, and (3) complex temporal dependencies in social interactions. To bridge this gap, we propose a multi-level benchmark TIME, designed for temporal reasoning in real-world scenarios. TIME consists of 38,522 QA pairs, covering 3 levels with 11 fine-grained sub-tasks. This benchmark encompasses 3 sub-datasets reflecting different real-world challenges: TIME-Wiki, TIME-News, and TIME-Dial. We conduct extensive experiments on reasoning models and non-reasoning models. And we conducted an in-depth analysis of temporal reasoning performance across diverse real-world scenarios and tasks, and summarized the impact of test-time scaling on temporal reasoning capabilities. Additionally, we release TIME-Lite, a human-annotated subset to foster future research and standardized evaluation in temporal reasoning. The code is available at https://github.com/sylvain-wei/TIME , the dataset is available at https://huggingface.co/datasets/SylvainWei/TIME , and the project page link is https://sylvain-wei.github.io/TIME/ .

Safety Subspaces are Not Linearly Distinct: A Fine-Tuning Case Study

arXiv:2505.14185v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) rely on safety alignment to produce socially acceptable responses. However, this behavior is known to be brittle: further fine-tuning, even on benign or lightly contaminated data, can degrade safety and reintroduce harmful behaviors. A growing body of work suggests that alignment may correspond to identifiable directions in weight space, forming subspaces that could, in principle, be isolated or preserved to defend against misalignment. In this work, we conduct a comprehensive empirical study of this perspective. We examine whether safety-relevant behavior is concentrated in specific linear subspaces, whether it can be separated from general-purpose learning, and whether harmfulness arises from distinguishable patterns in activations. Across both weight and activation spaces, our findings are consistent: subspaces that amplify safe behaviors also amplify useful ones, and prompts with different safety implications activate overlapping representations. Rather than residing in distinct directions, we show that safety is highly entangled with the general learning components of the model. This suggests that subspace-based defenses face fundamental limitations and underscores the need for alternative strategies to preserve safety under continued training. We corroborate these findings with multiple experiments on five open-source LLMs from the Llama and Qwen families. Our code is publicly available at: https://github.com/CERT-Lab/safety-subspaces.

AdaSTaR: Adaptive Data Sampling for Training Self-Taught Reasoners

arXiv:2505.16322v3 Announce Type: replace-cross Abstract: Self-Taught Reasoners (STaR), synonymously known as Rejection sampling Fine-Tuning (RFT), is an integral part of the training pipeline of self-improving reasoning Language Models (LMs). The self-improving mechanism often employs random observation (data) sampling. However, this results in trained observation imbalance; inefficiently over-training on solved examples while under-training on challenging ones. In response, we introduce Adaptive STaR (AdaSTaR), a novel algorithm that rectifies this by integrating two adaptive sampling principles: (1) Adaptive Sampling for Diversity: promoting balanced training across observations, and (2) Adaptive Sampling for Curriculum: dynamically adjusting data difficulty to match the model's evolving strength. Across six benchmarks, AdaSTaR achieves best test accuracy in all instances (6/6) and reduces training FLOPs by an average of 58.6% against an extensive list of baselines. These improvements in performance and efficiency generalize to different pre-trained LMs and larger models, paving the way for more efficient and effective self-improving LMs.

COUNTDOWN: Contextually Sparse Activation Filtering Out Unnecessary Weights in Down Projection

arXiv:2505.17701v2 Announce Type: replace-cross Abstract: The growing size of large language models has created significant computational inefficiencies. To address this challenge, sparse activation methods selectively deactivates non-essential parameters during inference, reducing computational costs in FFNN layers. While existing methods focus on non-linear gating mechanisms, we hypothesize that the sparsity of the FFNN layer lies globally in the form of a linear combination over its internal down projection matrix. Based on this insight, we propose two methods: M-COUNTDOWN, leveraging indirect coefficients, and D-COUNTDOWN, utilizing direct coefficients of the linear combination. Experimental results demonstrate that D-COUNTDOWN can omit 90% of computations with performance loss as low as 5.5% ideally, while M-COUNTDOWN provides a predictor-free solution with up to 29.4% better performance preservation compared to existing methods. Our specialized kernel implementations effectively realize these theoretical gains into substantial real-world acceleration.

How Many Parameters Does Your Task Really Need? Task Specific Pruning with LLM-Sieve

arXiv:2505.18350v2 Announce Type: replace-cross Abstract: As Large Language Models (LLMs) are increasingly deployed for narrow tasks in resource-constrained settings, a central question arises: how much of an LLM is truly necessary for a given task? We present LLM-Sieve, a framework that prunes LLMs down to the minimal parameter subset needed to preserve task performance. Our approach introduces two innovations: (i) output-aligned non-orthogonal projections, which yield more faithful low-rank approximations than traditional PCA/SVD by aligning directly with layer outputs; and (ii) adaptive pruning via a Genetic Algorithm, which automatically discovers matrix-specific pruning levels and exposes the uneven distribution of task-relevant knowledge. Across models from 3.8B to 70B parameters, LLM-Sieve removes 20-75% of weights with only 1-5% accuracy loss-substantially ahead of prior pruning methods. Beyond efficiency, our framework reveals bottleneck matrices that concentrate critical knowledge, suggesting architectural implications for future LLM design. LLM-Sieve integrates seamlessly with LoRA fine-tuning and quantization, enabling both efficient deployment and deeper understanding of knowledge organization in LLMs.

Scientific Paper Retrieval with LLM-Guided Semantic-Based Ranking

arXiv:2505.21815v2 Announce Type: replace-cross Abstract: Scientific paper retrieval is essential for supporting literature discovery and research. While dense retrieval methods demonstrate effectiveness in general-purpose tasks, they often fail to capture fine-grained scientific concepts that are essential for accurate understanding of scientific queries. Recent studies also use large language models (LLMs) for query understanding; however, these methods often lack grounding in corpus-specific knowledge and may generate unreliable or unfaithful content. To overcome these limitations, we propose SemRank, an effective and efficient paper retrieval framework that combines LLM-guided query understanding with a concept-based semantic index. Each paper is indexed using multi-granular scientific concepts, including general research topics and detailed key phrases. At query time, an LLM identifies core concepts derived from the corpus to explicitly capture the query's information need. These identified concepts enable precise semantic matching, significantly enhancing retrieval accuracy. Experiments show that SemRank consistently improves the performance of various base retrievers, surpasses strong existing LLM-based baselines, and remains highly efficient.

Identity resolution of software metadata using Large Language Models

arXiv:2505.23500v2 Announce Type: replace-cross Abstract: Software is an essential component of research. However, little attention has been paid to it compared with that paid to research data. Recently, there has been an increase in efforts to acknowledge and highlight the importance of software in research activities. Structured metadata from platforms like bio.tools, Bioconductor, and Galaxy ToolShed offers valuable insights into research software in the Life Sciences. Although originally intended to support discovery and integration, this metadata can be repurposed for large-scale analysis of software practices. However, its quality and completeness vary across platforms, reflecting diverse documentation practices. To gain a comprehensive view of software development and sustainability, consolidating this metadata is necessary, but requires robust mechanisms to address its heterogeneity and scale. This article presents an evaluation of instruction-tuned large language models for the task of software metadata identity resolution, a critical step in assembling a cohesive collection of research software. Such a collection is the reference component for the Software Observatory at OpenEBench, a platform that aggregates metadata to monitor the FAIRness of research software in the Life Sciences. We benchmarked multiple models against a human-annotated gold standard, examined their behavior on ambiguous cases, and introduced an agreement-based proxy for high-confidence automated decisions. The proxy achieved high precision and statistical robustness, while also highlighting the limitations of current models and the broader challenges of automating semantic judgment in FAIR-aligned software metadata across registries and repositories.

Human Empathy as Encoder: AI-Assisted Depression Assessment in Special Education

arXiv:2505.23631v3 Announce Type: replace-cross Abstract: Assessing student depression in sensitive environments like special education is challenging. Standardized questionnaires may not fully reflect students' true situations. Furthermore, automated methods often falter with rich student narratives, lacking the crucial, individualized insights stemming from teachers' empathetic connections with students. Existing methods often fail to address this ambiguity or effectively integrate educator understanding. To address these limitations by fostering a synergistic human-AI collaboration, this paper introduces Human Empathy as Encoder (HEAE), a novel, human-centered AI framework for transparent and socially responsible depression severity assessment. Our approach uniquely integrates student narrative text with a teacher-derived, 9-dimensional "Empathy Vector" (EV), its dimensions guided by the PHQ-9 framework,to explicitly translate tacit empathetic insight into a structured AI input enhancing rather than replacing human judgment. Rigorous experiments optimized the multimodal fusion, text representation, and classification architecture, achieving 82.74% accuracy for 7-level severity classification. This work demonstrates a path toward more responsible and ethical affective computing by structurally embedding human empathy

Rethinking Exact Unlearning under Exposure: Extracting Forgotten Data under Exact Unlearning in Large Language Model

arXiv:2505.24379v2 Announce Type: replace-cross Abstract: Large Language Models are typically trained on datasets collected from the web, which may inadvertently contain harmful or sensitive personal information. To address growing privacy concerns, unlearning methods have been proposed to remove the influence of specific data from trained models. Of these, exact unlearning -- which retrains the model from scratch without the target data -- is widely regarded the gold standard for mitigating privacy risks in deployment. In this paper, we revisit this assumption in a practical deployment setting where both the pre- and post-unlearning logits API are exposed, such as in open-weight scenarios. Targeting this setting, we introduce a novel data extraction attack that leverages signals from the pre-unlearning model to guide the post-unlearning model, uncovering patterns that reflect the removed data distribution. Combining model guidance with a token filtering strategy, our attack significantly improves extraction success rates -- doubling performance in some cases -- across common benchmarks such as MUSE, TOFU, and WMDP. Furthermore, we demonstrate our attack's effectiveness on a simulated medical diagnosis dataset to highlight real-world privacy risks associated with exact unlearning. In light of our findings, which suggest that unlearning may, in a contradictory way, increase the risk of privacy leakage during real-world deployments, we advocate for evaluation of unlearning methods to consider broader threat models that account not only for post-unlearning models but also for adversarial access to prior checkpoints. Code is publicly available at: https://github.com/Nicholas0228/unlearned_data_extraction_llm.

Contextual Integrity in LLMs via Reasoning and Reinforcement Learning

arXiv:2506.04245v2 Announce Type: replace-cross Abstract: As the era of autonomous agents making decisions on behalf of users unfolds, ensuring contextual integrity (CI) -- what is the appropriate information to share while carrying out a certain task -- becomes a central question to the field. We posit that CI demands a form of reasoning where the agent needs to reason about the context in which it is operating. To test this, we first prompt LLMs to reason explicitly about CI when deciding what information to disclose. We then extend this approach by developing a reinforcement learning (RL) framework that further instills in models the reasoning necessary to achieve CI. Using a synthetic, automatically created, dataset of only $\sim700$ examples but with diverse contexts and information disclosure norms, we show that our method substantially reduces inappropriate information disclosure while maintaining task performance across multiple model sizes and families. Importantly, improvements transfer from this synthetic dataset to established CI benchmarks such as PrivacyLens that has human annotations and evaluates privacy leakage of AI assistants in actions and tool calls.

Beyond Chunking: Discourse-Aware Hierarchical Retrieval for Long Document Question Answering

arXiv:2506.06313v3 Announce Type: replace-cross Abstract: Long document question answering systems typically process texts as flat sequences or use arbitrary segmentation, failing to capture discourse structures that guide human comprehension. We present a discourse-aware hierarchical framework that leverages rhetorical structure theory (RST) to enhance long document question answering. Our approach converts discourse trees into sentence-level representations and employs LLM-enhanced node representations to bridge structural and semantic information. The framework involves three key innovations: specialized discourse parsing for lengthy documents, LLM-based enhancement of discourse relation nodes, and structure-guided hierarchical retrieval. Comprehensive experiments on QASPER, QuALITY, and NarrativeQA demonstrate consistent improvements over existing approaches. Ablation studies confirm that incorporating discourse structure significantly enhances question answering across diverse document types.

Chasing Moving Targets with Online Self-Play Reinforcement Learning for Safer Language Models

arXiv:2506.07468v3 Announce Type: replace-cross Abstract: Conventional language model (LM) safety alignment relies on a reactive, disjoint procedure: attackers exploit a static model, followed by defensive fine-tuning to patch exposed vulnerabilities. This sequential approach creates a mismatch -- attackers overfit to obsolete defenses, while defenders perpetually lag behind emerging threats. To address this, we propose Self-RedTeam, an online self-play reinforcement learning algorithm where an attacker and defender agent co-evolve through continuous interaction. We cast safety alignment as a two-player zero-sum game, where a single model alternates between attacker and defender roles -- generating adversarial prompts and safeguarding against them -- while a reward LM adjudicates outcomes. This enables dynamic co-adaptation. Grounded in the game-theoretic framework of zero-sum games, we establish a theoretical safety guarantee which motivates the design of our method: if self-play converges to a Nash Equilibrium, the defender will reliably produce safe responses to any adversarial input. Empirically, Self-RedTeam uncovers more diverse attacks (+21.8% SBERT) compared to attackers trained against static defenders and achieves higher robustness on safety benchmarks (e.g., +65.5% on WildJailBreak) than defenders trained against static attackers. We further propose hidden Chain-of-Thought, allowing agents to plan privately, which boosts adversarial diversity and reduces over-refusals. Our results motivate a shift from reactive patching to proactive co-evolution in LM safety training, enabling scalable, autonomous, and robust self-improvement of LMs via multi-agent reinforcement learning (MARL).

Can We Infer Confidential Properties of Training Data from LLMs?

arXiv:2506.10364v3 Announce Type: replace-cross Abstract: Large language models (LLMs) are increasingly fine-tuned on domain-specific datasets to support applications in fields such as healthcare, finance, and law. These fine-tuning datasets often have sensitive and confidential dataset-level properties -- such as patient demographics or disease prevalence -- that are not intended to be revealed. While prior work has studied property inference attacks on discriminative models (e.g., image classification models) and generative models (e.g., GANs for image data), it remains unclear if such attacks transfer to LLMs. In this work, we introduce PropInfer, a benchmark task for evaluating property inference in LLMs under two fine-tuning paradigms: question-answering and chat-completion. Built on the ChatDoctor dataset, our benchmark includes a range of property types and task configurations. We further propose two tailored attacks: a prompt-based generation attack and a shadow-model attack leveraging word frequency signals. Empirical evaluations across multiple pretrained LLMs show the success of our attacks, revealing a previously unrecognized vulnerability in LLMs.

Trainable Dynamic Mask Sparse Attention

arXiv:2508.02124v4 Announce Type: replace-cross Abstract: In large language models, the demand for modeling long contexts is ever-increasing, yet the quadratic complexity of standard self-attention presents a significant bottleneck. While existing sparse attention mechanisms enhance efficiency, they often suffer from limitations such as static patterns and information loss. This paper introduces a Trainable Dynamic Mask Sparse Attention mechanism that addresses these challenges through three key innovations. First, it leverages value vectors to dynamically generate content-aware sparse masks, enabling the model to adaptively identify and focus on crucial information. Second, it implements a position-aware sparse attention computation that effectively skips unnecessary computational regions. Finally, we ensure that the introduced dynamic masks and sparse weights do not obstruct gradients, thereby supporting end-to-end training. This dual-sparsity design allows the model to retain complete information while significantly reducing computational complexity, achieving an excellent balance between efficiency and performance. We validate the performance of Dynamic Mask Attention through comprehensive experiments. Comparative studies demonstrate that our method consistently achieves Pareto dominance across various tasks, including scaling laws, multi-query associative recall, general benchmarks, and needle-in-a-haystack tests, delivering up to 10 times acceleration. These results highlight its capability to effectively balance model efficiency with long-context modeling. Our computational kernel is open-sourced at https://github.com/SmallDoges/flash-dmattn to facilitate further research and application within the community.

Non-Interactive Symbolic-Aided Chain-of-Thought for Logical Reasoning

arXiv:2508.12425v2 Announce Type: replace-cross Abstract: This work introduces Symbolic-Aided Chain-of-Thought (CoT), an improved approach to standard CoT, for logical reasoning in large language models (LLMs). The key idea is to integrate lightweight symbolic representations into few-shot prompts, structuring the inference steps with a consistent strategy to make reasoning patterns more explicit within a non-interactive reasoning process. By incorporating these symbolic structures, Symbolic-Aided CoT preserves the generalizability of standard prompting techniques while enhancing the transparency, interpretability, and analyzability of LLM logical reasoning. Extensive experiments on four well-known logical reasoning benchmarks -- ProofWriter, FOLIO, ProntoQA, and LogicalDeduction, which cover diverse reasoning tasks and scenarios -- demonstrate the effectiveness of the proposed approach, particularly in complex reasoning tasks that require navigating multiple constraints or rules. Notably, Symbolic-Aided CoT consistently improves LLMs' reasoning capabilities across various model sizes and significantly outperforms conventional CoT on three out of four datasets, ProofWriter, ProntoQA, and LogicalDeduction.

The Telephone Game: Evaluating Semantic Drift in Unified Models

arXiv:2509.04438v2 Announce Type: replace-cross Abstract: Employing a single, unified model (UM) for both visual understanding (image-to-text: I2T) and visual generation (text-to-image: T2I) has opened a new direction in Visual Language Model (VLM) research. While UMs can also support broader unimodal tasks (e.g., text-to-text, image-to-image), we focus on the core cross-modal pair T2I and I2T. Existing evaluation benchmarks consider these capabilities in isolation: FID and GenEval for T2I, and benchmarks such as MME, MMBench for I2T. These isolated single-pass metrics do not reveal cross-consistency: whether a model that "understands" a concept can also "render" it, nor whether semantic meaning is preserved when cycling between image and text modalities. To address this, we introduce the Semantic Drift Protocol (SDP) for Unified Models, a cyclic evaluation protocol that alternates I2T and T2I over multiple generations to quantify semantic drift. We propose two metrics: (i) Mean Cumulative Drift (MCD), an embedding-based measure of overall semantic drift; and (ii) Multi-Generation GenEval (MGG), an object-level compliance score extending GenEval. To assess generalization beyond COCO dataset, which is widely used in training; we create a new benchmark Nocaps+Docci400, sampled from NoCaps and DOCCI and evaluated on seven recent models. SDP reveals substantial variation in cross-modal stability: some models like BAGEL maintain semantic meaning over many alternations, whereas others like VILA-U drift quickly despite strong single-pass scores. Our results highlight SDP as a necessary complement to standard I2T and T2I evaluations. Code is available at https://github.com/mollahsabbir/Semantic-Drift-in-Unified-Models

ArcMemo: Abstract Reasoning Composition with Lifelong LLM Memory

arXiv:2509.04439v3 Announce Type: replace-cross Abstract: While inference-time scaling enables LLMs to carry out increasingly long and capable reasoning traces, the patterns and insights uncovered during these traces are immediately discarded once the context window is reset for a new query. External memory is a natural way to persist these discoveries, and recent work has shown clear benefits for reasoning-intensive tasks. We see an opportunity to make such memories more broadly reusable and scalable by moving beyond instance-based memory entries (e.g. exact query/response pairs, or summaries tightly coupled with the original problem context) toward concept-level memory: reusable, modular abstractions distilled from solution traces and stored in natural language. For future queries, relevant concepts are selectively retrieved and integrated into the prompt, enabling test-time continual learning without weight updates. Our design introduces new strategies for abstracting takeaways from rollouts and retrieving entries for new queries, promoting reuse and allowing memory to expand with additional experiences. We evaluate on ARC-AGI, a benchmark that stresses compositional generalization and abstract reasoning, making it a natural fit for concept memory. Our method yields a 7.5% relative gain over a strong no-memory baseline with performance continuing to scale with inference compute. We find abstract concepts to be the most consistent memory design, outscoring the baseline at all tested inference compute scales. Moreover, dynamically updating memory during test-time outperforms fixed settings, supporting the hypothesis that accumulating and abstracting patterns enables further solutions in a form of self-improvement. Code is available at https://github.com/matt-seb-ho/arc_memo.

FedMentor: Domain-Aware Differential Privacy for Heterogeneous Federated LLMs in Mental Health

arXiv:2509.14275v2 Announce Type: replace-cross Abstract: Privacy-preserving adaptation of Large Language Models (LLMs) in sensitive domains (e.g., mental health) requires balancing strict confidentiality with model utility and safety. We propose FedMentor, a federated fine-tuning framework that integrates Low-Rank Adaptation (LoRA) and domain-aware Differential Privacy (DP) to meet per-domain privacy budgets while maintaining performance. Each client (domain) applies a custom DP noise scale proportional to its data sensitivity, and the server adaptively reduces noise when utility falls below a threshold. In experiments on three mental health datasets, we show that FedMentor improves safety over standard Federated Learning (FL) without privacy, raising safe output rates by up to three points and lowering toxicity, while maintaining utility (BERTScore F1 and ROUGE-L) within 0.5% of the non-private baseline and close to the centralized upper bound. The framework scales to backbones with up to 1.7B parameters on single-GPU clients, requiring < 173 MB of communication per-round. FedMentor demonstrates a practical approach to privately fine-tune LLMs for safer deployments in healthcare and other sensitive fields.

CE-GPPO: Coordinating Entropy via Gradient-Preserving Clipping Policy Optimization in Reinforcement Learning

arXiv:2509.20712v3 Announce Type: replace-cross Abstract: Reinforcement learning (RL) has become a powerful paradigm for optimizing large language models (LLMs) to handle complex reasoning tasks. A core challenge in this process lies in managing policy entropy, which reflects the balance between exploration and exploitation during training. Existing methods, such as proximal policy optimization (PPO) and its variants, discard valuable gradient signals from low-probability tokens due to the clipping mechanism. We systematically analyze the entropy dynamics and reveal that these clipped tokens play a critical yet overlooked role in regulating entropy evolution. We propose \textbf{C}oordinating \textbf{E}ntropy via \textbf{G}radient-\textbf{P}reserving \textbf{P}olicy \textbf{O}ptimization (CE-GPPO), a novel algorithm that reintroduces gradients from clipped tokens in native PPO in a gentle and bounded manner. By controlling the magnitude of gradients from tokens outside the clipping interval, CE-GPPO is able to achieve an exploration-exploitation trade-off. We provide theoretical justification and empirical evidence showing that CE-GPPO effectively mitigates entropy instability. Extensive experiments on mathematical reasoning benchmarks show that CE-GPPO consistently outperforms strong baselines across different model scales.

RL Grokking Recipe: How Does RL Unlock and Transfer New Algorithms in LLMs?

arXiv:2509.21016v2 Announce Type: replace-cross Abstract: It remains an open question whether LLMs can acquire or generalize genuinely new reasoning strategies, beyond the sharpened skills encoded in their parameters during pre-training or post-training. To attempt to answer this debate, we introduce DELTA-Code -- Distributional Evaluation of Learnability and Transferrability in Algorithmic Coding -- a controlled benchmark of synthetic coding problem families designed to probe two fundamental aspects: learnability -- can LLMs, through reinforcement learning (RL), solve problem families where pretrained models exhibit failure with large enough attempts (pass@K=0)? -- and transferrability -- if learnability happens, can such skills transfer systematically to out-of-distribution (OOD) test sets? Unlike prior public coding datasets, DELTA isolates reasoning skills through templated problem generators and introduces fully OOD problem families that demand novel strategies rather than tool invocation or memorized patterns. Our experiments reveal a striking grokking phase transition: after an extended period with near-zero reward, RL-trained models abruptly climb to near-perfect accuracy. To enable learnability on previously unsolvable problem families, we explore key training ingredients such as staged warm-up with dense rewards, experience replay, curriculum training, and verification-in-the-loop. Beyond learnability, we use DELTA to evaluate transferability or generalization along exploratory, compositional, and transformative axes, as well as cross-family transfer. Results show solid gains within families and for recomposed skills, but persistent weaknesses in transformative cases. DELTA thus offers a clean testbed for probing the limits of RL-driven reasoning and for understanding how models can move beyond existing priors to acquire new algorithmic skills.

Your Models Have Thought Enough: Training Large Reasoning Models to Stop Overthinking

arXiv:2509.23392v2 Announce Type: replace-cross Abstract: Large Reasoning Models (LRMs) have achieved impressive performance on challenging tasks, yet their deep reasoning often incurs substantial computational costs. To achieve efficient reasoning, existing reinforcement learning methods still struggle to construct short reasoning path during the rollout stage, limiting effective learning. Inspired by Evidence Accumulation Models, we find that LRMs have accumulated sufficient information early in reasoning, making further reasoning steps redundant. Based on this insight, we propose Just-Enough Thinking (JET), which trains models to proactively terminate unnecessary reasoning. JET performs trajectory truncation during rollout to expose the model to short, distributionally consistent reasoning paths. Besides, it uses a quality-controlled length reward to better encourage concise reasoning while maintaining correctness. Extensive experiments demonstrate that JET significantly improves reasoning efficiency without sacrificing accuracy. Especially, DeepSeek-Distill-Qwen-1.5B achieves a 4.6% accuracy gain while reducing output length by 46.3% on the Olympiad benchmark. Our code is available in the GitHub.

Latent Visual Reasoning

arXiv:2509.24251v2 Announce Type: replace-cross Abstract: Multimodal Large Language Models (MLLMs) have achieved notable gains in various tasks by incorporating Chain-of-Thought (CoT) reasoning in language spaces. Recent work extends this direction by leveraging external tools for visual editing, thereby enhancing the visual signal along the reasoning trajectories. Nevertheless, these approaches remain fundamentally constrained: reasoning is still confined to the language space, with visual information treated as static preconditions. We introduce Latent Visual Reasoning (LVR), a new paradigm that enables autoregressive reasoning directly in the visual embedding space. A visual encoder first projects images into visual tokens within a joint semantic space shared with the language model. The language model is then trained to generate latent states that reconstruct key visual tokens critical for answering the query, constituting the process of latent visual reasoning. By interleaving LVR with standard text generation, our model achieves substantial gains on perception-intensive visual question answering tasks. In addition, we adapt the GRPO algorithm to conduct reinforcement learning on latent reasoning, further balancing LVR and textual generation. We show that LVR substantially improves fine-grained visual understanding and perception, achieving 71.67% on MMVP compared to 66.67% with Qwen2.5-VL. Code base and model weights will be released later.

Extracting the Structure of Press Releases for Predicting Earnings Announcement Returns

arXiv:2509.24254v2 Announce Type: replace-cross Abstract: We examine how textual features in earnings press releases predict stock returns on earnings announcement days. Using over 138,000 press releases from 2005 to 2023, we compare traditional bag-of-words and BERT-based embeddings. We find that press release content (soft information) is as informative as earnings surprise (hard information), with FinBERT yielding the highest predictive power. Combining models enhances explanatory strength and interpretability of the content of press releases. Stock prices fully reflect the content of press releases at market open. If press releases are leaked, it offers predictive advantage. Topic analysis reveals self-serving bias in managerial narratives. Our framework supports real-time return prediction through the integration of online learning, provides interpretability and reveals the nuanced role of language in price formation.

On the Self-awareness of Large Reasoning Models' Capability Boundaries

arXiv:2509.24711v2 Announce Type: replace-cross Abstract: Large Reasoning Models (LRMs) have shown impressive performance on complex reasoning tasks such as mathematics, yet they also display misbehaviors that expose their limitations. In particular, when faced with hard questions, LRMs often engage in unproductive reasoning until context limit, producing wrong answers while wasting substantial computation. This phenomenon reflects a fundamental issue: current answering paradigms overlook the relationship between questions and LRMs' capability boundaries. In this paper, we investigate whether LRMs possess self-awareness of capability boundaries. We begin by an observation that LRMs may know what they cannot solve through expressed reasoning confidence. For black-box models, we find that reasoning expressions reveal boundary signals, with accelerated growing confidence trajectory for solvable problems but convergent uncertainty trajectory for unsolvable ones. For white-box models, we show that hidden states of the last input token encode boundary information, with solvable and unsolvable problems linearly separable even before reasoning begins. Building on these findings, we propose two simple yet effective optimization strategies: reasoning expression monitoring and hidden states monitoring. Experiments demonstrate that these boundary-aware strategies enable LRMs to avoid unproductive reasoning without sacrificing accuracy, significantly improving reliability and efficiency by cutting token usage up to 62.7 - 93.6%.

Pushing LLMs to Their Logical Reasoning Bound: The Role of Data Reasoning Intensity

arXiv:2509.24836v3 Announce Type: replace-cross Abstract: Recent advances in large language models (LLMs) highlight the importance of training data structure and quality in shaping reasoning behavior. However, most existing approaches focus on transforming data formats while neglecting the internal reasoning complexity of training samples, leaving the reasoning potential of data under-explored and underutilized. In this work, we posit that LLM logical reasoning performance is jointly constrained by the potential of the training data and the cognitive capacity of the model. To make this relationship measurable, we introduce Data Reasoning Intensity (DRI), a novel metric that quantifies the latent logical reasoning complexity of samples by decomposing and aggregating their logical structures. This allows us to analyze how well current LLMs utilize logical reasoning signals and identify performance gaps relative to data potential. Based on this insight, we introduce a re-cognizing optimization strategy that systematically enhances the logical reasoning intensity of training data. Rather than increasing data volume, our method re-optimizes existing samples to better align with the LLM's logical reasoning boundary. Extensive experiments show that our approach significantly improves performance and generalization over data-centric strategies. We further validate our method under a reinforcement learning framework. Our results indicate that prioritizing reasoning complexity in data rather than sheer scale or superficial form is essential to realizing LLMs' full cognitive potential.

Learning to Reason as Action Abstractions with Scalable Mid-Training RL

arXiv:2509.25810v2 Announce Type: replace-cross Abstract: Large language models excel with reinforcement learning (RL), but fully unlocking this potential requires a mid-training stage. An effective mid-training phase should identify a compact set of useful actions and enable fast selection among them through online RL. We formalize this intuition by presenting the first theoretical result on how mid-training shapes post-training: it characterizes an action subspace that minimizes both the value approximation error from pruning and the RL error during subsequent planning. Our analysis reveals two key determinants of mid-training effectiveness: pruning efficiency, which shapes the prior of the initial RL policy, and its impact on RL convergence, which governs the extent to which that policy can be improved via online interactions. These results suggest that mid-training is most effective when the decision space is compact and the effective horizon is short, highlighting the importance of operating in the space of action abstractions rather than primitive actions. Building on these insights, we propose Reasoning as Action Abstractions (RA3), a scalable mid-training algorithm. Specifically, we derive a sequential variational lower bound and optimize it by iteratively discovering temporally-consistent latent structures via RL, followed by fine-tuning on the bootstrapped data. Experiments on code generation tasks demonstrate the effectiveness of our approach. Across multiple base models, RA3 improves the average performance on HumanEval and MBPP by 8 and 4 points over the base model and the next-token prediction baseline. Furthermore, RA3 achieves faster convergence and higher asymptotic performance in RLVR on HumanEval+, MBPP+, LiveCodeBench, and Codeforces.

Auto-ARGUE: LLM-Based Report Generation Evaluation

arXiv:2509.26184v3 Announce Type: replace-cross Abstract: Generation of long-form, citation-backed reports is a primary use case for retrieval augmented generation (RAG) systems. While open-source evaluation tools exist for various RAG tasks, ones tailored to report generation are lacking. Accordingly, we introduce Auto-ARGUE, a robust LLM-based implementation of the recent ARGUE framework for report generation evaluation. We present analysis of Auto-ARGUE on the report generation pilot task from the TREC 2024 NeuCLIR track, showing good system-level correlations with human judgments. We further release a web app for visualization of Auto-ARGUE outputs.

Is It Thinking or Cheating? Detecting Implicit Reward Hacking by Measuring Reasoning Effort

arXiv:2510.01367v2 Announce Type: replace-cross Abstract: Reward hacking, where a reasoning model exploits loopholes in a reward function to achieve high rewards without solving the intended task, poses a significant threat. This behavior may be explicit, i.e. verbalized in the model's chain-of-thought (CoT), or implicit, where the CoT appears benign thus bypasses CoT monitors. To detect implicit reward hacking, we propose TRACE (Truncated Reasoning AUC Evaluation). Our key observation is that hacking occurs when exploiting the loophole is easier than solving the actual task. This means that the model is using less 'effort' than required to achieve high reward. TRACE quantifies effort by measuring how early a model's reasoning becomes sufficient to obtain the reward. We progressively truncate a model's CoT at various lengths, force the model to answer, and estimate the expected reward at each cutoff. A hacking model, which takes a shortcut, will achieve a high expected reward with only a small fraction of its CoT, yielding a large area under the accuracy-vs-length curve. TRACE achieves over 65% gains over our strongest 72B CoT monitor in math reasoning, and over 30% gains over a 32B monitor in coding. We further show that TRACE can discover unknown loopholes during training. Overall, TRACE offers a scalable unsupervised approach for oversight where current monitoring methods prove ineffective.

Less LLM, More Documents: Searching for Improved RAG

arXiv:2510.02657v2 Announce Type: replace-cross Abstract: Retrieval-Augmented Generation (RAG) couples document retrieval with large language models (LLMs). While scaling generators improves accuracy, it also raises cost and limits deployability. We explore an orthogonal axis: enlarging the retriever's corpus to reduce reliance on large LLMs. Experimental results show that corpus scaling consistently strengthens RAG and can often serve as a substitute for increasing model size, though with diminishing returns at larger scales. Small- and mid-sized generators paired with larger corpora often rival much larger models with smaller corpora; mid-sized models tend to gain the most, while tiny and large models benefit less. Our analysis shows that improvements arise primarily from increased coverage of answer-bearing passages, while utilization efficiency remains largely unchanged. These findings establish a principled corpus-generator trade-off: investing in larger corpora offers an effective path to stronger RAG, often comparable to enlarging the LLM itself.

SoC-DT: Standard-of-Care Aligned Digital Twins for Patient-Specific Tumor Dynamics

arXiv:2510.03287v1 Announce Type: new Abstract: Accurate prediction of tumor trajectories under standard-of-care (SoC) therapies remains a major unmet need in oncology. This capability is essential for optimizing treatment planning and anticipating disease progression. Conventional reaction-diffusion models are limited in scope, as they fail to capture tumor dynamics under heterogeneous therapeutic paradigms. There is hence a critical need for computational frameworks that can realistically simulate SoC interventions while accounting for inter-patient variability in genomics, demographics, and treatment regimens. We introduce Standard-of-Care Digital Twin (SoC-DT), a differentiable framework that unifies reaction-diffusion tumor growth models, discrete SoC interventions (surgery, chemotherapy, radiotherapy) along with genomic and demographic personalization to predict post-treatment tumor structure on imaging. An implicit-explicit exponential time-differencing solver, IMEX-SoC, is also proposed, which ensures stability, positivity, and scalability in SoC treatment situations. Evaluated on both synthetic data and real world glioma data, SoC-DT consistently outperforms classical PDE baselines and purely data-driven neural models in predicting tumor dynamics. By bridging mechanistic interpretability with modern differentiable solvers, SoC-DT establishes a principled foundation for patient-specific digital twins in oncology, enabling biologically consistent tumor dynamics estimation. Code will be made available upon acceptance.

Visualizing Celebrity Dynamics in Video Content: A Proposed Approach Using Face Recognition Timestamp Data

arXiv:2510.03292v1 Announce Type: new Abstract: In an era dominated by video content, understanding its structure and dynamics has become increasingly important. This paper presents a hybrid framework that combines a distributed multi-GPU inference system with an interactive visualization platform for analyzing celebrity dynamics in video episodes. The inference framework efficiently processes large volumes of video data by leveraging optimized ONNX models, heterogeneous batch inference, and high-throughput parallelism, ensuring scalable generation of timestamped appearance records. These records are then transformed into a comprehensive suite of visualizations, including appearance frequency charts, duration analyses, pie charts, co-appearance matrices, network graphs, stacked area charts, seasonal comparisons, and heatmaps. Together, these visualizations provide multi-dimensional insights into video content, revealing patterns in celebrity prominence, screen-time distribution, temporal dynamics, co-appearance relationships, and intensity across episodes and seasons. The interactive nature of the system allows users to dynamically explore data, identify key moments, and uncover evolving relationships between individuals. By bridging distributed recognition with structured, visually-driven analytics, this work enables new possibilities for entertainment analytics, content creation strategies, and audience engagement studies.

Domain-Robust Marine Plastic Detection Using Vision Models

arXiv:2510.03294v1 Announce Type: new Abstract: Marine plastic pollution is a pressing environmental threat, making reliable automation for underwater debris detection essential. However, vision systems trained on one dataset often degrade on new imagery due to domain shift. This study benchmarks models for cross-domain robustness, training convolutional neural networks - CNNs (MobileNetV2, ResNet-18, EfficientNet-B0) and vision transformers (DeiT-Tiny, ViT-B16) on a labeled underwater dataset and then evaluates them on a balanced cross-domain test set built from plastic-positive images drawn from a different source and negatives from the training domain. Two zero-shot models were assessed, CLIP ViT-L14 and Google's Gemini 2.0 Flash, that leverage pretraining to classify images without fine-tuning. Results show the lightweight MobileNetV2 delivers the strongest cross-domain performance (F1 0.97), surpassing larger models. All fine-tuned models achieved high Precision (around 99%), but differ in Recall, indicating varying sensitivity to plastic instances. Zero-shot CLIP is comparatively sensitive (Recall around 80%) yet prone to false positives (Precision around 56%), whereas Gemini exhibits the inverse profile (Precision around 99%, Recall around 81%). Error analysis highlights recurring confusions with coral textures, suspended particulates, and specular glare. Overall, compact CNNs with supervised training can generalize effectively for cross-domain underwater detection, while large pretrained vision-language models provide complementary strengths.

Multimodal Arabic Captioning with Interpretable Visual Concept Integration

arXiv:2510.03295v1 Announce Type: new Abstract: We present VLCAP, an Arabic image captioning framework that integrates CLIP-based visual label retrieval with multimodal text generation. Rather than relying solely on end-to-end captioning, VLCAP grounds generation in interpretable Arabic visual concepts extracted with three multilingual encoders, mCLIP, AraCLIP, and Jina V4, each evaluated separately for label retrieval. A hybrid vocabulary is built from training captions and enriched with about 21K general domain labels translated from the Visual Genome dataset, covering objects, attributes, and scenes. The top-k retrieved labels are transformed into fluent Arabic prompts and passed along with the original image to vision-language models. In the second stage, we tested Qwen-VL and Gemini Pro Vision for caption generation, resulting in six encoder-decoder configurations. The results show that mCLIP + Gemini Pro Vision achieved the best BLEU-1 (5.34%) and cosine similarity (60.01%), while AraCLIP + Qwen-VL obtained the highest LLM-judge score (36.33%). This interpretable pipeline enables culturally coherent and contextually accurate Arabic captions.

Convolutional Neural Nets vs Vision Transformers: A SpaceNet Case Study with Balanced vs Imbalanced Regimes

arXiv:2510.03297v1 Announce Type: new Abstract: We present a controlled comparison of a convolutional neural network (EfficientNet-B0) and a Vision Transformer (ViT-Base) on SpaceNet under two label-distribution regimes: a naturally imbalanced five-class split and a balanced-resampled split with 700 images per class (70:20:10 train/val/test). With matched preprocessing (224x224, ImageNet normalization), lightweight augmentations, and a 40-epoch budget on a single NVIDIA P100, we report accuracy, macro-F1, balanced accuracy, per-class recall, and deployment metrics (model size and latency). On the imbalanced split, EfficientNet-B0 reaches 93% test accuracy with strong macro-F1 and lower latency; ViT-Base is competitive at 93% with a larger parameter count and runtime. On the balanced split, both models are strong; EfficientNet-B0 reaches 99% while ViT-Base remains competitive, indicating that balancing narrows architecture gaps while CNNs retain an efficiency edge. We release manifests, logs, and per-image predictions to support reproducibility.

A Comprehensive Review on Artificial Intelligence Empowered Solutions for Enhancing Pedestrian and Cyclist Safety

arXiv:2510.03314v1 Announce Type: new Abstract: Ensuring the safety of vulnerable road users (VRUs), such as pedestrians and cyclists, remains a critical global challenge, as conventional infrastructure-based measures often prove inadequate in dynamic urban environments. Recent advances in artificial intelligence (AI), particularly in visual perception and reasoning, open new opportunities for proactive and context-aware VRU protection. However, existing surveys on AI applications for VRUs predominantly focus on detection, offering limited coverage of other vision-based tasks that are essential for comprehensive VRU understanding and protection. This paper presents a state-of-the-art review of recent progress in camera-based AI sensing systems for VRU safety, with an emphasis on developments from the past five years and emerging research trends. We systematically examine four core tasks, namely detection and classification, tracking and reidentification, trajectory prediction, and intent recognition and prediction, which together form the backbone of AI-empowered proactive solutions for VRU protection in intelligent transportation systems. To guide future research, we highlight four major open challenges from the perspectives of data, model, and deployment. By linking advances in visual AI with practical considerations for real-world implementation, this survey aims to provide a foundational reference for the development of next-generation sensing systems to enhance VRU safety.

The View From Space: Navigating Instrumentation Differences with EOFMs

arXiv:2510.03316v1 Announce Type: new Abstract: Earth Observation Foundation Models (EOFMs) have exploded in prevalence as tools for processing the massive volumes of remotely sensed and other earth observation data, and for delivering impact on the many essential earth monitoring tasks. An emerging trend posits using the outputs of pre-trained models as 'embeddings' which summarize high dimensional data to be used for generic tasks such as similarity search and content-specific queries. However, most EOFM models are trained only on single modalities of data and then applied or benchmarked by matching bands across different modalities. It is not clear from existing work what impact diverse sensor architectures have on the internal representations of the present suite of EOFMs. We show in this work that the representation space of EOFMs is highly sensitive to sensor architecture and that understanding this difference gives a vital perspective on the pitfalls of current EOFM design and signals for how to move forward as model developers, users, and a community guided by robust remote-sensing science.

Photorealistic Inpainting for Perturbation-based Explanations in Ecological Monitoring

arXiv:2510.03317v1 Announce Type: new Abstract: Ecological monitoring is increasingly automated by vision models, yet opaque predictions limit trust and field adoption. We present an inpainting-guided, perturbation-based explanation technique that produces photorealistic, mask-localized edits that preserve scene context. Unlike masking or blurring, these edits stay in-distribution and reveal which fine-grained morphological cues drive predictions in tasks such as species recognition and trait attribution. We demonstrate the approach on a YOLOv9 detector fine-tuned for harbor seal detection in Glacier Bay drone imagery, using Segment-Anything-Model-refined masks to support two interventions: (i) object removal/replacement (e.g., replacing seals with plausible ice/water or boats) and (ii) background replacement with original animals composited onto new scenes. Explanations are assessed by re-scoring perturbed images (flip rate, confidence drop) and by expert review for ecological plausibility and interpretability. The resulting explanations localize diagnostic structures, avoid deletion artifacts common to traditional perturbations, and yield domain-relevant insights that support expert validation and more trustworthy deployment of AI in ecology.

Advances in Medical Image Segmentation: A Comprehensive Survey with a Focus on Lumbar Spine Applications

arXiv:2510.03318v1 Announce Type: new Abstract: Medical Image Segmentation (MIS) stands as a cornerstone in medical image analysis, playing a pivotal role in precise diagnostics, treatment planning, and monitoring of various medical conditions. This paper presents a comprehensive and systematic survey of MIS methodologies, bridging the gap between traditional image processing techniques and modern deep learning approaches. The survey encompasses thresholding, edge detection, region-based segmentation, clustering algorithms, and model-based techniques while also delving into state-of-the-art deep learning architectures such as Convolutional Neural Networks (CNNs), Fully Convolutional Networks (FCNs), and the widely adopted U-Net and its variants. Moreover, integrating attention mechanisms, semi-supervised learning, generative adversarial networks (GANs), and Transformer-based models is thoroughly explored. In addition to covering established methods, this survey highlights emerging trends, including hybrid architectures, cross-modality learning, federated and distributed learning frameworks, and active learning strategies, which aim to address challenges such as limited labeled datasets, computational complexity, and model generalizability across diverse imaging modalities. Furthermore, a specialized case study on lumbar spine segmentation is presented, offering insights into the challenges and advancements in this relatively underexplored anatomical region. Despite significant progress in the field, critical challenges persist, including dataset bias, domain adaptation, interpretability of deep learning models, and integration into real-world clinical workflows.

DECOR: Deep Embedding Clustering with Orientation Robustness

arXiv:2510.03328v1 Announce Type: new Abstract: In semiconductor manufacturing, early detection of wafer defects is critical for product yield optimization. However, raw wafer data from wafer quality tests are often complex, unlabeled, imbalanced and can contain multiple defects on a single wafer, making it crucial to design clustering methods that remain reliable under such imperfect data conditions. We introduce DECOR, a deep clustering with orientation robustness framework that groups complex defect patterns from wafer maps into consistent clusters. We evaluate our method on the open source MixedWM38 dataset, demonstrating its ability to discover clusters without manual tuning. DECOR explicitly accounts for orientation variations in wafer maps, ensuring that spatially similar defects are consistently clustered regardless of its rotation or alignment. Experiments indicate that our method outperforms existing clustering baseline methods, thus providing a reliable and scalable solution in automated visual inspection systems.

Error correction in multiclass image classification of facial emotion on unbalanced samples

arXiv:2510.03337v1 Announce Type: new Abstract: This paper considers the problem of error correction in multi-class classification of face images on unbalanced samples. The study is based on the analysis of a data frame containing images labeled by seven different emotional states of people of different ages. Particular attention is paid to the problem of class imbalance, in which some emotions significantly prevail over others. To solve the classification problem, a neural network model based on LSTM with an attention mechanism focusing on key areas of the face that are informative for emotion recognition is used. As part of the experiments, the model is trained on all possible configurations of subsets of six classes with subsequent error correction for the seventh class, excluded at the training stage. The results show that correction is possible for all classes, although the degree of success varies: some classes are better restored, others are worse. In addition, on the test sample, when correcting some classes, an increase in key quality metrics for small classes was recorded, which indicates the promise of the proposed approach in solving applied problems related to the search for rare events, for example, in anti-fraud systems. Thus, the proposed method can be effectively applied in facial expression analysis systems and in tasks requiring stable classification under skewed class distribution.

OpusAnimation: Code-Based Dynamic Chart Generation

arXiv:2510.03341v1 Announce Type: new Abstract: Dynamic Chart Generation (DCG) involves producing code-rendered animated visualizations as charts. While recent advances in multi-modal large language models (MLLMs) have significantly improved their capability on static chart generation and comprehension, MLLMs' potential for handling dynamic chart generation and understanding remains underexplored. To bridge this research gap, we introduce DCG-Bench (Dynamic Chart Generation Benchmark), the first benchmark evaluating MLLM's capability on dynamic chart generation tasks from three dimensions: Simple Text-to-Chart, Detailed Text-to-Chart, and Video-to-Chart tasks. We construct DCG-8K, a high-quality DCG dataset with annotations covering instruction-code-video triplets and QA pairs for both code and video evaluation. Based on DCG-8K, we explored a two-stage training recipe, proposing Joint-Code-Visual Reward for group relative policy optimization to construct expert MLLM Qwen2.5-VL-DCG-3B for the DCG task. Our benchmarking result reveals shortcomings of existing MLLMs in the visual-to-chart task, and our model beats the best open-sourced MLLM with an average 8.31% performance gain across three tasks, and shows on par performance against proprietary models with only 3B parameters, proving the effectiveness of our training recipe. Our code and dataset will be publicly available.

Visual Odometry with Transformers

arXiv:2510.03348v1 Announce Type: new Abstract: Modern monocular visual odometry methods typically combine pre-trained deep learning components with optimization modules, resulting in complex pipelines that rely heavily on camera calibration and hyperparameter tuning, and often struggle in unseen real-world scenarios. Recent large-scale 3D models trained on massive amounts of multi-modal data have partially alleviated these challenges, providing generalizable dense reconstruction and camera pose estimation. Still, they remain limited in handling long videos and providing accurate per-frame estimates, which are required for visual odometry. In this work, we demonstrate that monocular visual odometry can be addressed effectively in an end-to-end manner, thereby eliminating the need for handcrafted components such as bundle adjustment, feature matching, camera calibration, or dense 3D reconstruction. We introduce VoT, short for Visual odometry Transformer, which processes sequences of monocular frames by extracting features and modeling global relationships through temporal and spatial attention. Unlike prior methods, VoT directly predicts camera motion without estimating dense geometry and relies solely on camera poses for supervision. The framework is modular and flexible, allowing seamless integration of various pre-trained encoders as feature extractors. Experimental results demonstrate that VoT scales effectively with larger datasets, benefits substantially from stronger pre-trained backbones, generalizes across diverse camera motions and calibration settings, and outperforms traditional methods while running more than 3 times faster. The code will be released.

Inference-Time Search using Side Information for Diffusion-based Image Reconstruction

arXiv:2510.03352v1 Announce Type: new Abstract: Diffusion models have emerged as powerful priors for solving inverse problems. However, existing approaches typically overlook side information that could significantly improve reconstruction quality, especially in severely ill-posed settings. In this work, we propose a novel inference-time search algorithm that guides the sampling process using the side information in a manner that balances exploration and exploitation. This enables more accurate and reliable reconstructions, providing an alternative to the gradient-based guidance that is prone to reward-hacking artifacts. Our approach can be seamlessly integrated into a wide range of existing diffusion-based image reconstruction pipelines. Through extensive experiments on a number of inverse problems, such as box inpainting, super-resolution, and various deblurring tasks including motion, Gaussian, nonlinear, and blind deblurring, we show that our approach consistently improves the qualitative and quantitative performance of diffusion-based image reconstruction algorithms. We also show the superior performance of our approach with respect to other baselines, including reward gradient-based guidance algorithms. The code is available at \href{https://github.com/mhdfb/sideinfo-search-reconstruction}{this repository}.

Sonar Image Datasets: A Comprehensive Survey of Resources, Challenges, and Applications

arXiv:2510.03353v1 Announce Type: new Abstract: Sonar images are relevant for advancing underwater exploration, autonomous navigation, and ecosystem monitoring. However, the progress depends on data availability. The scarcity of publicly available, well-annotated sonar image datasets creates a significant bottleneck for the development of robust machine learning models. This paper presents a comprehensive and concise review of the current landscape of sonar image datasets, seeking not only to catalog existing resources but also to contextualize them, identify gaps, and provide a clear roadmap, serving as a base guide for researchers of any kind who wish to start or advance in the field of underwater acoustic data analysis. We mapped publicly accessible datasets across various sonar modalities, including Side Scan Sonar (SSS), Forward-Looking Sonar (FLS), Synthetic Aperture Sonar (SAS), Multibeam Echo Sounder (MBES), and Dual-Frequency Identification Sonar (DIDSON). An analysis was conducted on applications such as classification, detection, segmentation, and 3D reconstruction. This work focuses on state-of-the-art advancements, incorporating newly released datasets. The findings are synthesized into a master table and a chronological timeline, offering a clear and accessible comparison of characteristics, sizes, and annotation details datasets.

Learned Display Radiance Fields with Lensless Cameras

arXiv:2510.03356v1 Announce Type: new Abstract: Calibrating displays is a basic and regular task that content creators must perform to maintain optimal visual experience, yet it remains a troublesome issue. Measuring display characteristics from different viewpoints often requires specialized equipment and a dark room, making it inaccessible to most users. To avoid specialized hardware requirements in display calibrations, our work co-designs a lensless camera and an Implicit Neural Representation based algorithm for capturing display characteristics from various viewpoints. More specifically, our pipeline enables efficient reconstruction of light fields emitted from a display from a viewing cone of 46.6{\deg} X 37.6{\deg}. Our emerging pipeline paves the initial steps towards effortless display calibration and characterization.

Provenance Networks: End-to-End Exemplar-Based Explainability

arXiv:2510.03361v1 Announce Type: new Abstract: We introduce provenance networks, a novel class of neural models designed to provide end-to-end, training-data-driven explainability. Unlike conventional post-hoc methods, provenance networks learn to link each prediction directly to its supporting training examples as part of the model's normal operation, embedding interpretability into the architecture itself. Conceptually, the model operates similarly to a learned KNN, where each output is justified by concrete exemplars weighted by relevance in the feature space. This approach facilitates systematic investigations of the trade-off between memorization and generalization, enables verification of whether a given input was included in the training set, aids in the detection of mislabeled or anomalous data points, enhances resilience to input perturbations, and supports the identification of similar inputs contributing to the generation of a new data point. By jointly optimizing the primary task and the explainability objective, provenance networks offer insights into model behavior that traditional deep networks cannot provide. While the model introduces additional computational cost and currently scales to moderately sized datasets, it provides a complementary approach to existing explainability techniques. In particular, it addresses critical challenges in modern deep learning, including model opaqueness, hallucination, and the assignment of credit to data contributors, thereby improving transparency, robustness, and trustworthiness in neural models.

Unified Unsupervised Anomaly Detection via Matching Cost Filtering

arXiv:2510.03363v1 Announce Type: new Abstract: Unsupervised anomaly detection (UAD) aims to identify image- and pixel-level anomalies using only normal training data, with wide applications such as industrial inspection and medical analysis, where anomalies are scarce due to privacy concerns and cold-start constraints. Existing methods, whether reconstruction-based (restoring normal counterparts) or embedding-based (pretrained representations), fundamentally conduct image- or feature-level matching to generate anomaly maps. Nonetheless, matching noise has been largely overlooked, limiting their detection ability. Beyond earlier focus on unimodal RGB-based UAD, recent advances expand to multimodal scenarios, e.g., RGB--3D and RGB--Text, enabled by point cloud sensing and vision--language models. Despite shared challenges, these lines remain largely isolated, hindering a comprehensive understanding and knowledge transfer. In this paper, we advocate unified UAD for both unimodal and multimodal settings in the matching perspective. Under this insight, we present Unified Cost Filtering (UCF), a generic post-hoc refinement framework for refining anomaly cost volume of any UAD model. The cost volume is constructed by matching a test sample against normal samples from the same or different modalities, followed by a learnable filtering module with multi-layer attention guidance from the test sample, mitigating matching noise and highlighting subtle anomalies. Comprehensive experiments on 22 diverse benchmarks demonstrate the efficacy of UCF in enhancing a variety of UAD methods, consistently achieving new state-of-the-art results in both unimodal (RGB) and multimodal (RGB--3D, RGB--Text) UAD scenarios. Code and models will be released at https://github.com/ZHE-SAPI/CostFilter-AD.

Visual Language Model as a Judge for Object Detection in Industrial Diagrams

arXiv:2510.03376v1 Announce Type: new Abstract: Industrial diagrams such as piping and instrumentation diagrams (P&IDs) are essential for the design, operation, and maintenance of industrial plants. Converting these diagrams into digital form is an important step toward building digital twins and enabling intelligent industrial automation. A central challenge in this digitalization process is accurate object detection. Although recent advances have significantly improved object detection algorithms, there remains a lack of methods to automatically evaluate the quality of their outputs. This paper addresses this gap by introducing a framework that employs Visual Language Models (VLMs) to assess object detection results and guide their refinement. The approach exploits the multimodal capabilities of VLMs to identify missing or inconsistent detections, thereby enabling automated quality assessment and improving overall detection performance on complex industrial diagrams.

Spatial-ViLT: Enhancing Visual Spatial Reasoning through Multi-Task Learning

arXiv:2510.03441v1 Announce Type: new Abstract: Vision-language models (VLMs) have advanced multimodal reasoning but still face challenges in spatial reasoning for 3D scenes and complex object configurations. To address this, we introduce SpatialViLT, an enhanced VLM that integrates spatial features like depth maps, 3D coordinates, and edge maps through a multi-task learning framework. This approach enriches multimodal embeddings with spatial understanding. We propose two variants: SpatialViLT and MaskedSpatialViLT, focusing on full and masked object regions, respectively. Additionally, SpatialEnsemble combines both approaches, achieving state-of-the-art accuracy. Our models excel in spatial reasoning categories such as directional, topological, and proximity relations, as demonstrated on the challenging Visual Spatial Reasoning (VSR) dataset. This work represents a significant step in enhancing the spatial intelligence of AI systems, crucial for advanced multimodal understanding and real-world applications.

Denoising of Two-Phase Optically Sectioned Structured Illumination Reconstructions Using Encoder-Decoder Networks

arXiv:2510.03452v1 Announce Type: new Abstract: Structured illumination (SI) enhances image resolution and contrast by projecting patterned light onto a sample. In two-phase optical-sectioning SI (OS-SI), reduced acquisition time introduces residual artifacts that conventional denoising struggles to suppress. Deep learning offers an alternative to traditional methods; however, supervised training is limited by the lack of clean, optically sectioned ground-truth data. We investigate encoder-decoder networks for artifact reduction in two-phase OS-SI, using synthetic training pairs formed by applying real artifact fields to synthetic images. An asymmetrical denoising autoencoder (DAE) and a U-Net are trained on the synthetic data, then evaluated on real OS-SI images. Both networks improve image clarity, with each excelling against different artifact types. These results demonstrate that synthetic training enables supervised denoising of OS-SI images and highlight the potential of encoder-decoder networks to streamline reconstruction workflows.

PEaRL: Pathway-Enhanced Representation Learning for Gene and Pathway Expression Prediction from Histology

arXiv:2510.03455v1 Announce Type: new Abstract: Integrating histopathology with spatial transcriptomics (ST) provides a powerful opportunity to link tissue morphology with molecular function. Yet most existing multimodal approaches rely on a small set of highly variable genes, which limits predictive scope and overlooks the coordinated biological programs that shape tissue phenotypes. We present PEaRL (Pathway Enhanced Representation Learning), a multimodal framework that represents transcriptomics through pathway activation scores computed with ssGSEA. By encoding biologically coherent pathway signals with a transformer and aligning them with histology features via contrastive learning, PEaRL reduces dimensionality, improves interpretability, and strengthens cross-modal correspondence. Across three cancer ST datasets (breast, skin, and lymph node), PEaRL consistently outperforms SOTA methods, yielding higher accuracy for both gene- and pathway-level expression prediction (up to 58.9 percent and 20.4 percent increase in Pearson correlation coefficient compared to SOTA). These results demonstrate that grounding transcriptomic representation in pathways produces more biologically faithful and interpretable multimodal models, advancing computational pathology beyond gene-level embeddings.

DuPLUS: Dual-Prompt Vision-Language Framework for Universal Medical Image Segmentation and Prognosis

arXiv:2510.03483v1 Announce Type: new Abstract: Deep learning for medical imaging is hampered by task-specific models that lack generalizability and prognostic capabilities, while existing 'universal' approaches suffer from simplistic conditioning and poor medical semantic understanding. To address these limitations, we introduce DuPLUS, a deep learning framework for efficient multi-modal medical image analysis. DuPLUS introduces a novel vision-language framework that leverages hierarchical semantic prompts for fine-grained control over the analysis task, a capability absent in prior universal models. To enable extensibility to other medical tasks, it includes a hierarchical, text-controlled architecture driven by a unique dual-prompt mechanism. For segmentation, DuPLUS is able to generalize across three imaging modalities, ten different anatomically various medical datasets, encompassing more than 30 organs and tumor types. It outperforms the state-of-the-art task specific and universal models on 8 out of 10 datasets. We demonstrate extensibility of its text-controlled architecture by seamless integration of electronic health record (EHR) data for prognosis prediction, and on a head and neck cancer dataset, DuPLUS achieved a Concordance Index (CI) of 0.69. Parameter-efficient fine-tuning enables rapid adaptation to new tasks and modalities from varying centers, establishing DuPLUS as a versatile and clinically relevant solution for medical image analysis. The code for this work is made available at: https://anonymous.4open.science/r/DuPLUS-6C52

Real-Time Threaded Houbara Detection and Segmentation for Wildlife Conservation using Mobile Platforms

arXiv:2510.03501v1 Announce Type: new Abstract: Real-time animal detection and segmentation in natural environments are vital for wildlife conservation, enabling non-invasive monitoring through remote camera streams. However, these tasks remain challenging due to limited computational resources and the cryptic appearance of many species. We propose a mobile-optimized two-stage deep learning framework that integrates a Threading Detection Model (TDM) to parallelize YOLOv10-based detection and MobileSAM-based segmentation. Unlike prior YOLO+SAM pipelines, our approach improves real-time performance by reducing latency through threading. YOLOv10 handles detection while MobileSAM performs lightweight segmentation, both executed concurrently for efficient resource use. On the cryptic Houbara Bustard, a conservation-priority species, our model achieves mAP50 of 0.9627, mAP75 of 0.7731, mAP95 of 0.7178, and a MobileSAM mIoU of 0.7421. YOLOv10 operates at 43.7 ms per frame, confirming real-time readiness. We introduce a curated Houbara dataset of 40,000 annotated images to support model training and evaluation across diverse conditions. The code and dataset used in this study are publicly available on GitHub at https://github.com/LyesSaadSaoud/mobile-houbara-detseg. For interactive demos and additional resources, visit https://lyessaadsaoud.github.io/LyesSaadSaoud-Threaded-YOLO-SAM-Houbara.

Platonic Transformers: A Solid Choice For Equivariance

arXiv:2510.03511v1 Announce Type: new Abstract: While widespread, Transformers lack inductive biases for geometric symmetries common in science and computer vision. Existing equivariant methods often sacrifice the efficiency and flexibility that make Transformers so effective through complex, computationally intensive designs. We introduce the Platonic Transformer to resolve this trade-off. By defining attention relative to reference frames from the Platonic solid symmetry groups, our method induces a principled weight-sharing scheme. This enables combined equivariance to continuous translations and Platonic symmetries, while preserving the exact architecture and computational cost of a standard Transformer. Furthermore, we show that this attention is formally equivalent to a dynamic group convolution, which reveals that the model learns adaptive geometric filters and enables a highly scalable, linear-time convolutional variant. Across diverse benchmarks in computer vision (CIFAR-10), 3D point clouds (ScanObjectNN), and molecular property prediction (QM9, OMol25), the Platonic Transformer achieves competitive performance by leveraging these geometric constraints at no additional cost.

Domain Generalization for Semantic Segmentation: A Survey

arXiv:2510.03540v1 Announce Type: new Abstract: The generalization of deep neural networks to unknown domains is a major challenge despite their tremendous progress in recent years. For this reason, the dynamic area of domain generalization (DG) has emerged. In contrast to unsupervised domain adaptation, there is no access to or knowledge about the target domains, and DG methods aim to generalize across multiple different unseen target domains. Domain generalization is particularly relevant for the task semantic segmentation which is used in several areas such as biomedicine or automated driving. This survey provides a comprehensive overview of the rapidly evolving topic of domain generalized semantic segmentation. We cluster and review existing approaches and identify the paradigm shift towards foundation-model-based domain generalization. Finally, we provide an extensive performance comparison of all approaches, which highlights the significant influence of foundation models on domain generalization. This survey seeks to advance domain generalization research and inspire scientists to explore new research directions.

From Scope to Script: An Automated Report Generation Model for Gastrointestinal Endoscopy

arXiv:2510.03543v1 Announce Type: new Abstract: Endoscopic procedures such as esophagogastroduodenoscopy (EGD) and colonoscopy play a critical role in diagnosing and managing gastrointestinal (GI) disorders. However, the documentation burden associated with these procedures place significant strain on gastroenterologists, contributing to inefficiencies in clinical workflows and physician burnout. To address this challenge, we propose a novel automated report generation model that leverages a transformer-based vision encoder and text decoder within a two-stage training framework. In the first stage, both components are pre-trained on image/text caption pairs to capture generalized vision-language features, followed by fine-tuning on images/report pairs to generate clinically meaningful findings. Our approach not only streamlines the documentation process but also holds promise for reducing physician workload and improving patient care.

SketchPlan: Diffusion Based Drone Planning From Human Sketches

arXiv:2510.03545v1 Announce Type: new Abstract: We propose SketchPlan, a diffusion-based planner that interprets 2D hand-drawn sketches over depth images to generate 3D flight paths for drone navigation. SketchPlan comprises two components: a SketchAdapter that learns to map the human sketches to projected 2D paths, and DiffPath, a diffusion model that infers 3D trajectories from 2D projections and a first person view depth image. Our model achieves zero-shot sim-to-real transfer, generating accurate and safe flight paths in previously unseen real-world environments. To train the model, we build a synthetic dataset of 32k flight paths using a diverse set of photorealistic 3D Gaussian Splatting scenes. We automatically label the data by computing 2D projections of the 3D flight paths onto the camera plane, and use this to train the DiffPath diffusion model. However, since real human 2D sketches differ significantly from ideal 2D projections, we additionally label 872 of the 3D flight paths with real human sketches and use this to train the SketchAdapter to infer the 2D projection from the human sketch. We demonstrate SketchPlan's effectiveness in both simulated and real-world experiments, and show through ablations that training on a mix of human labeled and auto-labeled data together with a modular design significantly boosts its capabilities to correctly interpret human intent and infer 3D paths. In real-world drone tests, SketchPlan achieved 100\% success in low/medium clutter and 40\% in unseen high-clutter environments, outperforming key ablations by 20-60\% in task completion.

Unmasking Puppeteers: Leveraging Biometric Leakage to Disarm Impersonation in AI-based Videoconferencing

arXiv:2510.03548v1 Announce Type: new Abstract: AI-based talking-head videoconferencing systems reduce bandwidth by sending a compact pose-expression latent and re-synthesizing RGB at the receiver, but this latent can be puppeteered, letting an attacker hijack a victim's likeness in real time. Because every frame is synthetic, deepfake and synthetic video detectors fail outright. To address this security problem, we exploit a key observation: the pose-expression latent inherently contains biometric information of the driving identity. Therefore, we introduce the first biometric leakage defense without ever looking at the reconstructed RGB video: a pose-conditioned, large-margin contrastive encoder that isolates persistent identity cues inside the transmitted latent while cancelling transient pose and expression. A simple cosine test on this disentangled embedding flags illicit identity swaps as the video is rendered. Our experiments on multiple talking-head generation models show that our method consistently outperforms existing puppeteering defenses, operates in real-time, and shows strong generalization to out-of-distribution scenarios.

Streaming Drag-Oriented Interactive Video Manipulation: Drag Anything, Anytime!

arXiv:2510.03550v1 Announce Type: new Abstract: Achieving streaming, fine-grained control over the outputs of autoregressive video diffusion models remains challenging, making it difficult to ensure that they consistently align with user expectations. To bridge this gap, we propose \textbf{stReaming drag-oriEnted interactiVe vidEo manipuLation (REVEL)}, a new task that enables users to modify generated videos \emph{anytime} on \emph{anything} via fine-grained, interactive drag. Beyond DragVideo and SG-I2V, REVEL unifies drag-style video manipulation as editing and animating video frames with both supporting user-specified translation, deformation, and rotation effects, making drag operations versatile. In resolving REVEL, we observe: \emph{i}) drag-induced perturbations accumulate in latent space, causing severe latent distribution drift that halts the drag process; \emph{ii}) streaming drag is easily disturbed by context frames, thereby yielding visually unnatural outcomes. We thus propose a training-free approach, \textbf{DragStream}, comprising: \emph{i}) an adaptive distribution self-rectification strategy that leverages neighboring frames' statistics to effectively constrain the drift of latent embeddings; \emph{ii}) a spatial-frequency selective optimization mechanism, allowing the model to fully exploit contextual information while mitigating its interference via selectively propagating visual cues along generation. Our method can be seamlessly integrated into existing autoregressive video diffusion models, and extensive experiments firmly demonstrate the effectiveness of our DragStream.

GAS-MIL: Group-Aggregative Selection Multi-Instance Learning for Ensemble of Foundation Models in Digital Pathology Image Analysis

arXiv:2510.03555v1 Announce Type: new Abstract: Foundation models (FMs) have transformed computational pathology by providing powerful, general-purpose feature extractors. However, adapting and benchmarking individual FMs for specific diagnostic tasks is often time-consuming and resource-intensive, especially given their scale and diversity. To address this challenge, we introduce Group-Aggregative Selection Multi-Instance Learning (GAS-MIL), a flexible ensemble framework that seamlessly integrates features from multiple FMs, preserving their complementary strengths without requiring manual feature selection or extensive task-specific fine-tuning. Across classification tasks in three cancer datasets-prostate (PANDA), ovarian (UBC-OCEAN), and breast (TCGA-BrCa)-GAS-MIL consistently achieves superior or on-par performance relative to individual FMs and established MIL methods, demonstrating its robustness and generalizability. By enabling efficient integration of heterogeneous FMs, GAS-MIL streamlines model deployment for pathology and provides a scalable foundation for future multimodal and precision oncology applications.

Real-Time Assessment of Bystander Situation Awareness in Drone-Assisted First Aid

arXiv:2510.03558v1 Announce Type: new Abstract: Rapid naloxone delivery via drones offers a promising solution for responding to opioid overdose emergencies (OOEs), by extending lifesaving interventions to medically untrained bystanders before emergency medical services (EMS) arrive. Recognizing the critical role of bystander situational awareness (SA) in human-autonomy teaming (HAT), we address a key research gap in real-time SA assessment by introducing the Drone-Assisted Naloxone Delivery Simulation Dataset (DANDSD). This pioneering dataset captures HAT during simulated OOEs, where college students without medical training act as bystanders tasked with administering intranasal naloxone to a mock overdose victim. Leveraging this dataset, we propose a video-based real-time SA assessment framework that utilizes graph embeddings and transformer models to assess bystander SA in real time. Our approach integrates visual perception and comprehension cues--such as geometric, kinematic, and interaction graph features--and achieves high-performance SA prediction. It also demonstrates strong temporal segmentation accuracy, outperforming the FINCH baseline by 9% in Mean over Frames (MoF) and 5% in Intersection over Union (IoU). This work supports the development of adaptive drone systems capable of guiding bystanders effectively, ultimately improving emergency response outcomes and saving lives.

Evaluating OCR performance on food packaging labels in South Africa

arXiv:2510.03570v1 Announce Type: new Abstract: This study evaluates four open-source Optical Character Recognition (OCR) systems which are Tesseract, EasyOCR, PaddleOCR, and TrOCR on real world food packaging images. The aim is to assess their ability to extract ingredient lists and nutrition facts panels. Accurate OCR for packaging is important for compliance and nutrition monitoring but is challenging due to multilingual text, dense layouts, varied fonts, glare, and curved surfaces. A dataset of 231 products (1,628 images) was processed by all four models to assess speed and coverage, and a ground truth subset of 113 images (60 products) was created for accuracy evaluation. Metrics include Character Error Rate (CER), Word Error Rate (WER), BLEU, ROUGE-L, F1, coverage, and execution time. On the ground truth subset, Tesseract achieved the lowest CER (0.912) and the highest BLEU (0.245). EasyOCR provided a good balance between accuracy and multilingual support. PaddleOCR achieved near complete coverage but was slower because it ran on CPU only due to GPU incompatibility, and TrOCR produced the weakest results despite GPU acceleration. These results provide a packaging-specific benchmark, establish a baseline, and highlight directions for layout-aware methods and text localization.

FrameOracle: Learning What to See and How Much to See in Videos

arXiv:2510.03584v1 Announce Type: new Abstract: Vision-language models (VLMs) have advanced video understanding, but their performance is limited by the number of input frames they can process. Existing frame sampling strategies, such as uniform or fixed-budget selection, often fail to adapt to variations in information density or task complexity, resulting in inefficiency and information loss. To address this, we present FrameOracle, a lightweight and plug-and-play module that predicts both (1) which frames are most relevant to a given query and (2) how many frames are needed. FrameOracle is trained using a four-stage curriculum, with the first three stages relying on weak proxy signals such as cross-modal similarity. In the final stage, it leverages stronger supervision from a new dataset we introduce, FrameOracle-41K, the first large-scale VideoQA collection to provide keyframe annotations specifying the minimal set of frames required to answer each question. Extensive experiments across five VLMs and six benchmarks demonstrate that FrameOracle reduces 16-frame inputs to an average of 10.4 frames without any loss in accuracy. When starting from 64-frame candidates, it reduces the input to an average of 13.9 frames while improving accuracy by 1.4%, achieving state-of-the-art efficiency-accuracy trade-offs for scalable video understanding.

A Hybrid Co-Finetuning Approach for Visual Bug Detection in Video Games

arXiv:2510.03591v1 Announce Type: new Abstract: Manual identification of visual bugs in video games is a resource-intensive and costly process, often demanding specialized domain knowledge. While supervised visual bug detection models offer a promising solution, their reliance on extensive labeled datasets presents a significant challenge due to the infrequent occurrence of such bugs. To overcome this limitation, we propose a hybrid Co-FineTuning (CFT) method that effectively integrates both labeled and unlabeled data. Our approach leverages labeled samples from the target game and diverse co-domain games, additionally incorporating unlabeled data to enhance feature representation learning. This strategy maximizes the utility of all available data, substantially reducing the dependency on labeled examples from the specific target game. The developed framework demonstrates enhanced scalability and adaptability, facilitating efficient visual bug detection across various game titles. Our experimental results show the robustness of the proposed method for game visual bug detection, exhibiting superior performance compared to conventional baselines across multiple gaming environments. Furthermore, CFT maintains competitive performance even when trained with only 50% of the labeled data from the target game.

Exploring the Hierarchical Reasoning Model for Small Natural-Image Classification Without Augmentation

arXiv:2510.03598v1 Announce Type: new Abstract: This paper asks whether the Hierarchical Reasoning Model (HRM) with the two Transformer-style modules $(f_L,f_H)$, one step (DEQ-style) training, deep supervision, Rotary Position Embeddings, and RMSNorm can serve as a practical image classifier. It is evaluated on MNIST, CIFAR-10, and CIFAR-100 under a deliberately raw regime: no data augmentation, identical optimizer family with one-epoch warmup then cosine-floor decay, and label smoothing. HRM optimizes stably and performs well on MNIST ($\approx 98\%$ test accuracy), but on small natural images it overfits and generalizes poorly: on CIFAR-10, HRM reaches 65.0\% after 25 epochs, whereas a two-stage Conv--BN--ReLU baseline attains 77.2\% while training $\sim 30\times$ faster per epoch; on CIFAR-100, HRM achieves only 29.7\% test accuracy despite 91.5\% train accuracy, while the same CNN reaches 45.3\% test with 50.5\% train accuracy. Loss traces and error analyses indicate healthy optimization but insufficient image-specific inductive bias for HRM in this regime. It is concluded that, for small-resolution image classification without augmentation, HRM is not competitive with even simple convolutional architectures as the HRM currently exist but this does not exclude possibilities that modifications to the model may allow it to improve greatly.

Unsupervised Transformer Pre-Training for Images: Self-Distillation, Mean Teachers, and Random Crops

arXiv:2510.03606v1 Announce Type: new Abstract: Recent advances in self-supervised learning (SSL) have made it possible to learn general-purpose visual features that capture both the high-level semantics and the fine-grained spatial structure of images. Most notably, the recent DINOv2 has established a new state of the art by surpassing weakly supervised methods (WSL) like OpenCLIP on most benchmarks. In this survey, we examine the core ideas behind its approach, multi-crop view augmentation and self-distillation with a mean teacher, and trace their development in previous work. We then compare the performance of DINO and DINOv2 with other SSL and WSL methods across various downstream tasks, and highlight some remarkable emergent properties of their learned features with transformer backbones. We conclude by briefly discussing DINOv2's limitations, its impact, and future research directions.

Diffusion-Classifier Synergy: Reward-Aligned Learning via Mutual Boosting Loop for FSCIL

arXiv:2510.03608v1 Announce Type: new Abstract: Few-Shot Class-Incremental Learning (FSCIL) challenges models to sequentially learn new classes from minimal examples without forgetting prior knowledge, a task complicated by the stability-plasticity dilemma and data scarcity. Current FSCIL methods often struggle with generalization due to their reliance on limited datasets. While diffusion models offer a path for data augmentation, their direct application can lead to semantic misalignment or ineffective guidance. This paper introduces Diffusion-Classifier Synergy (DCS), a novel framework that establishes a mutual boosting loop between diffusion model and FSCIL classifier. DCS utilizes a reward-aligned learning strategy, where a dynamic, multi-faceted reward function derived from the classifier's state directs the diffusion model. This reward system operates at two levels: the feature level ensures semantic coherence and diversity using prototype-anchored maximum mean discrepancy and dimension-wise variance matching, while the logits level promotes exploratory image generation and enhances inter-class discriminability through confidence recalibration and cross-session confusion-aware mechanisms. This co-evolutionary process, where generated images refine the classifier and an improved classifier state yields better reward signals, demonstrably achieves state-of-the-art performance on FSCIL benchmarks, significantly enhancing both knowledge retention and new class learning.

MonitorVLM:A Vision Language Framework for Safety Violation Detection in Mining Operations

arXiv:2510.03666v1 Announce Type: new Abstract: Industrial accidents, particularly in high-risk domains such as surface and underground mining, are frequently caused by unsafe worker behaviors. Traditional manual inspection remains labor-intensive, error-prone, and insufficient for large-scale, dynamic environments, highlighting the urgent need for intelligent and automated safety monitoring. In this paper, we present MonitorVLM, a novel vision--language framework designed to detect safety violations directly from surveillance video streams. MonitorVLM introduces three key innovations: (1) a domain-specific violation dataset comprising 9,000 vision--question--answer (VQA) samples across 40 high-frequency mining regulations, enriched with augmentation and auxiliary detection cues; (2) a clause filter (CF) module that dynamically selects the Top-$K$ most relevant clauses, reducing inference latency by 13.56\% while maintaining accuracy; and (3) a behavior magnifier (BM) module that enhances worker regions to improve fine-grained action recognition, yielding additional gains of 3.45% in precision and 8.62% in recall. Experimental results demonstrate that MonitorVLM significantly outperforms baseline vision--language models, achieving improvements of 22.01% in precision, 34.22\% in recall, and 28.37% in F1 score over the 72B unfine-tuned baseline. A lightweight web-based interface further integrates MonitorVLM into practical workflows, enabling automatic violation reporting with video timestamping. This study highlights the potential of multimodal large models to enhance occupational safety monitoring in mining and beyond.

A Novel Cloud-Based Diffusion-Guided Hybrid Model for High-Accuracy Accident Detection in Intelligent Transportation Systems

arXiv:2510.03675v1 Announce Type: new Abstract: The integration of Diffusion Models into Intelligent Transportation Systems (ITS) is a substantial improvement in the detection of accidents. We present a novel hybrid model integrating guidance classification with diffusion techniques. By leveraging fine-tuned ExceptionNet architecture outputs as input for our proposed diffusion model and processing image tensors as our conditioning, our approach creates a robust classification framework. Our model consists of multiple conditional modules, which aim to modulate the linear projection of inputs using time embeddings and image covariate embeddings, allowing the network to adapt its behavior dynamically throughout the diffusion process. To address the computationally intensive nature of diffusion models, our implementation is cloud-based, enabling scalable and efficient processing. Our strategy overcomes the shortcomings of conventional classification approaches by leveraging diffusion models inherent capacity to effectively understand complicated data distributions. We investigate important diffusion characteristics, such as timestep schedulers, timestep encoding techniques, timestep count, and architectural design changes, using a thorough ablation study, and have conducted a comprehensive evaluation of the proposed model against the baseline models on a publicly available dataset. The proposed diffusion model performs best in image-based accident detection with an accuracy of 97.32%.

SAMSOD: Rethinking SAM Optimization for RGB-T Salient Object Detection

arXiv:2510.03689v1 Announce Type: new Abstract: RGB-T salient object detection (SOD) aims to segment attractive objects by combining RGB and thermal infrared images. To enhance performance, the Segment Anything Model has been fine-tuned for this task. However, the imbalance convergence of two modalities and significant gradient difference between high- and low- activations are ignored, thereby leaving room for further performance enhancement. In this paper, we propose a model called \textit{SAMSOD}, which utilizes unimodal supervision to enhance the learning of non-dominant modality and employs gradient deconfliction to reduce the impact of conflicting gradients on model convergence. The method also leverages two decoupled adapters to separately mask high- and low-activation neurons, emphasizing foreground objects by enhancing background learning. Fundamental experiments on RGB-T SOD benchmark datasets and generalizability experiments on scribble supervised RGB-T SOD, fully supervised RGB-D SOD datasets and full-supervised RGB-D rail surface defect detection all demonstrate the effectiveness of our proposed method.

Referring Expression Comprehension for Small Objects

arXiv:2510.03701v1 Announce Type: new Abstract: Referring expression comprehension (REC) aims to localize the target object described by a natural language expression. Recent advances in vision-language learning have led to significant performance improvements in REC tasks. However, localizing extremely small objects remains a considerable challenge despite its importance in real-world applications such as autonomous driving. To address this issue, we introduce a novel dataset and method for REC targeting small objects. First, we present the small object REC (SOREC) dataset, which consists of 100,000 pairs of referring expressions and corresponding bounding boxes for small objects in driving scenarios. Second, we propose the progressive-iterative zooming adapter (PIZA), an adapter module for parameter-efficient fine-tuning that enables models to progressively zoom in and localize small objects. In a series of experiments, we apply PIZA to GroundingDINO and demonstrate a significant improvement in accuracy on the SOREC dataset. Our dataset, codes and pre-trained models are publicly available on the project page.

Artery-Vein Segmentation from Fundus Images using Deep Learning

arXiv:2510.03717v1 Announce Type: new Abstract: Segmenting of clinically important retinal blood vessels into arteries and veins is a prerequisite for retinal vessel analysis. Such analysis can provide potential insights and bio-markers for identifying and diagnosing various retinal eye diseases. Alteration in the regularity and width of the retinal blood vessels can act as an indicator of the health of the vasculature system all over the body. It can help identify patients at high risk of developing vasculature diseases like stroke and myocardial infarction. Over the years, various Deep Learning architectures have been proposed to perform retinal vessel segmentation. Recently, attention mechanisms have been increasingly used in image segmentation tasks. The work proposes a new Deep Learning approach for artery-vein segmentation. The new approach is based on the Attention mechanism that is incorporated into the WNet Deep Learning model, and we call the model as Attention-WNet. The proposed approach has been tested on publicly available datasets such as HRF and DRIVE datasets. The proposed approach has outperformed other state-of-art models available in the literature.

Person-Centric Annotations of LAION-400M: Auditing Bias and Its Transfer to Models

arXiv:2510.03721v1 Announce Type: new Abstract: Vision-language models trained on large-scale multimodal datasets show strong demographic biases, but the role of training data in producing these biases remains unclear. A major barrier has been the lack of demographic annotations in web-scale datasets such as LAION-400M. We address this gap by creating person-centric annotations for the full dataset, including over 276 million bounding boxes, perceived gender and race/ethnicity labels, and automatically generated captions. These annotations are produced through validated automatic labeling pipelines combining object detection, multimodal captioning, and finetuned classifiers. Using them, we uncover demographic imbalances and harmful associations, such as the disproportionate linking of men and individuals perceived as Black or Middle Eastern with crime-related and negative content. We also show that 60-70% of gender bias in CLIP and Stable Diffusion can be linearly explained by direct co-occurrences in the data. Our resources establish the first large-scale empirical link between dataset composition and downstream model bias.

Mapping Rio de Janeiro's favelas: general-purpose vs. satellite-specific neural networks

arXiv:2510.03725v1 Announce Type: new Abstract: While deep learning methods for detecting informal settlements have already been developed, they have not yet fully utilized the potential offered by recent pretrained neural networks. We compare two types of pretrained neural networks for detecting the favelas of Rio de Janeiro: 1. Generic networks pretrained on large diverse datasets of unspecific images, 2. A specialized network pretrained on satellite imagery. While the latter is more specific to the target task, the former has been pretrained on significantly more images. Hence, this research investigates whether task specificity or data volume yields superior performance in urban informal settlement detection.

LoRA Patching: Exposing the Fragility of Proactive Defenses against Deepfakes

arXiv:2510.03747v1 Announce Type: new Abstract: Deepfakes pose significant societal risks, motivating the development of proactive defenses that embed adversarial perturbations in facial images to prevent manipulation. However, in this paper, we show that these preemptive defenses often lack robustness and reliability. We propose a novel approach, Low-Rank Adaptation (LoRA) patching, which injects a plug-and-play LoRA patch into Deepfake generators to bypass state-of-the-art defenses. A learnable gating mechanism adaptively controls the effect of the LoRA patch and prevents gradient explosions during fine-tuning. We also introduce a Multi-Modal Feature Alignment (MMFA) loss, encouraging the features of adversarial outputs to align with those of the desired outputs at the semantic level. Beyond bypassing, we present defensive LoRA patching, embedding visible warnings in the outputs as a complementary solution to mitigate this newly identified security vulnerability. With only 1,000 facial examples and a single epoch of fine-tuning, LoRA patching successfully defeats multiple proactive defenses. These results reveal a critical weakness in current paradigms and underscore the need for more robust Deepfake defense strategies. Our code is available at https://github.com/ZOMIN28/LoRA-Patching.

The Overlooked Value of Test-time Reference Sets in Visual Place Recognition

arXiv:2510.03751v1 Announce Type: new Abstract: Given a query image, Visual Place Recognition (VPR) is the task of retrieving an image of the same place from a reference database with robustness to viewpoint and appearance changes. Recent works show that some VPR benchmarks are solved by methods using Vision-Foundation-Model backbones and trained on large-scale and diverse VPR-specific datasets. Several benchmarks remain challenging, particularly when the test environments differ significantly from the usual VPR training datasets. We propose a complementary, unexplored source of information to bridge the train-test domain gap, which can further improve the performance of State-of-the-Art (SOTA) VPR methods on such challenging benchmarks. Concretely, we identify that the test-time reference set, the "map", contains images and poses of the target domain, and must be available before the test-time query is received in several VPR applications. Therefore, we propose to perform simple Reference-Set-Finetuning (RSF) of VPR models on the map, boosting the SOTA (~2.3% increase on average for Recall@1) on these challenging datasets. Finetuned models retain generalization, and RSF works across diverse test datasets.

Adaptively Sampling-Reusing-Mixing Decomposed Gradients to Speed Up Sharpness Aware Minimization

arXiv:2510.03763v1 Announce Type: new Abstract: Sharpness-Aware Minimization (SAM) improves model generalization but doubles the computational cost of Stochastic Gradient Descent (SGD) by requiring twice the gradient calculations per optimization step. To mitigate this, we propose Adaptively sampling-Reusing-mixing decomposed gradients to significantly accelerate SAM (ARSAM). Concretely, we firstly discover that SAM's gradient can be decomposed into the SGD gradient and the Projection of the Second-order gradient onto the First-order gradient (PSF). Furthermore, we observe that the SGD gradient and PSF dynamically evolve during training, emphasizing the growing role of the PSF to achieve a flat minima. Therefore, ARSAM is proposed to the reused PSF and the timely updated PSF still maintain the model's generalization ability. Extensive experiments show that ARSAM achieves state-of-the-art accuracies comparable to SAM across diverse network architectures. On CIFAR-10/100, ARSAM is comparable to SAM while providing a speedup of about 40\%. Moreover, ARSAM accelerates optimization for the various challenge tasks (\textit{e.g.}, human pose estimation, and model quantization) without sacrificing performance, demonstrating its broad practicality.% The code is publicly accessible at: https://github.com/ajiaaa/ARSAM.

CoPA: Hierarchical Concept Prompting and Aggregating Network for Explainable Diagnosis

arXiv:2510.03767v1 Announce Type: new Abstract: The transparency of deep learning models is essential for clinical diagnostics. Concept Bottleneck Model provides clear decision-making processes for diagnosis by transforming the latent space of black-box models into human-understandable concepts. However, concept-based methods still face challenges in concept capture capabilities. These methods often rely on encode features solely from the final layer, neglecting shallow and multiscale features, and lack effective guidance in concept encoding, hindering fine-grained concept extraction. To address these issues, we introduce Concept Prompting and Aggregating (CoPA), a novel framework designed to capture multilayer concepts under prompt guidance. This framework utilizes the Concept-aware Embedding Generator (CEG) to extract concept representations from each layer of the visual encoder. Simultaneously, these representations serve as prompts for Concept Prompt Tuning (CPT), steering the model towards amplifying critical concept-related visual cues. Visual representations from each layer are aggregated to align with textual concept representations. With the proposed method, valuable concept-wise information in the images is captured and utilized effectively, thus improving the performance of concept and disease prediction. Extensive experimental results demonstrate that CoPA outperforms state-of-the-art methods on three public datasets. Code is available at https://github.com/yihengd/CoPA.

Efficiency vs. Efficacy: Assessing the Compression Ratio-Dice Score Relationship through a Simple Benchmarking Framework for Cerebrovascular 3D Segmentation

arXiv:2510.03769v1 Announce Type: new Abstract: The increasing size and complexity of medical imaging datasets, particularly in 3D formats, present significant barriers to collaborative research and transferability. This study investigates whether the ZFP compression technique can mitigate these challenges without compromising the performance of automated cerebrovascular segmentation, a critical first step in intracranial aneurysm detection. We apply ZFP in both its error tolerance and fixed-rate modes to a large scale, and one of the most recent, datasets in the literature, 3D medical dataset containing ground-truth vascular segmentations. The segmentation quality on the compressed volumes is rigorously compared to the uncompressed baseline (Dice approximately equals 0.8774). Our findings reveal that ZFP can achieve substantial data reduction--up to a 22.89:1 ratio in error tolerance mode--while maintaining a high degree of fidelity, with the mean Dice coefficient remaining high at 0.87656. These results demonstrate that ZFP is a viable and powerful tool for enabling more efficient and accessible research on large-scale medical datasets, fostering broader collaboration across the community.

MambaCAFU: Hybrid Multi-Scale and Multi-Attention Model with Mamba-Based Fusion for Medical Image Segmentation

arXiv:2510.03786v1 Announce Type: new Abstract: In recent years, deep learning has shown near-expert performance in segmenting complex medical tissues and tumors. However, existing models are often task-specific, with performance varying across modalities and anatomical regions. Balancing model complexity and performance remains challenging, particularly in clinical settings where both accuracy and efficiency are critical. To address these issues, we propose a hybrid segmentation architecture featuring a three-branch encoder that integrates CNNs, Transformers, and a Mamba-based Attention Fusion (MAF) mechanism to capture local, global, and long-range dependencies. A multi-scale attention-based CNN decoder reconstructs fine-grained segmentation maps while preserving contextual consistency. Additionally, a co-attention gate enhances feature selection by emphasizing relevant spatial and semantic information across scales during both encoding and decoding, improving feature interaction and cross-scale communication. Extensive experiments on multiple benchmark datasets show that our approach outperforms state-of-the-art methods in accuracy and generalization, while maintaining comparable computational complexity. By effectively balancing efficiency and effectiveness, our architecture offers a practical and scalable solution for diverse medical imaging tasks. Source code and trained models will be publicly released upon acceptance to support reproducibility and further research.

Road Damage and Manhole Detection using Deep Learning for Smart Cities: A Polygonal Annotation Approach

arXiv:2510.03797v1 Announce Type: new Abstract: Urban safety and infrastructure maintenance are critical components of smart city development. Manual monitoring of road damages is time-consuming, highly costly, and error-prone. This paper presents a deep learning approach for automated road damage and manhole detection using the YOLOv9 algorithm with polygonal annotations. Unlike traditional bounding box annotation, we employ polygonal annotations for more precise localization of road defects. We develop a novel dataset comprising more than one thousand images which are mostly collected from Dhaka, Bangladesh. This dataset is used to train a YOLO-based model for three classes, namely Broken, Not Broken, and Manhole. We achieve 78.1% overall image-level accuracy. The YOLOv9 model demonstrates strong performance for Broken (86.7% F1-score) and Not Broken (89.2% F1-score) classes, with challenges in Manhole detection (18.2% F1-score) due to class imbalance. Our approach offers an efficient and scalable solution for monitoring urban infrastructure in developing countries.

Contrastive-SDE: Guiding Stochastic Differential Equations with Contrastive Learning for Unpaired Image-to-Image Translation

arXiv:2510.03821v1 Announce Type: new Abstract: Unpaired image-to-image translation involves learning mappings between source domain and target domain in the absence of aligned or corresponding samples. Score based diffusion models have demonstrated state-of-the-art performance in generative tasks. Their ability to approximate complex data distributions through stochastic differential equations (SDEs) enables them to generate high-fidelity and diverse outputs, making them particularly well-suited for unpaired I2I settings. In parallel, contrastive learning provides a powerful framework for learning semantic similarities without the need for explicit supervision or paired data. By pulling together representations of semantically similar samples and pushing apart dissimilar ones, contrastive methods are inherently aligned with the objectives of unpaired translation. Its ability to selectively enforce semantic consistency at the feature level makes contrastive learning particularly effective for guiding generation in unpaired scenarios. In this work, we propose a time-dependent contrastive learning approach where a model is trained with SimCLR by considering an image and its domain invarient feature as a positive pair, enabling the preservation of domain-invariant features and the discarding of domain-specific ones. The learned contrastive model then guides the inference of a pretrained SDE for the I2I translation task. We empirically compare Contrastive-SDE with several baselines across three common unpaired I2I tasks, using four metrics for evaluation. Constrastive-SDE achieves comparable results to the state-of-the-art on several metrics. Furthermore, we observe that our model converges significantly faster and requires no label supervision or classifier training, making it a more efficient alternative for this task.

LIBERO-PRO: Towards Robust and Fair Evaluation of Vision-Language-Action Models Beyond Memorization

arXiv:2510.03827v1 Announce Type: new Abstract: LIBERO has emerged as a widely adopted benchmark for evaluating Vision-Language-Action (VLA) models; however, its current training and evaluation settings are problematic, often leading to inflated performance estimates and preventing fair model comparison. To address these issues, we introduce LIBERO-PRO, an extended LIBERO benchmark that systematically evaluates model performance under reasonable perturbations across four dimensions: manipulated objects, initial states, task instructions, and environments. Experimental results reveal that, although existing models achieve over 90% accuracy under the standard LIBERO evaluation, their performance collapses to 0.0% under our generalized setting. Crucially, this discrepancy exposes the models' reliance on rote memorization of action sequences and environment layouts from the training set, rather than genuine task understanding or environmental perception. For instance, models persist in executing grasping actions when the target object is replaced with irrelevant items, and their outputs remain unchanged even when given corrupted instructions or even messy tokens. These findings expose the severe flaws in current evaluation practices, and we call on the community to abandon misleading methodologies in favor of robust assessments of model generalization and comprehension. Our code is available at: https://github.com/Zxy-MLlab/LIBERO-PRO.

Mirage: Unveiling Hidden Artifacts in Synthetic Images with Large Vision-Language Models

arXiv:2510.03840v1 Announce Type: new Abstract: Recent advances in image generation models have led to models that produce synthetic images that are increasingly difficult for standard AI detectors to identify, even though they often remain distinguishable by humans. To identify this discrepancy, we introduce \textbf{Mirage}, a curated dataset comprising a diverse range of AI-generated images exhibiting visible artifacts, where current state-of-the-art detection methods largely fail. Furthermore, we investigate whether Large Vision-Language Models (LVLMs), which are increasingly employed as substitutes for human judgment in various tasks, can be leveraged for explainable AI image detection. Our experiments on both Mirage and existing benchmark datasets demonstrate that while LVLMs are highly effective at detecting AI-generated images with visible artifacts, their performance declines when confronted with images lacking such cues.

UGround: Towards Unified Visual Grounding with Unrolled Transformers

arXiv:2510.03853v1 Announce Type: new Abstract: We present UGround, a \textbf{U}nified visual \textbf{Ground}ing paradigm that dynamically selects intermediate layers across \textbf{U}nrolled transformers as mask as prompt'', diverging from the prevailing pipeline that leverages the fixed last hidden layer as\texttt{} as prompt''. UGround addresses two primary challenges posed by the prevailing paradigm: (1) its reliance on the fixed last hidden layer, which sequentially amplifies cumulative errors arising from layer-by-layer propagation without intermediate correction, and (2) its use of \texttt{} as a prompt, which implicitly projects textual embeddings into visual space without explicit spatial cues (\eg, coordinates). Central to UGround is Policy-Prompted Masking, which comprises two key components: Stochastic Skip Connection (SSC) and Mask as Prompt (MasP). SSC is a reinforcement learning policy that, via stochastic sampling, allows each \texttt{} token to slide across unrolled transformer layers, enabling dynamic layer selection at which it connects to the vision model (\eg, SAM) in a skip-connection fashion. Given the selected hidden layer, MasP uses the similarity map derived from the \texttt{} token and image tokens as a soft logit mask to prompt SAM for mask generation, offering explicit spatial cues through its activation regions. To validate the effectiveness of UGround, we, for the first time, have unified visual grounding within a single framework from an attribute perspective, spanning from traditional refer expression segmentation to newly proposed reasoning segmentation, single-target to multi-target, positive query to false premise (empty target). All codes and models are publicly available at \href{https://github.com/rui-qian/UGround}{https://github.com/rui-qian/UGround}.

Optimized Minimal 4D Gaussian Splatting

arXiv:2510.03857v1 Announce Type: new Abstract: 4D Gaussian Splatting has emerged as a new paradigm for dynamic scene representation, enabling real-time rendering of scenes with complex motions. However, it faces a major challenge of storage overhead, as millions of Gaussians are required for high-fidelity reconstruction. While several studies have attempted to alleviate this memory burden, they still face limitations in compression ratio or visual quality. In this work, we present OMG4 (Optimized Minimal 4D Gaussian Splatting), a framework that constructs a compact set of salient Gaussians capable of faithfully representing 4D Gaussian models. Our method progressively prunes Gaussians in three stages: (1) Gaussian Sampling to identify primitives critical to reconstruction fidelity, (2) Gaussian Pruning to remove redundancies, and (3) Gaussian Merging to fuse primitives with similar characteristics. In addition, we integrate implicit appearance compression and generalize Sub-Vector Quantization (SVQ) to 4D representations, further reducing storage while preserving quality. Extensive experiments on standard benchmark datasets demonstrate that OMG4 significantly outperforms recent state-of-the-art methods, reducing model sizes by over 60% while maintaining reconstruction quality. These results position OMG4 as a significant step forward in compact 4D scene representation, opening new possibilities for a wide range of applications. Our source code is available at https://minshirley.github.io/OMG4/.

Cross-View Open-Vocabulary Object Detection in Aerial Imagery

arXiv:2510.03858v1 Announce Type: new Abstract: Traditional object detection models are typically trained on a fixed set of classes, limiting their flexibility and making it costly to incorporate new categories. Open-vocabulary object detection addresses this limitation by enabling models to identify unseen classes without explicit training. Leveraging pretrained models contrastively trained on abundantly available ground-view image-text classification pairs provides a strong foundation for open-vocabulary object detection in aerial imagery. Domain shifts, viewpoint variations, and extreme scale differences make direct knowledge transfer across domains ineffective, requiring specialized adaptation strategies. In this paper, we propose a novel framework for adapting open-vocabulary representations from ground-view images to solve object detection in aerial imagery through structured domain alignment. The method introduces contrastive image-to-image alignment to enhance the similarity between aerial and ground-view embeddings and employs multi-instance vocabulary associations to align aerial images with text embeddings. Extensive experiments on the xView, DOTAv2, VisDrone, DIOR, and HRRSD datasets are used to validate our approach. Our open-vocabulary model achieves improvements of +6.32 mAP on DOTAv2, +4.16 mAP on VisDrone (Images), and +3.46 mAP on HRRSD in the zero-shot setting when compared to finetuned closed-vocabulary dataset-specific model performance, thus paving the way for more flexible and scalable object detection systems in aerial applications.

Exploring the Challenge and Value of Deep Learning in Automated Skin Disease Diagnosis

arXiv:2510.03869v1 Announce Type: new Abstract: Skin cancer is one of the most prevalent and deadly forms of cancer worldwide, which highlights the critical importance of early detection and diagnosis in improving patient outcomes. Deep learning (DL) has shown significant promise in enhancing the accuracy and efficiency of automated skin disease diagnosis, particularly in detecting and evaluating skin lesions and classification. However, there are still several challenges for DL-based skin cancer diagnosis, including complex features, image noise, intra-class variation, inter-class similarity, and data imbalance. By synthesizing recent research, this review discusses innovative approaches to cope with these challenges, such as data augmentation, hybrid models, and feature fusion, etc. Furthermore, the review highlights the integration of DL models into clinical workflows, offering insights into the potential of deep learning to revolutionize skin disease diagnosis and improve clinical decision-making. This article follows a comprehensive methodology based on the PRISMA framework and emphasizes the need for continued advancements to fully unlock the transformative potential of DL in dermatological care.

SDAKD: Student Discriminator Assisted Knowledge Distillation for Super-Resolution Generative Adversarial Networks

arXiv:2510.03870v1 Announce Type: new Abstract: Generative Adversarial Networks (GANs) achieve excellent performance in generative tasks, such as image super-resolution, but their computational requirements make difficult their deployment on resource-constrained devices. While knowledge distillation is a promising research direction for GAN compression, effectively training a smaller student generator is challenging due to the capacity mismatch between the student generator and the teacher discriminator. In this work, we propose Student Discriminator Assisted Knowledge Distillation (SDAKD), a novel GAN distillation methodology that introduces a student discriminator to mitigate this capacity mismatch. SDAKD follows a three-stage training strategy, and integrates an adapted feature map distillation approach in its last two training stages. We evaluated SDAKD on two well-performing super-resolution GANs, GCFSR and Real-ESRGAN. Our experiments demonstrate consistent improvements over the baselines and SOTA GAN knowledge distillation methods. The SDAKD source code will be made openly available upon acceptance of the paper.

PoseGaze-AHP: A Knowledge-Based 3D Dataset for AI-Driven Ocular and Postural Diagnosis

arXiv:2510.03873v1 Announce Type: new Abstract: Diagnosing ocular-induced abnormal head posture (AHP) requires a comprehensive analysis of both head pose and ocular movements. However, existing datasets focus on these aspects separately, limiting the development of integrated diagnostic approaches and restricting AI-driven advancements in AHP analysis. To address this gap, we introduce PoseGaze-AHP, a novel 3D dataset that synchronously captures head pose and gaze movement information for ocular-induced AHP assessment. Structured clinical data were extracted from medical literature using large language models (LLMs) through an iterative process with the Claude 3.5 Sonnet model, combining stepwise, hierarchical, and complex prompting strategies. The extracted records were systematically imputed and transformed into 3D representations using the Neural Head Avatar (NHA) framework. The dataset includes 7,920 images generated from two head textures, covering a broad spectrum of ocular conditions. The extraction method achieved an overall accuracy of 91.92%, demonstrating its reliability for clinical dataset construction. PoseGaze-AHP is the first publicly available resource tailored for AI-driven ocular-induced AHP diagnosis, supporting the development of accurate and privacy-compliant diagnostic tools.

DHQA-4D: Perceptual Quality Assessment of Dynamic 4D Digital Human

arXiv:2510.03874v1 Announce Type: new Abstract: With the rapid development of 3D scanning and reconstruction technologies, dynamic digital human avatars based on 4D meshes have become increasingly popular. A high-precision dynamic digital human avatar can be applied to various fields such as game production, animation generation, and remote immersive communication. However, these 4D human avatar meshes are prone to being degraded by various types of noise during the processes of collection, compression, and transmission, thereby affecting the viewing experience of users. In light of this fact, quality assessment of dynamic 4D digital humans becomes increasingly important. In this paper, we first propose a large-scale dynamic digital human quality assessment dataset, DHQA-4D, which contains 32 high-quality real-scanned 4D human mesh sequences, 1920 distorted textured 4D human meshes degraded by 11 textured distortions, as well as their corresponding textured and non-textured mean opinion scores (MOSs). Equipped with DHQA-4D dataset, we analyze the influence of different types of distortion on human perception for textured dynamic 4D meshes and non-textured dynamic 4D meshes. Additionally, we propose DynaMesh-Rater, a novel large multimodal model (LMM) based approach that is able to assess both textured 4D meshes and non-textured 4D meshes. Concretely, DynaMesh-Rater elaborately extracts multi-dimensional features, including visual features from a projected 2D video, motion features from cropped video clips, and geometry features from the 4D human mesh to provide comprehensive quality-related information. Then we utilize a LMM model to integrate the multi-dimensional features and conduct a LoRA-based instruction tuning technique to teach the LMM model to predict the quality scores. Extensive experimental results on the DHQA-4D dataset demonstrate the superiority of our DynaMesh-Rater method over previous quality assessment methods.

Skin Lesion Classification Based on ResNet-50 Enhanced With Adaptive Spatial Feature Fusion

arXiv:2510.03876v1 Announce Type: new Abstract: Skin cancer classification remains a challenging problem due to high inter-class similarity, intra-class variability, and image noise in dermoscopic images. To address these issues, we propose an improved ResNet-50 model enhanced with Adaptive Spatial Feature Fusion (ASFF), which adaptively integrates multi-scale semantic and surface features to improve feature representation and reduce overfitting. The ResNet-50 model is enhanced with an adaptive feature fusion mechanism to achieve more effective multi-scale feature extraction and improve overall performance. Specifically, a dual-branch design fuses high-level semantic and mid-level detail features, which are processed through global average pooling and fully connected layers to generate adaptive weights for weighted fusion, thereby strengthening feature learning and reducing the impact of noise on classification. The method is evaluated on a subset of the ISIC 2020 dataset containing 3297 benign and malignant skin lesion images. Experimental results show that the proposed ASFF-based ResNet-50 achieves the best overall performance compared with 5 classic convolutional neural networks (CNNs) models. The proposed model reached an accuracy of 93.18% along with higher precision, recall, specificity, and F1 score. The improved model achieves an AUC value of 0.9670 and 0.9717 in the P-R and ROC curve, respectively. Then, the evaluation based on Grad-CAM further proved that the improved model adaptively focuses on lesion-relevant regions while suppressing irrelevant background information, thereby validating its enhanced feature learning capability from a deep representation perspective. These findings demonstrate that the proposed approach provides a more effective and efficient solution for computer-aided skin cancer diagnosis.

Multi-Modal Oral Cancer Detection Using Weighted Ensemble Convolutional Neural Networks

arXiv:2510.03878v1 Announce Type: new Abstract: Aims Late diagnosis of Oral Squamous Cell Carcinoma (OSCC) contributes significantly to its high global mortality rate, with over 50\% of cases detected at advanced stages and a 5-year survival rate below 50\% according to WHO statistics. This study aims to improve early detection of OSCC by developing a multimodal deep learning framework that integrates clinical, radiological, and histopathological images using a weighted ensemble of DenseNet-121 convolutional neural networks (CNNs). Material and Methods A retrospective study was conducted using publicly available datasets representing three distinct medical imaging modalities. Each modality-specific dataset was used to train a DenseNet-121 CNN via transfer learning. Augmentation and modality-specific preprocessing were applied to increase robustness. Predictions were fused using a validation-weighted ensemble strategy. Evaluation was performed using accuracy, precision, recall, F1-score. Results High validation accuracy was achieved for radiological (100\%) and histopathological (95.12\%) modalities, with clinical images performing lower (63.10\%) due to visual heterogeneity. The ensemble model demonstrated improved diagnostic robustness with an overall accuracy of 84.58\% on a multimodal validation dataset of 55 samples. Conclusion The multimodal ensemble framework bridges gaps in the current diagnostic workflow by offering a non-invasive, AI-assisted triage tool that enhances early identification of high-risk lesions. It supports clinicians in decision-making, aligning with global oncology guidelines to reduce diagnostic delays and improve patient outcomes.

Exploring Instruction Data Quality for Explainable Image Quality Assessment

arXiv:2510.03880v1 Announce Type: new Abstract: In recent years, with the rapid development of powerful multimodal large language models (MLLMs), explainable image quality assessment (IQA) has gradually become popular, aiming at providing quality-related descriptions and answers of images. To achieve this goal, recent methods seek to construct a large-scale instruction tuning dataset to empower the MLLM with quality perception ability following the well-known scaling law. However, a large amount of instruction tuning data may cause substantial computational costs and redundant data, which in turn will cause harm to the performance of the model. To cope with this problem, in this paper, we challenge the scaling law and systematically investigate the role of data quality of the instruction tuning dataset for explainable IQA. Using a powerful pre-trained MLLM, we first investigate the changes in model performance after fine-tuning with different sizes of instruction tuning data. We find that selecting a subset of the data set randomly using an appropriate ratio can even lead to better results than training with the entire instruction tuning dataset, demonstrating the redundancy of current explainable IQA instruction tuning data. Beyond randomly sampling a subset, we propose a clustering-based data selection framework with three stages: clustering feature extraction, cluster quota allocation, and cluster sampling strategy. Then we systematically analyze the choices of each stage and propose a simple but efficient data selection method IQA-Select for explainable IQA. The experimental results demonstrate that IQA-Select can achieve 102.1% and 103.7% performance of full fine-tuning using only 10% selected data in Q-Bench and AesBench respectively, significantly reducing computational costs while achieving better performance.

Bridge Thinking and Acting: Unleashing Physical Potential of VLM with Generalizable Action Expert

arXiv:2510.03896v1 Announce Type: new Abstract: Although Vision-Language Models (VLM) have demonstrated impressive planning and reasoning capabilities, translating these abilities into the physical world introduces significant challenges. Conventional Vision-Language-Action (VLA) models, which integrate reasoning and action into a monolithic architecture, generalize poorly because they are constrained by scarce, narrow-domain data. While recent dual-system approaches attempt to decouple "thinking" from "acting", they are often constrained by semantic ambiguities within the action module. This ambiguity makes large-scale, cross-task training infeasible. Consequently, these systems typically necessitate fine-tuning on newly collected data when deployed to novel environments, and the cooperation mechanism between the two systems remains ill-defined. To address these limitations, we introduce, for the first time, a framework centered around a generalizable action expert. Our approach utilizes sparse 3D trajectories as an intermediate representation, effectively bridging the high-level planning capabilities of the VLM with the low-level physical action module. During the planning phase, the VLM is only required to generate coarse 3D waypoints. These waypoints are then processed by our generalizable action expert, which refines them into dense, executable action sequences by sampling real-time point cloud observations of the environment. To promote training efficiency and robust generalization, we introduce a novel "Action Pre-training, Pointcloud Fine-tuning" paradigm. Our method combines the broad generalization capabilities of VLMs in visual understanding and planning with the fine-grained, action-level generalization of action expert.

Zero-Shot Fine-Grained Image Classification Using Large Vision-Language Models

arXiv:2510.03903v1 Announce Type: new Abstract: Large Vision-Language Models (LVLMs) have demonstrated impressive performance on vision-language reasoning tasks. However, their potential for zero-shot fine-grained image classification, a challenging task requiring precise differentiation between visually similar categories, remains underexplored. We present a novel method that transforms zero-shot fine-grained image classification into a visual question-answering framework, leveraging LVLMs' comprehensive understanding capabilities rather than relying on direct class name generation. We enhance model performance through a novel attention intervention technique. We also address a key limitation in existing datasets by developing more comprehensive and precise class description benchmarks. We validate the effectiveness of our method through extensive experimentation across multiple fine-grained image classification benchmarks. Our proposed method consistently outperforms the current state-of-the-art (SOTA) approach, demonstrating both the effectiveness of our method and the broader potential of LVLMs for zero-shot fine-grained classification tasks. Code and Datasets: https://github.com/Atabuzzaman/Fine-grained-classification

From Filters to VLMs: Benchmarking Defogging Methods through Object Detection and Segmentation Performance

arXiv:2510.03906v1 Announce Type: new Abstract: Autonomous driving perception systems are particularly vulnerable in foggy conditions, where light scattering reduces contrast and obscures fine details critical for safe operation. While numerous defogging methods exist-from handcrafted filters to learned restoration models-improvements in image fidelity do not consistently translate into better downstream detection and segmentation. Moreover, prior evaluations often rely on synthetic data, leaving questions about real-world transferability. We present a structured empirical study that benchmarks a comprehensive set of pipelines, including (i) classical filters, (ii) modern defogging networks, (iii) chained variants (filter$\rightarrow$model, model$\rightarrow$filter), and (iv) prompt-driven visual--language image editing models (VLM) applied directly to foggy images. Using Foggy Cityscapes, we assess both image quality and downstream performance on object detection (mAP) and segmentation (PQ, RQ, SQ). Our analysis reveals when defogging helps, when chaining yields synergy or degradation, and how VLM-based editors compare to dedicated approaches. In addition, we evaluate qualitative rubric-based scores from a VLM judge and quantify their alignment with task metrics, showing strong correlations with mAP. Together, these results establish a transparent, task-oriented benchmark for defogging methods and highlight the conditions under which preprocessing genuinely improves autonomous perception in adverse weather.

Generating Human Motion Videos using a Cascaded Text-to-Video Framework

arXiv:2510.03909v1 Announce Type: new Abstract: Human video generation is becoming an increasingly important task with broad applications in graphics, entertainment, and embodied AI. Despite the rapid progress of video diffusion models (VDMs), their use for general-purpose human video generation remains underexplored, with most works constrained to image-to-video setups or narrow domains like dance videos. In this work, we propose CAMEO, a cascaded framework for general human motion video generation. It seamlessly bridges Text-to-Motion (T2M) models and conditional VDMs, mitigating suboptimal factors that may arise in this process across both training and inference through carefully designed components. Specifically, we analyze and prepare both textual prompts and visual conditions to effectively train the VDM, ensuring robust alignment between motion descriptions, conditioning signals, and the generated videos. Furthermore, we introduce a camera-aware conditioning module that connects the two stages, automatically selecting viewpoints aligned with the input text to enhance coherence and reduce manual intervention. We demonstrate the effectiveness of our approach on both the MovieGen benchmark and a newly introduced benchmark tailored to the T2M-VDM combination, while highlighting its versatility across diverse use cases.

OpenFLAME: Federated Visual Positioning System to Enable Large-Scale Augmented Reality Applications

arXiv:2510.03915v1 Announce Type: new Abstract: World-scale augmented reality (AR) applications need a ubiquitous 6DoF localization backend to anchor content to the real world consistently across devices. Large organizations such as Google and Niantic are 3D scanning outdoor public spaces in order to build their own Visual Positioning Systems (VPS). These centralized VPS solutions fail to meet the needs of many future AR applications -- they do not cover private indoor spaces because of privacy concerns, regulations, and the labor bottleneck of updating and maintaining 3D scans. In this paper, we present OpenFLAME, a federated VPS backend that allows independent organizations to 3D scan and maintain a separate VPS service for their own spaces. This enables access control of indoor 3D scans, distributed maintenance of the VPS backend, and encourages larger coverage. Sharding of VPS services introduces several unique challenges -- coherency of localization results across spaces, quality control of VPS services, selection of the right VPS service for a location, and many others. We introduce the concept of federated image-based localization and provide reference solutions for managing and merging data across maps without sharing private data.

Talking Tennis: Language Feedback from 3D Biomechanical Action Recognition

arXiv:2510.03921v1 Announce Type: new Abstract: Automated tennis stroke analysis has advanced significantly with the integration of biomechanical motion cues alongside deep learning techniques, enhancing stroke classification accuracy and player performance evaluation. Despite these advancements, existing systems often fail to connect biomechanical insights with actionable language feedback that is both accessible and meaningful to players and coaches. This research project addresses this gap by developing a novel framework that extracts key biomechanical features (such as joint angles, limb velocities, and kinetic chain patterns) from motion data using Convolutional Neural Network Long Short-Term Memory (CNN-LSTM)-based models. These features are analyzed for relationships influencing stroke effectiveness and injury risk, forming the basis for feedback generation using large language models (LLMs). Leveraging the THETIS dataset and feature extraction techniques, our approach aims to produce feedback that is technically accurate, biomechanically grounded, and actionable for end-users. The experimental setup evaluates this framework on classification performance and interpretability, bridging the gap between explainable AI and sports biomechanics.

Harnessing Synthetic Preference Data for Enhancing Temporal Understanding of Video-LLMs

arXiv:2510.03955v1 Announce Type: new Abstract: While Video Large Language Models (Video-LLMs) have demonstrated remarkable performance across general video understanding benchmarks-particularly in video captioning and descriptive tasks-they consistently underperform on tasks that require fine-grained temporal understanding. This limitation arises due to the lack of visual complexity and temporal nuance in current fine-tuning datasets, leading these models to rely heavily on language-based reasoning rather than truly understanding video dynamics. In this work, we propose TimeWarp, a systematic method to create a targeted synthetic temporal dataset to fine-tune the model's responses to encourage it to focus on the given input video. We introduce a large-scale preference dataset, created using TimeWarp, that captures intricate temporal dynamics often overlooked, grounding the model's responses to visual and temporal information. We demonstrate that when our method is applied to existing models, it significantly improves performance on temporal understanding benchmarks, highlighting the effectiveness of our proposed datasets in advancing temporal understanding in Video-LLMs, resulting in an absolute improvement in performance across seven benchmarks. Code is available at https://github.com/sameepv21/timewarp.

No Tokens Wasted: Leveraging Long Context in Biomedical Vision-Language Models

arXiv:2510.03978v1 Announce Type: new Abstract: Embedding vision-language models (VLMs) are typically pretrained with short text windows (<77 tokens), which forces the truncation of long-format captions. Yet, the distribution of biomedical captions from large-scale open source literature reveals that a huge portion of captions far exceed 77 tokens. To this end, we investigate the impact of pretraining on long-format biomedical captions by extending the context length of text encoders in VLMs. We find that longer context (thus, enabling additional supervision provided in long-format captions) correlates with better retrieval and classification performance. Given this finding, we introduce BIOMEDICA-LongCAP, a dataset of 1M image-caption pairs enriched with context-aware descriptions from full-text articles, providing longer and additional textual supervision. Using BIOMEDICA-LongCAP, we train BMC-LongCLIP, a long-context biomedical VLM with a text encoder supporting windows of up to 512 tokens. Our model extends context capacity by 6.6x, reducing token waste from 55% to just 2.2%. On long-caption retrieval benchmarks, BMC-LongCLIP achieves up to +30% absolute gains in Recall@1 and +2% average improvements in classification, while also converging faster than short-context. Our results demonstrate that long-context modeling is a promising direction for advancing biomedical VLMs.

Keep It on a Leash: Controllable Pseudo-label Generation Towards Realistic Long-Tailed Semi-Supervised Learning

arXiv:2510.03993v1 Announce Type: new Abstract: Current long-tailed semi-supervised learning methods assume that labeled data exhibit a long-tailed distribution, and unlabeled data adhere to a typical predefined distribution (i.e., long-tailed, uniform, or inverse long-tailed). However, the distribution of the unlabeled data is generally unknown and may follow an arbitrary distribution. To tackle this challenge, we propose a Controllable Pseudo-label Generation (CPG) framework, expanding the labeled dataset with the progressively identified reliable pseudo-labels from the unlabeled dataset and training the model on the updated labeled dataset with a known distribution, making it unaffected by the unlabeled data distribution. Specifically, CPG operates through a controllable self-reinforcing optimization cycle: (i) at each training step, our dynamic controllable filtering mechanism selectively incorporates reliable pseudo-labels from the unlabeled dataset into the labeled dataset, ensuring that the updated labeled dataset follows a known distribution; (ii) we then construct a Bayes-optimal classifier using logit adjustment based on the updated labeled data distribution; (iii) this improved classifier subsequently helps identify more reliable pseudo-labels in the next training step. We further theoretically prove that this optimization cycle can significantly reduce the generalization error under some conditions. Additionally, we propose a class-aware adaptive augmentation module to further improve the representation of minority classes, and an auxiliary branch to maximize data utilization by leveraging all labeled and unlabeled samples. Comprehensive evaluations on various commonly used benchmark datasets show that CPG achieves consistent improvements, surpassing state-of-the-art methods by up to \textbf{15.97\%} in accuracy. The code is available at https://github.com/yaxinhou/CPG.

Enhancing OCR for Sino-Vietnamese Language Processing via Fine-tuned PaddleOCRv5

arXiv:2510.04003v1 Announce Type: new Abstract: Recognizing and processing Classical Chinese (Han-Nom) texts play a vital role in digitizing Vietnamese historical documents and enabling cross-lingual semantic research. However, existing OCR systems struggle with degraded scans, non-standard glyphs, and handwriting variations common in ancient sources. In this work, we propose a fine-tuning approach for PaddleOCRv5 to improve character recognition on Han-Nom texts. We retrain the text recognition module using a curated subset of ancient Vietnamese Chinese manuscripts, supported by a full training pipeline covering preprocessing, LMDB conversion, evaluation, and visualization. Experimental results show a significant improvement over the base model, with exact accuracy increasing from 37.5 percent to 50.0 percent, particularly under noisy image conditions. Furthermore, we develop an interactive demo that visually compares pre- and post-fine-tuning recognition results, facilitating downstream applications such as Han-Vietnamese semantic alignment, machine translation, and historical linguistics research. The demo is available at https://huggingface.co/spaces/MinhDS/Fine-tuned-PaddleOCRv5.

Fit Pixels, Get Labels: Meta-learned Implicit Networks for Image Segmentation

arXiv:2510.04021v1 Announce Type: new Abstract: Implicit neural representations (INRs) have achieved remarkable successes in learning expressive yet compact signal representations. However, they are not naturally amenable to predictive tasks such as segmentation, where they must learn semantic structures over a distribution of signals. In this study, we introduce MetaSeg, a meta-learning framework to train INRs for medical image segmentation. MetaSeg uses an underlying INR that simultaneously predicts per pixel intensity values and class labels. It then uses a meta-learning procedure to find optimal initial parameters for this INR over a training dataset of images and segmentation maps, such that the INR can simply be fine-tuned to fit pixels of an unseen test image, and automatically decode its class labels. We evaluated MetaSeg on 2D and 3D brain MRI segmentation tasks and report Dice scores comparable to commonly used U-Net models, but with $90\%$ fewer parameters. MetaSeg offers a fresh, scalable alternative to traditional resource-heavy architectures such as U-Nets and vision transformers for medical image segmentation. Our project is available at https://kushalvyas.github.io/metaseg.html .

Video-in-the-Loop: Span-Grounded Long Video QA with Interleaved Reasoning

arXiv:2510.04022v1 Announce Type: new Abstract: We present \emph{Video-in-the-Loop} (ViTL), a two-stage long-video QA framework that preserves a fixed token budget by first \emph{localizing} question-relevant interval(s) with a low-fps skim and then \emph{answering} via span-aware reallocation of visual tokens at higher effective frame rate, emitting an interleaved output with both spans and the final option for direct attribution. We also introduce \dataname{}, which converts description based event graphs into \emph{span-grounded} multiple-choice QA by pairing each question with \emph{ground-truth} time span(s) and related reasoning. ViTL is trained end-to-end with an interleaved group-relative objective that couples temporal IoU for localization with answer correctness, allowing credit to flow from answers back to spans without increasing compute. Under fixed token budgets, ViTL attains up to 8.6% with 50% less frame input on long-video QA and temporal grounding (e.g., Charades-STA, ActivityNet-Captions) and ablations show that span-aware token reallocation consistently surpasses uniform sampling. Together, \dataname{} and ViTL provide an interpretable, compute-efficient recipe for scalable long-video QA.

Enhancing Fake News Video Detection via LLM-Driven Creative Process Simulation

arXiv:2510.04024v1 Announce Type: new Abstract: The emergence of fake news on short video platforms has become a new significant societal concern, necessitating automatic video-news-specific detection. Current detectors primarily rely on pattern-based features to separate fake news videos from real ones. However, limited and less diversified training data lead to biased patterns and hinder their performance. This weakness stems from the complex many-to-many relationships between video material segments and fabricated news events in real-world scenarios: a single video clip can be utilized in multiple ways to create different fake narratives, while a single fabricated event often combines multiple distinct video segments. However, existing datasets do not adequately reflect such relationships due to the difficulty of collecting and annotating large-scale real-world data, resulting in sparse coverage and non-comprehensive learning of the characteristics of potential fake news video creation. To address this issue, we propose a data augmentation framework, AgentAug, that generates diverse fake news videos by simulating typical creative processes. AgentAug implements multiple LLM-driven pipelines of four fabrication categories for news video creation, combined with an active learning strategy based on uncertainty sampling to select the potentially useful augmented samples during training. Experimental results on two benchmark datasets demonstrate that AgentAug consistently improves the performance of short video fake news detectors.

Prompt-to-Prompt: Text-Based Image Editing Via Cross-Attention Mechanisms -- The Research of Hyperparameters and Novel Mechanisms to Enhance Existing Frameworks

arXiv:2510.04034v1 Announce Type: new Abstract: Recent advances in image editing have shifted from manual pixel manipulation to employing deep learning methods like stable diffusion models, which now leverage cross-attention mechanisms for text-driven control. This transition has simplified the editing process but also introduced variability in results, such as inconsistent hair color changes. Our research aims to enhance the precision and reliability of prompt-to-prompt image editing frameworks by exploring and optimizing hyperparameters. We present a comprehensive study of the "word swap" method, develop an "attention re-weight method" for better adaptability, and propose the "CL P2P" framework to address existing limitations like cycle inconsistency. This work contributes to understanding and improving the interaction between hyperparameter settings and the architectural choices of neural network models, specifically their attention mechanisms, which significantly influence the composition and quality of the generated images.

\textsc{GUI-Spotlight}: Adaptive Iterative Focus Refinement for Enhanced GUI Visual Grounding

arXiv:2510.04039v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) have markedly expanded the competence of graphical user-interface (GUI) systems, propelling them beyond controlled simulations into complex, real-world environments across diverse platforms. However, practical usefulness is still bounded by the reliability of visual grounding, i.e., mapping textual references to exact on-screen elements. This limitation prevents the system from accurately performing pointer-level actions such as clicking or dragging. To address it, we introduce GUI-Spotlight -- a model trained for image-grounded reasoning that dynamically invokes multiple specialized tools to iteratively narrow its focus to the relevant region of the screen, thereby substantially improving visual grounding accuracy. On the ScreenSpot-Pro benchmark, GUI-Spotlight trained with only 18.5K training samples achieves 52.8\% accuracy, surpassing V2P-7B (50.6\% with 9.6M training samples) and GTA-1-7B (50.1\% with 1.56M training samples).

Quantization Range Estimation for Convolutional Neural Networks

arXiv:2510.04044v1 Announce Type: new Abstract: Post-training quantization for reducing the storage of deep neural network models has been demonstrated to be an effective way in various tasks. However, low-bit quantization while maintaining model accuracy is a challenging problem. In this paper, we present a range estimation method to improve the quantization performance for post-training quantization. We model the range estimation into an optimization problem of minimizing quantization errors by layer-wise local minima. We prove this problem is locally convex and present an efficient search algorithm to find the optimal solution. We propose the application of the above search algorithm to the transformed weights space to do further improvement in practice. Our experiments demonstrate that our method outperforms state-of-the-art performance generally on top-1 accuracy for image classification tasks on the ResNet series models and Inception-v3 model. The experimental results show that the proposed method has almost no loss of top-1 accuracy in 8-bit and 6-bit settings for image classifications, and the accuracy of 4-bit quantization is also significantly improved. The code is available at https://github.com/codeiscommitting/REQuant.

MetaFind: Scene-Aware 3D Asset Retrieval for Coherent Metaverse Scene Generation

arXiv:2510.04057v1 Announce Type: new Abstract: We present MetaFind, a scene-aware tri-modal compositional retrieval framework designed to enhance scene generation in the metaverse by retrieving 3D assets from large-scale repositories. MetaFind addresses two core challenges: (i) inconsistent asset retrieval that overlooks spatial, semantic, and stylistic constraints, and (ii) the absence of a standardized retrieval paradigm specifically tailored for 3D asset retrieval, as existing approaches mainly rely on general-purpose 3D shape representation models. Our key innovation is a flexible retrieval mechanism that supports arbitrary combinations of text, image, and 3D modalities as queries, enhancing spatial reasoning and style consistency by jointly modeling object-level features (including appearance) and scene-level layout structures. Methodologically, MetaFind introduces a plug-and-play equivariant layout encoder ESSGNN that captures spatial relationships and object appearance features, ensuring retrieved 3D assets are contextually and stylistically coherent with the existing scene, regardless of coordinate frame transformations. The framework supports iterative scene construction by continuously adapting retrieval results to current scene updates. Empirical evaluations demonstrate the improved spatial and stylistic consistency of MetaFind in various retrieval tasks compared to baseline methods.

Ordinal Encoding as a Regularizer in Binary Loss for Solar Flare Prediction

arXiv:2510.04063v1 Announce Type: new Abstract: The prediction of solar flares is typically formulated as a binary classification task, distinguishing events as either Flare (FL) or No-Flare (NF) according to a specified threshold (for example, greater than or equal to C-class, M-class, or X-class). However, this binary framework neglects the inherent ordinal relationships among the sub-classes contained within each category (FL and NF). Several studies on solar flare prediction have empirically shown that the most frequent misclassifications occur near this prediction threshold. This suggests that the models struggle to differentiate events that are similar in intensity but fall on opposite sides of the binary threshold. To mitigate this limitation, we propose a modified loss function that integrates the ordinal information among the sub-classes of the binarized flare labels into the conventional binary cross-entropy (BCE) loss. This approach serves as an ordinality-aware, data-driven regularization method that penalizes the incorrect predictions of flare events in close proximity to the prediction threshold more heavily than those away from the boundary during model optimization. By incorporating ordinal weighting into the loss function, we aim to enhance the model's learning process by leveraging the ordinal characteristics of the data, thereby improving its overall performance.

QuantDemoire: Quantization with Outlier Aware for Image Demoir\'eing

arXiv:2510.04066v1 Announce Type: new Abstract: Demoir\'eing aims to remove moir\'e artifacts that often occur in images. While recent deep learning-based methods have achieved promising results, they typically require substantial computational resources, limiting their deployment on edge devices. Model quantization offers a compelling solution. However, directly applying existing quantization methods to demoir\'eing models introduces severe performance degradation. The main reasons are distribution outliers and weakened representations in smooth regions. To address these issues, we propose QuantDemoire, a post-training quantization framework tailored to demoir\'eing. It contains two key components. First}, we introduce an outlier-aware quantizer to reduce errors from outliers. It uses sampling-based range estimation to reduce activation outliers, and keeps a few extreme weights in FP16 with negligible cost. Second, we design a frequency-aware calibration strategy. It emphasizes low- and mid-frequency components during fine-tuning, which mitigates banding artifacts caused by low-bit quantization. Extensive experiments validate that our QuantDemoire achieves large reductions in parameters and computation while maintaining quality. Meanwhile, it outperforms existing quantization methods by over 4 dB** on W4A4. Code is released at: https://github.com/zhengchen1999/QuantDemoire.

Diffusion Low Rank Hybrid Reconstruction for Sparse View Medical Imaging

arXiv:2510.04069v1 Announce Type: new Abstract: This work presents TV-LoRA, a novel method for low-dose sparse-view CT reconstruction that combines a diffusion generative prior (NCSN++ with SDE modeling) and multi-regularization constraints, including anisotropic TV and nuclear norm (LoRA), within an ADMM framework. To address ill-posedness and texture loss under extremely sparse views, TV-LoRA integrates generative and physical constraints, and utilizes a 2D slice-based strategy with FFT acceleration and tensor-parallel optimization for efficient inference. Experiments on AAPM-2016, CTHD, and LIDC datasets with $N_{\mathrm{view}}=8,4,2$ show that TV-LoRA consistently surpasses benchmarks in SSIM, texture recovery, edge clarity, and artifact suppression, demonstrating strong robustness and generalizability. Ablation studies confirm the complementary effects of LoRA regularization and diffusion priors, while the FFT-PCG module provides a speedup. Overall, Diffusion + TV-LoRA achieves high-fidelity, efficient 3D CT reconstruction and broad clinical applicability in low-dose, sparse-sampling scenarios.

TOPO-Bench: An Open-Source Topological Mapping Evaluation Framework with Quantifiable Perceptual Aliasing

arXiv:2510.04100v1 Announce Type: new Abstract: Topological mapping offers a compact and robust representation for navigation, but progress in the field is hindered by the lack of standardized evaluation metrics, datasets, and protocols. Existing systems are assessed using different environments and criteria, preventing fair and reproducible comparisons. Moreover, a key challenge - perceptual aliasing - remains under-quantified, despite its strong influence on system performance. We address these gaps by (1) formalizing topological consistency as the fundamental property of topological maps and showing that localization accuracy provides an efficient and interpretable surrogate metric, and (2) proposing the first quantitative measure of dataset ambiguity to enable fair comparisons across environments. To support this protocol, we curate a diverse benchmark dataset with calibrated ambiguity levels, implement and release deep-learned baseline systems, and evaluate them alongside classical methods. Our experiments and analysis yield new insights into the limitations of current approaches under perceptual aliasing. All datasets, baselines, and evaluation tools are fully open-sourced to foster consistent and reproducible research in topological mapping.

Learning Efficient Meshflow and Optical Flow from Event Cameras

arXiv:2510.04111v1 Announce Type: new Abstract: In this paper, we explore the problem of event-based meshflow estimation, a novel task that involves predicting a spatially smooth sparse motion field from event cameras. To start, we review the state-of-the-art in event-based flow estimation, highlighting two key areas for further research: i) the lack of meshflow-specific event datasets and methods, and ii) the underexplored challenge of event data density. First, we generate a large-scale High-Resolution Event Meshflow (HREM) dataset, which showcases its superiority by encompassing the merits of high resolution at 1280x720, handling dynamic objects and complex motion patterns, and offering both optical flow and meshflow labels. These aspects have not been fully explored in previous works. Besides, we propose Efficient Event-based MeshFlow (EEMFlow) network, a lightweight model featuring a specially crafted encoder-decoder architecture to facilitate swift and accurate meshflow estimation. Furthermore, we upgrade EEMFlow network to support dense event optical flow, in which a Confidence-induced Detail Completion (CDC) module is proposed to preserve sharp motion boundaries. We conduct comprehensive experiments to show the exceptional performance and runtime efficiency (30x faster) of our EEMFlow model compared to the recent state-of-the-art flow method. As an extension, we expand HREM into HREM+, a multi-density event dataset contributing to a thorough study of the robustness of existing methods across data with varying densities, and propose an Adaptive Density Module (ADM) to adjust the density of input event data to a more optimal range, enhancing the model's generalization ability. We empirically demonstrate that ADM helps to significantly improve the performance of EEMFlow and EEMFlow+ by 8% and 10%, respectively. Code and dataset are released at https://github.com/boomluo02/EEMFlowPlus.

Joint Learning of Pose Regression and Denoising Diffusion with Score Scaling Sampling for Category-level 6D Pose Estimation

arXiv:2510.04125v1 Announce Type: new Abstract: Latest diffusion models have shown promising results in category-level 6D object pose estimation by modeling the conditional pose distribution with depth image input. The existing methods, however, suffer from slow convergence during training, learning its encoder with the diffusion denoising network in end-to-end fashion, and require an additional network that evaluates sampled pose hypotheses to filter out low-quality pose candidates. In this paper, we propose a novel pipeline that tackles these limitations by two key components. First, the proposed method pretrains the encoder with the direct pose regression head, and jointly learns the networks via the regression head and the denoising diffusion head, significantly accelerating training convergence while achieving higher accuracy. Second, sampling guidance via time-dependent score scaling is proposed s.t. the exploration-exploitation trade-off is effectively taken, eliminating the need for the additional evaluation network. The sampling guidance maintains multi-modal characteristics of symmetric objects at early denoising steps while ensuring high-quality pose generation at final steps. Extensive experiments on multiple benchmarks including REAL275, HouseCat6D, and ROPE, demonstrate that the proposed method, simple yet effective, achieves state-of-the-art accuracies even with single-pose inference, while being more efficient in both training and inference.

Learning from All: Concept Alignment for Autonomous Distillation from Multiple Drifting MLLMs

arXiv:2510.04142v1 Announce Type: new Abstract: This paper identifies a critical yet underexplored challenge in distilling from multimodal large language models (MLLMs): the reasoning trajectories generated by multiple drifting teachers exhibit concept drift, whereby their reasoning distributions evolve unpredictably and transmit biases to the student model, ultimately compromising its performance. To tackle this issue, we pioneer a theoretical connection between concept drift and knowledge distillation, casting the non-stationary reasoning dynamics from multiple MLLM teachers as next-token prediction of multi-stream reasoning trajectories.Guided by concept drift, we introduce the "learn, compare, critique" paradigm, culminating in autonomous preference optimization (APO). Under the active guidance of the teachers, the student model first learns and self-distils preferred thinking by comparing multiple teachers. It then engages in critical reflection over the drifting inference from teachers, performing concept alignment through APO, ultimately yielding a robust, consistent, and generalizable model.Extensive experiments demonstrate our superior performance of consistency, robustness and generalization within knowledge distillation. Besides, we also contributed a large-scale dataset, CXR-MAX (Multi-teachers Alignment X-rays), comprising 170,982 distilled reasoning trajectories derived from publicly accessible MLLMs based on MIMIC-CXR. Our code and data are public at: https://anonymous.4open.science/r/Autonomous-Distillation/.

Automating construction safety inspections using a multi-modal vision-language RAG framework

arXiv:2510.04145v1 Announce Type: new Abstract: Conventional construction safety inspection methods are often inefficient as they require navigating through large volume of information. Recent advances in large vision-language models (LVLMs) provide opportunities to automate safety inspections through enhanced visual and linguistic understanding. However, existing applications face limitations including irrelevant or unspecific responses, restricted modal inputs and hallucinations. Utilisation of Large Language Models (LLMs) for this purpose is constrained by availability of training data and frequently lack real-time adaptability. This study introduces SiteShield, a multi-modal LVLM-based Retrieval-Augmented Generation (RAG) framework for automating construction safety inspection reports by integrating visual and audio inputs. Using real-world data, SiteShield outperformed unimodal LLMs without RAG with an F1 score of 0.82, hamming loss of 0.04, precision of 0.76, and recall of 0.96. The findings indicate that SiteShield offers a novel pathway to enhance information retrieval and efficiency in generating safety reports.

BLADE: Bias-Linked Adaptive DEbiasing

arXiv:2510.04174v1 Announce Type: new Abstract: Neural networks have revolutionized numerous fields, yet they remain vulnerable to a critical flaw: the tendency to learn implicit biases, spurious correlations between certain attributes and target labels in training data. These biases are often more prevalent and easier to learn, causing models to rely on superficial patterns rather than task-relevant features necessary for generalization. Existing methods typically rely on strong assumptions, such as prior knowledge of these biases or access to bias-conflicting samples, i.e., samples that contradict spurious correlations and counterbalance bias-aligned samples, samples that conform to these spurious correlations. However, such assumptions are often impractical in real-world settings. We propose BLADE ({B}ias-{L}inked {A}daptive {DE}biasing), a generative debiasing framework that requires no prior knowledge of bias or bias-conflicting samples. BLADE first trains a generative model to translate images across bias domains while preserving task-relevant features. Then, it adaptively refines each image with its synthetic counterpart based on the image's susceptibility to bias. To encourage robust representations, BLADE aligns an image with its bias-translated synthetic counterpart that shares task-relevant features but differs in bias, while misaligning it with samples sharing the same bias. We evaluate BLADE on multiple benchmark datasets and show that it significantly outperforms state-of-the-art methods. Notably, it exceeds the closest baseline by an absolute margin of around 18% on the corrupted CIFAR-10 dataset under the worst group setting, establishing a new benchmark in bias mitigation and demonstrating its potential for developing more robust deep learning models without explicit supervision.

From Segments to Concepts: Interpretable Image Classification via Concept-Guided Segmentation

arXiv:2510.04180v1 Announce Type: new Abstract: Deep neural networks have achieved remarkable success in computer vision; however, their black-box nature in decision-making limits interpretability and trust, particularly in safety-critical applications. Interpretability is crucial in domains where errors have severe consequences. Existing models not only lack transparency but also risk exploiting unreliable or misleading features, which undermines both robustness and the validity of their explanations. Concept Bottleneck Models (CBMs) aim to improve transparency by reasoning through human-interpretable concepts. Still, they require costly concept annotations and lack spatial grounding, often failing to identify which regions support each concept. We propose SEG-MIL-CBM, a novel framework that integrates concept-guided image segmentation into an attention-based multiple instance learning (MIL) framework, where each segmented region is treated as an instance and the model learns to aggregate evidence across them. By reasoning over semantically meaningful regions aligned with high-level concepts, our model highlights task-relevant evidence, down-weights irrelevant cues, and produces spatially grounded, concept-level explanations without requiring annotations of concepts or groups. SEG-MIL-CBM achieves robust performance across settings involving spurious correlations (unintended dependencies between background and label), input corruptions (perturbations that degrade visual quality), and large-scale benchmarks, while providing transparent, concept-level explanations.

Let Features Decide Their Own Solvers: Hybrid Feature Caching for Diffusion Transformers

arXiv:2510.04188v1 Announce Type: new Abstract: Diffusion Transformers offer state-of-the-art fidelity in image and video synthesis, but their iterative sampling process remains a major bottleneck due to the high cost of transformer forward passes at each timestep. To mitigate this, feature caching has emerged as a training-free acceleration technique that reuses or forecasts hidden representations. However, existing methods often apply a uniform caching strategy across all feature dimensions, ignoring their heterogeneous dynamic behaviors. Therefore, we adopt a new perspective by modeling hidden feature evolution as a mixture of ODEs across dimensions, and introduce HyCa, a Hybrid ODE solver inspired caching framework that applies dimension-wise caching strategies. HyCa achieves near-lossless acceleration across diverse domains and models, including 5.55 times speedup on FLUX, 5.56 times speedup on HunyuanVideo, 6.24 times speedup on Qwen-Image and Qwen-Image-Edit without retraining.

World-To-Image: Grounding Text-to-Image Generation with Agent-Driven World Knowledge

arXiv:2510.04201v1 Announce Type: new Abstract: While text-to-image (T2I) models can synthesize high-quality images, their performance degrades significantly when prompted with novel or out-of-distribution (OOD) entities due to inherent knowledge cutoffs. We introduce World-To-Image, a novel framework that bridges this gap by empowering T2I generation with agent-driven world knowledge. We design an agent that dynamically searches the web to retrieve images for concepts unknown to the base model. This information is then used to perform multimodal prompt optimization, steering powerful generative backbones toward an accurate synthesis. Critically, our evaluation goes beyond traditional metrics, utilizing modern assessments like LLMGrader and ImageReward to measure true semantic fidelity. Our experiments show that World-To-Image substantially outperforms state-of-the-art methods in both semantic alignment and visual aesthetics, achieving +8.1% improvement in accuracy-to-prompt on our curated NICE benchmark. Our framework achieves these results with high efficiency in less than three iterations, paving the way for T2I systems that can better reflect the ever-changing real world. Our demo code is available here\footnote{https://github.com/mhson-kyle/World-To-Image}.

MASC: Boosting Autoregressive Image Generation with a Manifold-Aligned Semantic Clustering

arXiv:2510.04220v1 Announce Type: new Abstract: Autoregressive (AR) models have shown great promise in image generation, yet they face a fundamental inefficiency stemming from their core component: a vast, unstructured vocabulary of visual tokens. This conventional approach treats tokens as a flat vocabulary, disregarding the intrinsic structure of the token embedding space where proximity often correlates with semantic similarity. This oversight results in a highly complex prediction task, which hinders training efficiency and limits final generation quality. To resolve this, we propose Manifold-Aligned Semantic Clustering (MASC), a principled framework that constructs a hierarchical semantic tree directly from the codebook's intrinsic structure. MASC employs a novel geometry-aware distance metric and a density-driven agglomerative construction to model the underlying manifold of the token embeddings. By transforming the flat, high-dimensional prediction task into a structured, hierarchical one, MASC introduces a beneficial inductive bias that significantly simplifies the learning problem for the AR model. MASC is designed as a plug-and-play module, and our extensive experiments validate its effectiveness: it accelerates training by up to 57% and significantly improves generation quality, reducing the FID of LlamaGen-XL from 2.87 to 2.58. MASC elevates existing AR frameworks to be highly competitive with state-of-the-art methods, establishing that structuring the prediction space is as crucial as architectural innovation for scalable generative modeling.

Zoom-In to Sort AI-Generated Images Out

arXiv:2510.04225v1 Announce Type: new Abstract: The rapid growth of AI-generated imagery has blurred the boundary between real and synthetic content, raising critical concerns for digital integrity. Vision-language models (VLMs) offer interpretability through explanations but often fail to detect subtle artifacts in high-quality synthetic images. We propose ZoomIn, a two-stage forensic framework that improves both accuracy and interpretability. Mimicking human visual inspection, ZoomIn first scans an image to locate suspicious regions and then performs a focused analysis on these zoomed-in areas to deliver a grounded verdict. To support training, we introduce MagniFake, a dataset of 20,000 real and high-quality synthetic images annotated with bounding boxes and forensic explanations, generated through an automated VLM-based pipeline. Our method achieves 96.39% accuracy with robust generalization, while providing human-understandable explanations grounded in visual evidence.

A Recursive Pyramidal Algorithm for Solving the Image Registration Problem

arXiv:2510.04231v1 Announce Type: new Abstract: The problem of image registration is finding a transformation that aligns two images, such that the corresponding points are in the same location. This paper introduces a simple, end-to-end trainable algorithm that is implementable in a few lines of Python code. The approach is shown to work with very little training data and training time, while achieving accurate results in some settings. An example application to stereo vision was trained from 74 images on a 19x15 input window. With just a dozen lines of Python code this algorithm excels in brevity and may serve as a good start in related scenarios with limitations to training data, training time or code complexity.

Detection of retinal diseases using an accelerated reused convolutional network

arXiv:2510.04232v1 Announce Type: new Abstract: Convolutional neural networks are continually evolving, with some efforts aimed at improving accuracy, others at increasing speed, and some at enhancing accessibility. Improving accessibility broadens the application of neural networks across a wider range of tasks, including the detection of eye diseases. Early diagnosis of eye diseases and consulting an ophthalmologist can prevent many vision disorders. Given the importance of this issue, various datasets have been collected from the cornea to facilitate the process of making neural network models. However, most of the methods introduced in the past are computationally complex. In this study, we tried to increase the accessibility of deep neural network models. We did this at the most fundamental level, specifically by redesigning and optimizing the convolutional layers. By doing so, we created a new general model that incorporates our novel convolutional layer named ArConv layers. Thanks to the efficient performance of this new layer, the model has suitable complexity for use in mobile phones and can perform the task of diagnosing the presence of disease with high accuracy. The final model we present contains only 1.3 million parameters. In comparison to the MobileNetV2 model, which has 2.2 million parameters, our model demonstrated better accuracy when trained and evaluated on the RfMiD dataset under identical conditions, achieving an accuracy of 0.9328 versus 0.9266 on the RfMiD test set.

Scaling Sequence-to-Sequence Generative Neural Rendering

arXiv:2510.04236v1 Announce Type: new Abstract: We present Kaleido, a family of generative models designed for photorealistic, unified object- and scene-level neural rendering. Kaleido operates on the principle that 3D can be regarded as a specialised sub-domain of video, expressed purely as a sequence-to-sequence image synthesis task. Through a systemic study of scaling sequence-to-sequence generative neural rendering, we introduce key architectural innovations that enable our model to: i) perform generative view synthesis without explicit 3D representations; ii) generate any number of 6-DoF target views conditioned on any number of reference views via a masked autoregressive framework; and iii) seamlessly unify 3D and video modelling within a single decoder-only rectified flow transformer. Within this unified framework, Kaleido leverages large-scale video data for pre-training, which significantly improves spatial consistency and reduces reliance on scarce, camera-labelled 3D datasets -- all without any architectural modifications. Kaleido sets a new state-of-the-art on a range of view synthesis benchmarks. Its zero-shot performance substantially outperforms other generative methods in few-view settings, and, for the first time, matches the quality of per-scene optimisation methods in many-view settings.

The best performance in the CARE 2025 -- Liver Task (LiSeg-Contrast): Contrast-Aware Semi-Supervised Segmentation with Domain Generalization and Test-Time Adaptation

arXiv:2510.04243v1 Announce Type: new Abstract: Accurate liver segmentation from contrast-enhanced MRI is essential for diagnosis, treatment planning, and disease monitoring. However, it remains challenging due to limited annotated data, heterogeneous enhancement protocols, and significant domain shifts across scanners and institutions. Traditional image-to-image translation frameworks have made great progress in domain generalization, but their application is not straightforward. For example, Pix2Pix requires image registration, and cycle-GAN cannot be integrated seamlessly into segmentation pipelines. Meanwhile, these methods are originally used to deal with cross-modality scenarios, and often introduce structural distortions and suffer from unstable training, which may pose drawbacks in our single-modality scenario. To address these challenges, we propose CoSSeg-TTA, a compact segmentation framework for the GED4 (Gd-EOB-DTPA enhanced hepatobiliary phase MRI) modality built upon nnU-Netv2 and enhanced with a semi-supervised mean teacher scheme to exploit large amounts of unlabeled volumes. A domain adaptation module, incorporating a randomized histogram-based style appearance transfer function and a trainable contrast-aware network, enriches domain diversity and mitigates cross-center variability. Furthermore, a continual test-time adaptation strategy is employed to improve robustness during inference. Extensive experiments demonstrate that our framework consistently outperforms the nnU-Netv2 baseline, achieving superior Dice score and Hausdorff Distance while exhibiting strong generalization to unseen domains under low-annotation conditions.

Concept-Based Masking: A Patch-Agnostic Defense Against Adversarial Patch Attacks

arXiv:2510.04245v1 Announce Type: new Abstract: Adversarial patch attacks pose a practical threat to deep learning models by forcing targeted misclassifications through localized perturbations, often realized in the physical world. Existing defenses typically assume prior knowledge of patch size or location, limiting their applicability. In this work, we propose a patch-agnostic defense that leverages concept-based explanations to identify and suppress the most influential concept activation vectors, thereby neutralizing patch effects without explicit detection. Evaluated on Imagenette with a ResNet-50, our method achieves higher robust and clean accuracy than the state-of-the-art PatchCleanser, while maintaining strong performance across varying patch sizes and locations. Our results highlight the promise of combining interpretability with robustness and suggest concept-driven defenses as a scalable strategy for securing machine learning models against adversarial patch attacks.

Flexible and Efficient Spatio-Temporal Transformer for Sequential Visual Place Recognition

arXiv:2510.04282v1 Announce Type: new Abstract: Sequential Visual Place Recognition (Seq-VPR) leverages transformers to capture spatio-temporal features effectively; however, existing approaches prioritize performance at the expense of flexibility and efficiency. In practice, a transformer-based Seq-VPR model should be flexible to the number of frames per sequence (seq-length), deliver fast inference, and have low memory usage to meet real-time constraints. To our knowledge, no existing transformer-based Seq-VPR method achieves both flexibility and efficiency. To address this gap, we propose Adapt-STformer, a Seq-VPR method built around our novel Recurrent Deformable Transformer Encoder (Recurrent-DTE), which uses an iterative recurrent mechanism to fuse information from multiple sequential frames. This design naturally supports variable seq-lengths, fast inference, and low memory usage. Experiments on the Nordland, Oxford, and NuScenes datasets show that Adapt-STformer boosts recall by up to 17% while reducing sequence extraction time by 36% and lowering memory usage by 35% compared to the second-best baseline.

ChronoEdit: Towards Temporal Reasoning for Image Editing and World Simulation

arXiv:2510.04290v1 Announce Type: new Abstract: Recent advances in large generative models have significantly advanced image editing and in-context image generation, yet a critical gap remains in ensuring physical consistency, where edited objects must remain coherent. This capability is especially vital for world simulation related tasks. In this paper, we present ChronoEdit, a framework that reframes image editing as a video generation problem. First, ChronoEdit treats the input and edited images as the first and last frames of a video, allowing it to leverage large pretrained video generative models that capture not only object appearance but also the implicit physics of motion and interaction through learned temporal consistency. Second, ChronoEdit introduces a temporal reasoning stage that explicitly performs editing at inference time. Under this setting, the target frame is jointly denoised with reasoning tokens to imagine a plausible editing trajectory that constrains the solution space to physically viable transformations. The reasoning tokens are then dropped after a few steps to avoid the high computational cost of rendering a full video. To validate ChronoEdit, we introduce PBench-Edit, a new benchmark of image-prompt pairs for contexts that require physical consistency, and demonstrate that ChronoEdit surpasses state-of-the-art baselines in both visual fidelity and physical plausibility. Code and models for both the 14B and 2B variants of ChronoEdit will be released on the project page: https://research.nvidia.com/labs/toronto-ai/chronoedit

CARE-PD: A Multi-Site Anonymized Clinical Dataset for Parkinson's Disease Gait Assessment

arXiv:2510.04312v1 Announce Type: new Abstract: Objective gait assessment in Parkinson's Disease (PD) is limited by the absence of large, diverse, and clinically annotated motion datasets. We introduce CARE-PD, the largest publicly available archive of 3D mesh gait data for PD, and the first multi-site collection spanning 9 cohorts from 8 clinical centers. All recordings (RGB video or motion capture) are converted into anonymized SMPL meshes via a harmonized preprocessing pipeline. CARE-PD supports two key benchmarks: supervised clinical score prediction (estimating Unified Parkinson's Disease Rating Scale, UPDRS, gait scores) and unsupervised motion pretext tasks (2D-to-3D keypoint lifting and full-body 3D reconstruction). Clinical prediction is evaluated under four generalization protocols: within-dataset, cross-dataset, leave-one-dataset-out, and multi-dataset in-domain adaptation. To assess clinical relevance, we compare state-of-the-art motion encoders with a traditional gait-feature baseline, finding that encoders consistently outperform handcrafted features. Pretraining on CARE-PD reduces MPJPE (from 60.8mm to 7.5mm) and boosts PD severity macro-F1 by 17 percentage points, underscoring the value of clinically curated, diverse training data. CARE-PD and all benchmark code are released for non-commercial research at https://neurips2025.care-pd.ca/.

GenAR: Next-Scale Autoregressive Generation for Spatial Gene Expression Prediction

arXiv:2510.04315v1 Announce Type: new Abstract: Spatial Transcriptomics (ST) offers spatially resolved gene expression but remains costly. Predicting expression directly from widely available Hematoxylin and Eosin (H&E) stained images presents a cost-effective alternative. However, most computational approaches (i) predict each gene independently, overlooking co-expression structure, and (ii) cast the task as continuous regression despite expression being discrete counts. This mismatch can yield biologically implausible outputs and complicate downstream analyses. We introduce GenAR, a multi-scale autoregressive framework that refines predictions from coarse to fine. GenAR clusters genes into hierarchical groups to expose cross-gene dependencies, models expression as codebook-free discrete token generation to directly predict raw counts, and conditions decoding on fused histological and spatial embeddings. From an information-theoretic perspective, the discrete formulation avoids log-induced biases and the coarse-to-fine factorization aligns with a principled conditional decomposition. Extensive experimental results on four Spatial Transcriptomics datasets across different tissue types demonstrate that GenAR achieves state-of-the-art performance, offering potential implications for precision medicine and cost-effective molecular profiling. Code is publicly available at https://github.com/oyjr/genar.

RAP: 3D Rasterization Augmented End-to-End Planning

arXiv:2510.04333v1 Announce Type: new Abstract: Imitation learning for end-to-end driving trains policies only on expert demonstrations. Once deployed in a closed loop, such policies lack recovery data: small mistakes cannot be corrected and quickly compound into failures. A promising direction is to generate alternative viewpoints and trajectories beyond the logged path. Prior work explores photorealistic digital twins via neural rendering or game engines, but these methods are prohibitively slow and costly, and thus mainly used for evaluation. In this work, we argue that photorealism is unnecessary for training end-to-end planners. What matters is semantic fidelity and scalability: driving depends on geometry and dynamics, not textures or lighting. Motivated by this, we propose 3D Rasterization, which replaces costly rendering with lightweight rasterization of annotated primitives, enabling augmentations such as counterfactual recovery maneuvers and cross-agent view synthesis. To transfer these synthetic views effectively to real-world deployment, we introduce a Raster-to-Real feature-space alignment that bridges the sim-to-real gap. Together, these components form Rasterization Augmented Planning (RAP), a scalable data augmentation pipeline for planning. RAP achieves state-of-the-art closed-loop robustness and long-tail generalization, ranking first on four major benchmarks: NAVSIM v1/v2, Waymo Open Dataset Vision-based E2E Driving, and Bench2Drive. Our results show that lightweight rasterization with feature alignment suffices to scale E2E training, offering a practical alternative to photorealistic rendering. Project page: https://alan-lanfeng.github.io/RAP/.

Diffusion^2: Dual Diffusion Model with Uncertainty-Aware Adaptive Noise for Momentary Trajectory Prediction

arXiv:2510.04365v1 Announce Type: new Abstract: Accurate pedestrian trajectory prediction is crucial for ensuring safety and efficiency in autonomous driving and human-robot interaction scenarios. Earlier studies primarily utilized sufficient observational data to predict future trajectories. However, in real-world scenarios, such as pedestrians suddenly emerging from blind spots, sufficient observational data is often unavailable (i.e. momentary trajectory), making accurate prediction challenging and increasing the risk of traffic accidents. Therefore, advancing research on pedestrian trajectory prediction under extreme scenarios is critical for enhancing traffic safety. In this work, we propose a novel framework termed Diffusion^2, tailored for momentary trajectory prediction. Diffusion^2 consists of two sequentially connected diffusion models: one for backward prediction, which generates unobserved historical trajectories, and the other for forward prediction, which forecasts future trajectories. Given that the generated unobserved historical trajectories may introduce additional noise, we propose a dual-head parameterization mechanism to estimate their aleatoric uncertainty and design a temporally adaptive noise module that dynamically modulates the noise scale in the forward diffusion process. Empirically, Diffusion^2 sets a new state-of-the-art in momentary trajectory prediction on ETH/UCY and Stanford Drone datasets.

MorphoSim: An Interactive, Controllable, and Editable Language-guided 4D World Simulator

arXiv:2510.04390v1 Announce Type: new Abstract: World models that support controllable and editable spatiotemporal environments are valuable for robotics, enabling scalable training data, repro ducible evaluation, and flexible task design. While recent text-to-video models generate realistic dynam ics, they are constrained to 2D views and offer limited interaction. We introduce MorphoSim, a language guided framework that generates 4D scenes with multi-view consistency and object-level controls. From natural language instructions, MorphoSim produces dynamic environments where objects can be directed, recolored, or removed, and scenes can be observed from arbitrary viewpoints. The framework integrates trajectory-guided generation with feature field dis tillation, allowing edits to be applied interactively without full re-generation. Experiments show that Mor phoSim maintains high scene fidelity while enabling controllability and editability. The code is available at https://github.com/eric-ai-lab/Morph4D.

Your Vision-Language Model Can't Even Count to 20: Exposing the Failures of VLMs in Compositional Counting

arXiv:2510.04401v1 Announce Type: new Abstract: Vision-Language Models (VLMs) have become a central focus of today's AI community, owing to their impressive abilities gained from training on large-scale vision-language data from the Web. These models have demonstrated strong performance across diverse tasks, including image understanding, video understanding, complex visual reasoning, and embodied AI. Despite these noteworthy successes, a fundamental question remains: Can VLMs count objects correctly? In this paper, we introduce a simple yet effective benchmark, VLMCountBench, designed under a minimalist setting with only basic geometric shapes (e.g., triangles, circles) and their compositions, focusing exclusively on counting tasks without interference from other factors. We adopt strict independent variable control and systematically study the effects of simple properties such as color, size, and prompt refinement in a controlled ablation. Our empirical results reveal that while VLMs can count reliably when only one shape type is present, they exhibit substantial failures when multiple shape types are combined (i.e., compositional counting). This highlights a fundamental empirical limitation of current VLMs and motivates important directions for future research.

CodeFormer++: Blind Face Restoration Using Deformable Registration and Deep Metric Learning

arXiv:2510.04410v1 Announce Type: new Abstract: Blind face restoration (BFR) has attracted increasing attention with the rise of generative methods. Most existing approaches integrate generative priors into the restoration pro- cess, aiming to jointly address facial detail generation and identity preservation. However, these methods often suffer from a trade-off between visual quality and identity fidelity, leading to either identity distortion or suboptimal degradation removal. In this paper, we present CodeFormer++, a novel framework that maximizes the utility of generative priors for high-quality face restoration while preserving identity. We decompose BFR into three sub-tasks: (i) identity- preserving face restoration, (ii) high-quality face generation, and (iii) dynamic fusion of identity features with realistic texture details. Our method makes three key contributions: (1) a learning-based deformable face registration module that semantically aligns generated and restored faces; (2) a texture guided restoration network to dynamically extract and transfer the texture of generated face to boost the quality of identity-preserving restored face; and (3) the integration of deep metric learning for BFR with the generation of informative positive and hard negative samples to better fuse identity- preserving and generative features. Extensive experiments on real-world and synthetic datasets demonstrate that, the pro- posed CodeFormer++ achieves superior performance in terms of both visual fidelity and identity consistency.

A.I.R.: Enabling Adaptive, Iterative, and Reasoning-based Frame Selection For Video Question Answering

arXiv:2510.04428v1 Announce Type: new Abstract: Effectively applying Vision-Language Models (VLMs) to Video Question Answering (VideoQA) hinges on selecting a concise yet comprehensive set of frames, as processing entire videos is computationally infeasible. However, current frame selection methods face a critical trade-off: approaches relying on lightweight similarity models, such as CLIP, often fail to capture the nuances of complex queries, resulting in inaccurate similarity scores that cannot reflect the authentic query-frame relevance, which further undermines frame selection. Meanwhile, methods that leverage a VLM for deeper analysis achieve higher accuracy but incur prohibitive computational costs. To address these limitations, we propose A.I.R., a training-free approach for Adaptive, Iterative, and Reasoning-based frame selection. We leverage a powerful VLM to perform deep, semantic analysis on complex queries, and this analysis is deployed within a cost-effective iterative loop that processes only a small batch of the most high-potential frames at a time. Extensive experiments on various VideoQA benchmarks demonstrate that our approach outperforms existing frame selection methods, significantly boosts the performance of the foundation VLM, and achieves substantial gains in computational efficiency over other VLM-based techniques.

REAR: Rethinking Visual Autoregressive Models via Generator-Tokenizer Consistency Regularization

arXiv:2510.04450v1 Announce Type: new Abstract: Visual autoregressive (AR) generation offers a promising path toward unifying vision and language models, yet its performance remains suboptimal against diffusion models. Prior work often attributes this gap to tokenizer limitations and rasterization ordering. In this work, we identify a core bottleneck from the perspective of generator-tokenizer inconsistency, i.e., the AR-generated tokens may not be well-decoded by the tokenizer. To address this, we propose reAR, a simple training strategy introducing a token-wise regularization objective: when predicting the next token, the causal transformer is also trained to recover the visual embedding of the current token and predict the embedding of the target token under a noisy context. It requires no changes to the tokenizer, generation order, inference pipeline, or external models. Despite its simplicity, reAR substantially improves performance. On ImageNet, it reduces gFID from 3.02 to 1.86 and improves IS to 316.9 using a standard rasterization-based tokenizer. When applied to advanced tokenizers, it achieves a gFID of 1.42 with only 177M parameters, matching the performance with larger state-of-the-art diffusion models (675M).

SPEGNet: Synergistic Perception-Guided Network for Camouflaged Object Detection

arXiv:2510.04472v1 Announce Type: new Abstract: Camouflaged object detection segments objects with intrinsic similarity and edge disruption. Current detection methods rely on accumulated complex components. Each approach adds components such as boundary modules, attention mechanisms, and multi-scale processors independently. This accumulation creates a computational burden without proportional gains. To manage this complexity, they process at reduced resolutions, eliminating fine details essential for camouflage. We present SPEGNet, addressing fragmentation through a unified design. The architecture integrates multi-scale features via channel calibration and spatial enhancement. Boundaries emerge directly from context-rich representations, maintaining semantic-spatial alignment. Progressive refinement implements scale-adaptive edge modulation with peak influence at intermediate resolutions. This design strikes a balance between boundary precision and regional consistency. SPEGNet achieves 0.887 $S_\alpha$ on CAMO, 0.890 on COD10K, and 0.895 on NC4K, with real-time inference speed. Our approach excels across scales, from tiny, intricate objects to large, pattern-similar ones, while handling occlusion and ambiguous boundaries. Code, model weights, and results are available on \href{https://github.com/Baber-Jan/SPEGNet}{https://github.com/Baber-Jan/SPEGNet}.

MedCLM: Learning to Localize and Reason via a CoT-Curriculum in Medical Vision-Language Models

arXiv:2510.04477v1 Announce Type: new Abstract: Bridging clinical diagnostic reasoning with AI remains a central challenge in medical imaging. We introduce MedCLM, an automated pipeline that converts detection datasets into large-scale medical visual question answering (VQA) data with Chain-of-Thought (CoT) reasoning by linking lesion boxes to organ segmentation and structured rationales. These contextual signals enable medical vision-language models to generate question-answer pairs with step-by-step reasoning. To utilize this data effectively, we propose an Integrated CoT-Curriculum Strategy composed of an Easy stage with explicit lesion boxes for visual grounding, a Medium stage that encourages implicit localization, and a Hard stage for weakly supervised reasoning. Experimental results demonstrate that MedCLM attains state-of-the-art performance on several medical VQA benchmarks, providing a scalable framework for developing clinically aligned medical vision-language models.

VaseVQA-3D: Benchmarking 3D VLMs on Ancient Greek Pottery

arXiv:2510.04479v1 Announce Type: new Abstract: Vision-Language Models (VLMs) have achieved significant progress in multimodal understanding tasks, demonstrating strong capabilities particularly in general tasks such as image captioning and visual reasoning. However, when dealing with specialized cultural heritage domains like 3D vase artifacts, existing models face severe data scarcity issues and insufficient domain knowledge limitations. Due to the lack of targeted training data, current VLMs struggle to effectively handle such culturally significant specialized tasks. To address these challenges, we propose the VaseVQA-3D dataset, which serves as the first 3D visual question answering dataset for ancient Greek pottery analysis, collecting 664 ancient Greek vase 3D models with corresponding question-answer data and establishing a complete data construction pipeline. We further develop the VaseVLM model, enhancing model performance in vase artifact analysis through domain-adaptive training. Experimental results validate the effectiveness of our approach, where we improve by 12.8% on R@1 metrics and by 6.6% on lexical similarity compared with previous state-of-the-art on the VaseVQA-3D dataset, significantly improving the recognition and understanding of 3D vase artifacts, providing new technical pathways for digital heritage preservation research.

TBStar-Edit: From Image Editing Pattern Shifting to Consistency Enhancement

arXiv:2510.04483v1 Announce Type: new Abstract: Recent advances in image generation and editing technologies have enabled state-of-the-art models to achieve impressive results in general domains. However, when applied to e-commerce scenarios, these general models often encounter consistency limitations. To address this challenge, we introduce TBStar-Edit, an new image editing model tailored for the e-commerce domain. Through rigorous data engineering, model architecture design and training strategy, TBStar-Edit achieves precise and high-fidelity image editing while maintaining the integrity of product appearance and layout. Specifically, for data engineering, we establish a comprehensive data construction pipeline, encompassing data collection, construction, filtering, and augmentation, to acquire high-quality, instruction-following, and strongly consistent editing data to support model training. For model architecture design, we design a hierarchical model framework consisting of a base model, pattern shifting modules, and consistency enhancement modules. For model training, we adopt a two-stage training strategy to enhance the consistency preservation: first stage for editing pattern shifting, and second stage for consistency enhancement. Each stage involves training different modules with separate datasets. Finally, we conduct extensive evaluations of TBStar-Edit on a self-proposed e-commerce benchmark, and the results demonstrate that TBStar-Edit outperforms existing general-domain editing models in both objective metrics (VIE Score) and subjective user preference.

Asynchronous Denoising Diffusion Models for Aligning Text-to-Image Generation

arXiv:2510.04504v1 Announce Type: new Abstract: Diffusion models have achieved impressive results in generating high-quality images. Yet, they often struggle to faithfully align the generated images with the input prompts. This limitation arises from synchronous denoising, where all pixels simultaneously evolve from random noise to clear images. As a result, during generation, the prompt-related regions can only reference the unrelated regions at the same noise level, failing to obtain clear context and ultimately impairing text-to-image alignment. To address this issue, we propose asynchronous diffusion models -- a novel framework that allocates distinct timesteps to different pixels and reformulates the pixel-wise denoising process. By dynamically modulating the timestep schedules of individual pixels, prompt-related regions are denoised more gradually than unrelated regions, thereby allowing them to leverage clearer inter-pixel context. Consequently, these prompt-related regions achieve better alignment in the final images. Extensive experiments demonstrate that our asynchronous diffusion models can significantly improve text-to-image alignment across diverse prompts. The code repository for this work is available at https://github.com/hu-zijing/AsynDM.

TAG:Tangential Amplifying Guidance for Hallucination-Resistant Diffusion Sampling

arXiv:2510.04533v1 Announce Type: new Abstract: Recent diffusion models achieve the state-of-the-art performance in image generation, but often suffer from semantic inconsistencies or hallucinations. While various inference-time guidance methods can enhance generation, they often operate indirectly by relying on external signals or architectural modifications, which introduces additional computational overhead. In this paper, we propose Tangential Amplifying Guidance (TAG), a more efficient and direct guidance method that operates solely on trajectory signals without modifying the underlying diffusion model. TAG leverages an intermediate sample as a projection basis and amplifies the tangential components of the estimated scores with respect to this basis to correct the sampling trajectory. We formalize this guidance process by leveraging a first-order Taylor expansion, which demonstrates that amplifying the tangential component steers the state toward higher-probability regions, thereby reducing inconsistencies and enhancing sample quality. TAG is a plug-and-play, architecture-agnostic module that improves diffusion sampling fidelity with minimal computational addition, offering a new perspective on diffusion guidance.

Conditional Representation Learning for Customized Tasks

arXiv:2510.04564v1 Announce Type: new Abstract: Conventional representation learning methods learn a universal representation that primarily captures dominant semantics, which may not always align with customized downstream tasks. For instance, in animal habitat analysis, researchers prioritize scene-related features, whereas universal embeddings emphasize categorical semantics, leading to suboptimal results. As a solution, existing approaches resort to supervised fine-tuning, which however incurs high computational and annotation costs. In this paper, we propose Conditional Representation Learning (CRL), aiming to extract representations tailored to arbitrary user-specified criteria. Specifically, we reveal that the semantics of a space are determined by its basis, thereby enabling a set of descriptive words to approximate the basis for a customized feature space. Building upon this insight, given a user-specified criterion, CRL first employs a large language model (LLM) to generate descriptive texts to construct the semantic basis, then projects the image representation into this conditional feature space leveraging a vision-language model (VLM). The conditional representation better captures semantics for the specific criterion, which could be utilized for multiple customized tasks. Extensive experiments on classification and retrieval tasks demonstrate the superiority and generality of the proposed CRL. The code is available at https://github.com/XLearning-SCU/2025-NeurIPS-CRL.

Pathology-CoT: Learning Visual Chain-of-Thought Agent from Expert Whole Slide Image Diagnosis Behavior

arXiv:2510.04587v1 Announce Type: new Abstract: Diagnosing a whole-slide image is an interactive, multi-stage process involving changes in magnification and movement between fields. Although recent pathology foundation models are strong, practical agentic systems that decide what field to examine next, adjust magnification, and deliver explainable diagnoses are still lacking. The blocker is data: scalable, clinically aligned supervision of expert viewing behavior that is tacit and experience-based, not written in textbooks or online, and therefore absent from large language model training. We introduce the AI Session Recorder, which works with standard WSI viewers to unobtrusively record routine navigation and convert the viewer logs into standardized behavioral commands (inspect or peek at discrete magnifications) and bounding boxes. A lightweight human-in-the-loop review turns AI-drafted rationales into the Pathology-CoT dataset, a form of paired "where to look" and "why it matters" supervision produced at roughly six times lower labeling time. Using this behavioral data, we build Pathologist-o3, a two-stage agent that first proposes regions of interest and then performs behavior-guided reasoning. On gastrointestinal lymph-node metastasis detection, it achieved 84.5% precision, 100.0% recall, and 75.4% accuracy, exceeding the state-of-the-art OpenAI o3 model and generalizing across backbones. To our knowledge, this constitutes one of the first behavior-grounded agentic systems in pathology. Turning everyday viewer logs into scalable, expert-validated supervision, our framework makes agentic pathology practical and establishes a path to human-aligned, upgradeable clinical AI.

A Spatial-Spectral-Frequency Interactive Network for Multimodal Remote Sensing Classification

arXiv:2510.04628v1 Announce Type: new Abstract: Deep learning-based methods have achieved significant success in remote sensing Earth observation data analysis. Numerous feature fusion techniques address multimodal remote sensing image classification by integrating global and local features. However, these techniques often struggle to extract structural and detail features from heterogeneous and redundant multimodal images. With the goal of introducing frequency domain learning to model key and sparse detail features, this paper introduces the spatial-spectral-frequency interaction network (S$^2$Fin), which integrates pairwise fusion modules across the spatial, spectral, and frequency domains. Specifically, we propose a high-frequency sparse enhancement transformer that employs sparse spatial-spectral attention to optimize the parameters of the high-frequency filter. Subsequently, a two-level spatial-frequency fusion strategy is introduced, comprising an adaptive frequency channel module that fuses low-frequency structures with enhanced high-frequency details, and a high-frequency resonance mask that emphasizes sharp edges via phase similarity. In addition, a spatial-spectral attention fusion module further enhances feature extraction at intermediate layers of the network. Experiments on four benchmark multimodal datasets with limited labeled data demonstrate that S$^2$Fin performs superior classification, outperforming state-of-the-art methods. The code is available at https://github.com/HaoLiu-XDU/SSFin.

SFANet: Spatial-Frequency Attention Network for Deepfake Detection

arXiv:2510.04630v1 Announce Type: new Abstract: Detecting manipulated media has now become a pressing issue with the recent rise of deepfakes. Most existing approaches fail to generalize across diverse datasets and generation techniques. We thus propose a novel ensemble framework, combining the strengths of transformer-based architectures, such as Swin Transformers and ViTs, and texture-based methods, to achieve better detection accuracy and robustness. Our method introduces innovative data-splitting, sequential training, frequency splitting, patch-based attention, and face segmentation techniques to handle dataset imbalances, enhance high-impact regions (e.g., eyes and mouth), and improve generalization. Our model achieves state-of-the-art performance when tested on the DFWild-Cup dataset, a diverse subset of eight deepfake datasets. The ensemble benefits from the complementarity of these approaches, with transformers excelling in global feature extraction and texturebased methods providing interpretability. This work demonstrates that hybrid models can effectively address the evolving challenges of deepfake detection, offering a robust solution for real-world applications.

Do Superpixel Segmentation Methods Influence Deforestation Image Classification?

arXiv:2510.04645v1 Announce Type: new Abstract: Image segmentation is a crucial step in various visual applications, including environmental monitoring through remote sensing. In the context of the ForestEyes project, which combines citizen science and machine learning to detect deforestation in tropical forests, image segments are used for labeling by volunteers and subsequent model training. Traditionally, the Simple Linear Iterative Clustering (SLIC) algorithm is adopted as the segmentation method. However, recent studies have indicated that other superpixel-based methods outperform SLIC in remote sensing image segmentation, and might suggest that they are more suitable for the task of detecting deforested areas. In this sense, this study investigated the impact of the four best segmentation methods, together with SLIC, on the training of classifiers for the target application. Initially, the results showed little variation in performance among segmentation methods, even when selecting the top five classifiers using the PyCaret AutoML library. However, by applying a classifier fusion approach (ensemble of classifiers), noticeable improvements in balanced accuracy were observed, highlighting the importance of both the choice of segmentation method and the combination of machine learning-based models for deforestation detection tasks.

EduPersona: Benchmarking Subjective Ability Boundaries of Virtual Student Agents

arXiv:2510.04648v1 Announce Type: new Abstract: As large language models are increasingly integrated into education, virtual student agents are becoming vital for classroom simulation and teacher training. Yet their classroom-oriented subjective abilities remain largely unassessed, limiting understanding of model boundaries and hindering trustworthy deployment. We present EduPersona, a large-scale benchmark spanning two languages, three subjects, and ten persona types based on the Big Five theory. The dataset contains 1,308 authentic classroom dialogue rounds, corresponding to 12,814 teacher-student Q&A turns, and is further expanded through persona stylization into roughly 10 times larger scale (128k turns), providing a solid foundation for evaluation. Building on this resource, we decompose hard-to-quantify subjective performance into three progressive tasks: TASK1 basic coherence (whether behavior, emotion, expression, and voice align with classroom context), TASK2 student realism, and TASK3 long-term persona consistency, thereby establishing an evaluation framework grounded in educational theory and research value. We conduct systematic experiments on three representative LLMs, comparing their original versions with ten persona-fine-tuned variants trained on EduPersona. Results show consistent and significant average improvements across all tasks: TASK1 +33.6%, TASK2 +30.6%, and TASK3 +14.9%. These improvements highlight the dataset's effectiveness and research value, while also revealing the heterogeneous difficulty of persona modeling. In summary, EduPersona delivers the first classroom benchmark centered on subjective abilities, establishes a decoupled and verifiable research paradigm, and we will open-source both the dataset and the framework to support the broader research community in advancing trustworthy and human-like AI for education.

MoME: Estimating Psychological Traits from Gait with Multi-Stage Mixture of Movement Experts

arXiv:2510.04654v1 Announce Type: new Abstract: Gait encodes rich biometric and behavioural information, yet leveraging the manner of walking to infer psychological traits remains a challenging and underexplored problem. We introduce a hierarchical Multi-Stage Mixture of Movement Experts (MoME) architecture for multi-task prediction of psychological attributes from gait sequences represented as 2D poses. MoME processes the walking cycle in four stages of movement complexity, employing lightweight expert models to extract spatio-temporal features and task-specific gating modules to adaptively weight experts across traits and stages. Evaluated on the PsyMo benchmark covering 17 psychological traits, our method outperforms state-of-the-art gait analysis models, achieving a 37.47% weighted F1 score at the run level and 44.6% at the subject level. Our experiments show that integrating auxiliary tasks such as identity recognition, gender prediction, and BMI estimation further improves psychological trait estimation. Our findings demonstrate the viability of multi-task gait-based learning for psychological trait estimation and provide a foundation for future research on movement-informed psychological inference.

ConceptSplit: Decoupled Multi-Concept Personalization of Diffusion Models via Token-wise Adaptation and Attention Disentanglement

arXiv:2510.04668v1 Announce Type: new Abstract: In recent years, multi-concept personalization for text-to-image (T2I) diffusion models to represent several subjects in an image has gained much more attention. The main challenge of this task is "concept mixing", where multiple learned concepts interfere or blend undesirably in the output image. To address this issue, in this paper, we present ConceptSplit, a novel framework to split the individual concepts through training and inference. Our framework comprises two key components. First, we introduce Token-wise Value Adaptation (ToVA), a merging-free training method that focuses exclusively on adapting the value projection in cross-attention. Based on our empirical analysis, we found that modifying the key projection, a common approach in existing methods, can disrupt the attention mechanism and lead to concept mixing. Second, we propose Latent Optimization for Disentangled Attention (LODA), which alleviates attention entanglement during inference by optimizing the input latent. Through extensive qualitative and quantitative experiments, we demonstrate that ConceptSplit achieves robust multi-concept personalization, mitigating unintended concept interference. Code is available at https://github.com/KU-VGI/ConceptSplit

Label-Efficient Cross-Modality Generalization for Liver Segmentation in Multi-Phase MRI

arXiv:2510.04705v1 Announce Type: new Abstract: Accurate liver segmentation in multi-phase MRI is vital for liver fibrosis assessment, yet labeled data is often scarce and unevenly distributed across imaging modalities and vendor systems. We propose a label-efficient segmentation approach that promotes cross-modality generalization under real-world conditions, where GED4 hepatobiliary-phase annotations are limited, non-contrast sequences (T1WI, T2WI, DWI) are unlabeled, and spatial misalignment and missing phases are common. Our method integrates a foundation-scale 3D segmentation backbone adapted via fine-tuning, co-training with cross pseudo supervision to leverage unlabeled volumes, and a standardized preprocessing pipeline. Without requiring spatial registration, the model learns to generalize across MRI phases and vendors, demonstrating robust segmentation performance in both labeled and unlabeled domains. Our results exhibit the effectiveness of our proposed label-efficient baseline for liver segmentation in multi-phase, multi-vendor MRI and highlight the potential of combining foundation model adaptation with co-training for real-world clinical imaging tasks.

ID-Consistent, Precise Expression Generation with Blendshape-Guided Diffusion

arXiv:2510.04706v1 Announce Type: new Abstract: Human-centric generative models designed for AI-driven storytelling must bring together two core capabilities: identity consistency and precise control over human performance. While recent diffusion-based approaches have made significant progress in maintaining facial identity, achieving fine-grained expression control without compromising identity remains challenging. In this work, we present a diffusion-based framework that faithfully reimagines any subject under any particular facial expression. Building on an ID-consistent face foundation model, we adopt a compositional design featuring an expression cross-attention module guided by FLAME blendshape parameters for explicit control. Trained on a diverse mixture of image and video data rich in expressive variation, our adapter generalizes beyond basic emotions to subtle micro-expressions and expressive transitions, overlooked by prior works. In addition, a pluggable Reference Adapter enables expression editing in real images by transferring the appearance from a reference frame during synthesis. Extensive quantitative and qualitative evaluations show that our model outperforms existing methods in tailored and identity-consistent expression generation. Code and models can be found at https://github.com/foivospar/Arc2Face.

ReactDiff: Fundamental Multiple Appropriate Facial Reaction Diffusion Model

arXiv:2510.04712v1 Announce Type: new Abstract: The automatic generation of diverse and human-like facial reactions in dyadic dialogue remains a critical challenge for human-computer interaction systems. Existing methods fail to model the stochasticity and dynamics inherent in real human reactions. To address this, we propose ReactDiff, a novel temporal diffusion framework for generating diverse facial reactions that are appropriate for responding to any given dialogue context. Our key insight is that plausible human reactions demonstrate smoothness, and coherence over time, and conform to constraints imposed by human facial anatomy. To achieve this, ReactDiff incorporates two vital priors (spatio-temporal facial kinematics) into the diffusion process: i) temporal facial behavioral kinematics and ii) facial action unit dependencies. These two constraints guide the model toward realistic human reaction manifolds, avoiding visually unrealistic jitters, unstable transitions, unnatural expressions, and other artifacts. Extensive experiments on the REACT2024 dataset demonstrate that our approach not only achieves state-of-the-art reaction quality but also excels in diversity and reaction appropriateness.

Object-Centric Representation Learning for Enhanced 3D Scene Graph Prediction

arXiv:2510.04714v1 Announce Type: new Abstract: 3D Semantic Scene Graph Prediction aims to detect objects and their semantic relationships in 3D scenes, and has emerged as a crucial technology for robotics and AR/VR applications. While previous research has addressed dataset limitations and explored various approaches including Open-Vocabulary settings, they frequently fail to optimize the representational capacity of object and relationship features, showing excessive reliance on Graph Neural Networks despite insufficient discriminative capability. In this work, we demonstrate through extensive analysis that the quality of object features plays a critical role in determining overall scene graph accuracy. To address this challenge, we design a highly discriminative object feature encoder and employ a contrastive pretraining strategy that decouples object representation learning from the scene graph prediction. This design not only enhances object classification accuracy but also yields direct improvements in relationship prediction. Notably, when plugging in our pretrained encoder into existing frameworks, we observe substantial performance improvements across all evaluation metrics. Additionally, whereas existing approaches have not fully exploited the integration of relationship information, we effectively combine both geometric and semantic features to achieve superior relationship prediction. Comprehensive experiments on the 3DSSG dataset demonstrate that our approach significantly outperforms previous state-of-the-art methods. Our code is publicly available at https://github.com/VisualScienceLab-KHU/OCRL-3DSSG-Codes.

Benchmark on Monocular Metric Depth Estimation in Wildlife Setting

arXiv:2510.04723v1 Announce Type: new Abstract: Camera traps are widely used for wildlife monitoring, but extracting accurate distance measurements from monocular images remains challenging due to the lack of depth information. While monocular depth estimation (MDE) methods have advanced significantly, their performance in natural wildlife environments has not been systematically evaluated. This work introduces the first benchmark for monocular metric depth estimation in wildlife monitoring conditions. We evaluate four state-of-the-art MDE methods (Depth Anything V2, ML Depth Pro, ZoeDepth, and Metric3D) alongside a geometric baseline on 93 camera trap images with ground truth distances obtained using calibrated ChARUCO patterns. Our results demonstrate that Depth Anything V2 achieves the best overall performance with a mean absolute error of 0.454m and correlation of 0.962, while methods like ZoeDepth show significant degradation in outdoor natural environments (MAE: 3.087m). We find that median-based depth extraction consistently outperforms mean-based approaches across all deep learning methods. Additionally, we analyze computational efficiency, with ZoeDepth being fastest (0.17s per image) but least accurate, while Depth Anything V2 provides an optimal balance of accuracy and speed (0.22s per image). This benchmark establishes performance baselines for wildlife applications and provides practical guidance for implementing depth estimation in conservation monitoring systems.

ExposureEngine: Oriented Logo Detection and Sponsor Visibility Analytics in Sports Broadcasts

arXiv:2510.04739v1 Announce Type: new Abstract: Quantifying sponsor visibility in sports broadcasts is a critical marketing task traditionally hindered by manual, subjective, and unscalable analysis methods. While automated systems offer an alternative, their reliance on axis-aligned Horizontal Bounding Box (HBB) leads to inaccurate exposuremetrics when logos appear rotated or skewed due to dynamic camera angles and perspective distortions. This paper introduces ExposureEngine, an end-to-end system designed for accurate, rotation-aware sponsor visibility analytics in sports broadcasts, demonstrated in a soccer case study. Our approach predicts Oriented Bounding Box (OBB) to provide a geometrically precise fit to each logo regardless of the orientation on-screen. To train and evaluate our detector, we developed a new dataset comprising 1,103 frames from Swedish elite soccer, featuring 670 unique sponsor logos annotated with OBBs. Our model achieves a mean Average Precision (mAP@0.5) of 0.859, with a precision of 0.96 and recall of 0.87, demonstrating robust performance in localizing logos under diverse broadcast conditions. The system integrates these detections into an analytical pipeline that calculates precise visibility metrics, such as exposure duration and on-screen coverage. Furthermore, we incorporate a language-driven agentic layer, enabling users to generate reports, summaries, and media content through natural language queries. The complete system, including the dataset and the analytics dashboard, provides a comprehensive solution for auditable and interpretable sponsor measurement in sports media. An overview of the ExposureEngine is available online: https://youtu.be/tRw6OBISuW4 .

Anomaly-Aware YOLO: A Frugal yet Robust Approach to Infrared Small Target Detection

arXiv:2510.04741v1 Announce Type: new Abstract: Infrared Small Target Detection (IRSTD) is a challenging task in defense applications, where complex backgrounds and tiny target sizes often result in numerous false alarms using conventional object detectors. To overcome this limitation, we propose Anomaly-Aware YOLO (AA-YOLO), which integrates a statistical anomaly detection test into its detection head. By treating small targets as unexpected patterns against the background, AA-YOLO effectively controls the false alarm rate. Our approach not only achieves competitive performance on several IRSTD benchmarks, but also demonstrates remarkable robustness in scenarios with limited training data, noise, and domain shifts. Furthermore, since only the detection head is modified, our design is highly generic and has been successfully applied across various YOLO backbones, including lightweight models. It also provides promising results when integrated into an instance segmentation YOLO. This versatility makes AA-YOLO an attractive solution for real-world deployments where resources are constrained. The code will be publicly released.

Beyond Appearance: Transformer-based Person Identification from Conversational Dynamics

arXiv:2510.04753v1 Announce Type: new Abstract: This paper investigates the performance of transformer-based architectures for person identification in natural, face-to-face conversation scenario. We implement and evaluate a two-stream framework that separately models spatial configurations and temporal motion patterns of 133 COCO WholeBody keypoints, extracted from a subset of the CANDOR conversational corpus. Our experiments compare pre-trained and from-scratch training, investigate the use of velocity features, and introduce a multi-scale temporal transformer for hierarchical motion modeling. Results demonstrate that domain-specific training significantly outperforms transfer learning, and that spatial configurations carry more discriminative information than temporal dynamics. The spatial transformer achieves 95.74% accuracy, while the multi-scale temporal transformer achieves 93.90%. Feature-level fusion pushes performance to 98.03%, confirming that postural and dynamic information are complementary. These findings highlight the potential of transformer architectures for person identification in natural interactions and provide insights for future multimodal and cross-cultural studies.

Progressive Gaussian Transformer with Anisotropy-aware Sampling for Open Vocabulary Occupancy Prediction

arXiv:2510.04759v1 Announce Type: new Abstract: The 3D occupancy prediction task has witnessed remarkable progress in recent years, playing a crucial role in vision-based autonomous driving systems. While traditional methods are limited to fixed semantic categories, recent approaches have moved towards predicting text-aligned features to enable open-vocabulary text queries in real-world scenes. However, there exists a trade-off in text-aligned scene modeling: sparse Gaussian representation struggles to capture small objects in the scene, while dense representation incurs significant computational overhead. To address these limitations, we present PG-Occ, an innovative Progressive Gaussian Transformer Framework that enables open-vocabulary 3D occupancy prediction. Our framework employs progressive online densification, a feed-forward strategy that gradually enhances the 3D Gaussian representation to capture fine-grained scene details. By iteratively enhancing the representation, the framework achieves increasingly precise and detailed scene understanding. Another key contribution is the introduction of an anisotropy-aware sampling strategy with spatio-temporal fusion, which adaptively assigns receptive fields to Gaussians at different scales and stages, enabling more effective feature aggregation and richer scene information capture. Through extensive evaluations, we demonstrate that PG-Occ achieves state-of-the-art performance with a relative 14.3% mIoU improvement over the previous best performing method. Code and pretrained models will be released upon publication on our project page: https://yanchi-3dv.github.io/PG-Occ

Beyond the Seen: Bounded Distribution Estimation for Open-Vocabulary Learning

arXiv:2510.04770v1 Announce Type: new Abstract: Open-vocabulary learning requires modeling the data distribution in open environments, which consists of both seen-class and unseen-class data. Existing methods estimate the distribution in open environments using seen-class data, where the absence of unseen classes makes the estimation error inherently unidentifiable. Intuitively, learning beyond the seen classes is crucial for distribution estimation to bound the estimation error. We theoretically demonstrate that the distribution can be effectively estimated by generating unseen-class data, through which the estimation error is upper-bounded. Building on this theoretical insight, we propose a novel open-vocabulary learning method, which generates unseen-class data for estimating the distribution in open environments. The method consists of a class-domain-wise data generation pipeline and a distribution alignment algorithm. The data generation pipeline generates unseen-class data under the guidance of a hierarchical semantic tree and domain information inferred from the seen-class data, facilitating accurate distribution estimation. With the generated data, the distribution alignment algorithm estimates and maximizes the posterior probability to enhance generalization in open-vocabulary learning. Extensive experiments on $11$ datasets demonstrate that our method outperforms baseline approaches by up to $14\%$, highlighting its effectiveness and superiority.

Federated Learning for Surgical Vision in Appendicitis Classification: Results of the FedSurg EndoVis 2024 Challenge

arXiv:2510.04772v1 Announce Type: new Abstract: Purpose: The FedSurg challenge was designed to benchmark the state of the art in federated learning for surgical video classification. Its goal was to assess how well current methods generalize to unseen clinical centers and adapt through local fine-tuning while enabling collaborative model development without sharing patient data. Methods: Participants developed strategies to classify inflammation stages in appendicitis using a preliminary version of the multi-center Appendix300 video dataset. The challenge evaluated two tasks: generalization to an unseen center and center-specific adaptation after fine-tuning. Submitted approaches included foundation models with linear probing, metric learning with triplet loss, and various FL aggregation schemes (FedAvg, FedMedian, FedSAM). Performance was assessed using F1-score and Expected Cost, with ranking robustness evaluated via bootstrapping and statistical testing. Results: In the generalization task, performance across centers was limited. In the adaptation task, all teams improved after fine-tuning, though ranking stability was low. The ViViT-based submission achieved the strongest overall performance. The challenge highlighted limitations in generalization, sensitivity to class imbalance, and difficulties in hyperparameter tuning in decentralized training, while spatiotemporal modeling and context-aware preprocessing emerged as promising strategies. Conclusion: The FedSurg Challenge establishes the first benchmark for evaluating FL strategies in surgical video classification. Findings highlight the trade-off between local personalization and global robustness, and underscore the importance of architecture choice, preprocessing, and loss design. This benchmarking offers a reference point for future development of imbalance-aware, adaptive, and robust FL methods in clinical surgical AI.

Hands-Free Heritage: Automated 3D Scanning for Cultural Heritage Digitization

arXiv:2510.04781v1 Announce Type: new Abstract: High-fidelity 3D scanning is essential for preserving cultural heritage artefacts, supporting documentation, analysis, and long-term conservation. However, conventional methods typically require specialized expertise and manual intervention to maintain optimal scanning conditions and coverage. We present an automated two-robot scanning system that eliminates the need for handheld or semi-automatic workflows by combining coordinated robotic manipulation with high-resolution 3D scanning. Our system parameterizes the scanning space into distinct regions, enabling coordinated motion planning between a scanner-equipped robot and a tray-handling robot. Optimized trajectory planning and waypoint distribution ensure comprehensive surface coverage, minimize occlusions, and balance reconstruction accuracy with system efficiency. Experimental results show that our approach achieves significantly lower Chamfer Distance and higher F-score compared to baseline methods, offering superior geometric accuracy, improved digitization efficiency, and reduced reliance on expert operators.

A Comparative Study of Vision Transformers and CNNs for Few-Shot Rigid Transformation and Fundamental Matrix Estimation

arXiv:2510.04794v1 Announce Type: new Abstract: Vision-transformers (ViTs) and large-scale convolution-neural-networks (CNNs) have reshaped computer vision through pretrained feature representations that enable strong transfer learning for diverse tasks. However, their efficiency as backbone architectures for geometric estimation tasks involving image deformations in low-data regimes remains an open question. This work considers two such tasks: 1) estimating 2D rigid transformations between pairs of images and 2) predicting the fundamental matrix for stereo image pairs, an important problem in various applications, such as autonomous mobility, robotics, and 3D scene reconstruction. Addressing this intriguing question, this work systematically compares large-scale CNNs (ResNet, EfficientNet, CLIP-ResNet) with ViT-based foundation models (CLIP-ViT variants and DINO) in various data size settings, including few-shot scenarios. These pretrained models are optimized for classification or contrastive learning, encouraging them to focus mostly on high-level semantics. The considered tasks require balancing local and global features differently, challenging the straightforward adoption of these models as the backbone. Empirical comparative analysis shows that, similar to training from scratch, ViTs outperform CNNs during refinement in large downstream-data scenarios. However, in small data scenarios, the inductive bias and smaller capacity of CNNs improve their performance, allowing them to match that of a ViT. Moreover, ViTs exhibit stronger generalization in cross-domain evaluation where the data distribution changes. These results emphasize the importance of carefully selecting model architectures for refinement, motivating future research towards hybrid architectures that balance local and global representations.

DiT-VTON: Diffusion Transformer Framework for Unified Multi-Category Virtual Try-On and Virtual Try-All with Integrated Image Editing

arXiv:2510.04797v1 Announce Type: new Abstract: The rapid growth of e-commerce has intensified the demand for Virtual Try-On (VTO) technologies, enabling customers to realistically visualize products overlaid on their own images. Despite recent advances, existing VTO models face challenges with fine-grained detail preservation, robustness to real-world imagery, efficient sampling, image editing capabilities, and generalization across diverse product categories. In this paper, we present DiT-VTON, a novel VTO framework that leverages a Diffusion Transformer (DiT), renowned for its performance on text-conditioned image generation, adapted here for the image-conditioned VTO task. We systematically explore multiple DiT configurations, including in-context token concatenation, channel concatenation, and ControlNet integration, to determine the best setup for VTO image conditioning. To enhance robustness, we train the model on an expanded dataset encompassing varied backgrounds, unstructured references, and non-garment categories, demonstrating the benefits of data scaling for VTO adaptability. DiT-VTON also redefines the VTO task beyond garment try-on, offering a versatile Virtual Try-All (VTA) solution capable of handling a wide range of product categories and supporting advanced image editing functionalities such as pose preservation, localized editing, texture transfer, and object-level customization. Experimental results show that our model surpasses state-of-the-art methods on VITON-HD, achieving superior detail preservation and robustness without reliance on additional condition encoders. It also outperforms models with VTA and image editing capabilities on a diverse dataset spanning thousands of product categories.

Did you just see that? Arbitrary view synthesis for egocentric replay of operating room workflows from ambient sensors

arXiv:2510.04802v1 Announce Type: new Abstract: Observing surgical practice has historically relied on fixed vantage points or recollections, leaving the egocentric visual perspectives that guide clinical decisions undocumented. Fixed-camera video can capture surgical workflows at the room-scale, but cannot reconstruct what each team member actually saw. Thus, these videos only provide limited insights into how decisions that affect surgical safety, training, and workflow optimization are made. Here we introduce EgoSurg, the first framework to reconstruct the dynamic, egocentric replays for any operating room (OR) staff directly from wall-mounted fixed-camera video, and thus, without intervention to clinical workflow. EgoSurg couples geometry-driven neural rendering with diffusion-based view enhancement, enabling high-visual fidelity synthesis of arbitrary and egocentric viewpoints at any moment. In evaluation across multi-site surgical cases and controlled studies, EgoSurg reconstructs person-specific visual fields and arbitrary viewpoints with high visual quality and fidelity. By transforming existing OR camera infrastructure into a navigable dynamic 3D record, EgoSurg establishes a new foundation for immersive surgical data science, enabling surgical practice to be visualized, experienced, and analyzed from every angle.

Visual Representations inside the Language Model

arXiv:2510.04819v1 Announce Type: new Abstract: Despite interpretability work analyzing VIT encoders and transformer activations, we don't yet understand why Multimodal Language Models (MLMs) struggle on perception-heavy tasks. We offer an under-studied perspective by examining how popular MLMs (LLaVA-OneVision, Qwen2.5-VL, and Llama-3-LLaVA-NeXT) process their visual key-value tokens. We first study the flow of visual information through the language model, finding that image value tokens encode sufficient information to perform several perception-heavy tasks zero-shot: segmentation, semantic correspondence, temporal correspondence, and referring expression detection. We find that while the language model does augment the visual information received from the projection of input visual encodings-which we reveal correlates with overall MLM perception capability-it contains less visual information on several tasks than the equivalent visual encoder (SigLIP) that has not undergone MLM finetuning. Further, we find that the visual information corresponding to input-agnostic image key tokens in later layers of language models contains artifacts which reduce perception capability of the overall MLM. Next, we discuss controlling visual information in the language model, showing that adding a text prefix to the image input improves perception capabilities of visual representations. Finally, we reveal that if language models were able to better control their visual information, their perception would significantly improve; e.g., in 33.3% of Art Style questions in the BLINK benchmark, perception information present in the language model is not surfaced to the output! Our findings reveal insights into the role of key-value tokens in multimodal systems, paving the way for deeper mechanistic interpretability of MLMs and suggesting new directions for training their visual encoder and language model components.

AvatarVTON: 4D Virtual Try-On for Animatable Avatars

arXiv:2510.04822v1 Announce Type: new Abstract: We propose AvatarVTON, the first 4D virtual try-on framework that generates realistic try-on results from a single in-shop garment image, enabling free pose control, novel-view rendering, and diverse garment choices. Unlike existing methods, AvatarVTON supports dynamic garment interactions under single-view supervision, without relying on multi-view garment captures or physics priors. The framework consists of two key modules: (1) a Reciprocal Flow Rectifier, a prior-free optical-flow correction strategy that stabilizes avatar fitting and ensures temporal coherence; and (2) a Non-Linear Deformer, which decomposes Gaussian maps into view-pose-invariant and view-pose-specific components, enabling adaptive, non-linear garment deformations. To establish a benchmark for 4D virtual try-on, we extend existing baselines with unified modules for fair qualitative and quantitative comparisons. Extensive experiments show that AvatarVTON achieves high fidelity, diversity, and dynamic garment realism, making it well-suited for AR/VR, gaming, and digital-human applications.

Flow Matching for Conditional MRI-CT and CBCT-CT Image Synthesis

arXiv:2510.04823v1 Announce Type: new Abstract: Generating synthetic CT (sCT) from MRI or CBCT plays a crucial role in enabling MRI-only and CBCT-based adaptive radiotherapy, improving treatment precision while reducing patient radiation exposure. To address this task, we adopt a fully 3D Flow Matching (FM) framework, motivated by recent work demonstrating FM's efficiency in producing high-quality images. In our approach, a Gaussian noise volume is transformed into an sCT image by integrating a learned FM velocity field, conditioned on features extracted from the input MRI or CBCT using a lightweight 3D encoder. We evaluated the method on the SynthRAD2025 Challenge benchmark, training separate models for MRI $\rightarrow$ sCT and CBCT $\rightarrow$ sCT across three anatomical regions: abdomen, head and neck, and thorax. Validation and testing were performed through the challenge submission system. The results indicate that the method accurately reconstructs global anatomical structures; however, preservation of fine details was limited, primarily due to the relatively low training resolution imposed by memory and runtime constraints. Future work will explore patch-based training and latent-space flow models to improve resolution and local structural fidelity.

Beyond Random: Automatic Inner-loop Optimization in Dataset Distillation

arXiv:2510.04838v1 Announce Type: new Abstract: The growing demand for efficient deep learning has positioned dataset distillation as a pivotal technique for compressing training dataset while preserving model performance. However, existing inner-loop optimization methods for dataset distillation typically rely on random truncation strategies, which lack flexibility and often yield suboptimal results. In this work, we observe that neural networks exhibit distinct learning dynamics across different training stages-early, middle, and late-making random truncation ineffective. To address this limitation, we propose Automatic Truncated Backpropagation Through Time (AT-BPTT), a novel framework that dynamically adapts both truncation positions and window sizes according to intrinsic gradient behavior. AT-BPTT introduces three key components: (1) a probabilistic mechanism for stage-aware timestep selection, (2) an adaptive window sizing strategy based on gradient variation, and (3) a low-rank Hessian approximation to reduce computational overhead. Extensive experiments on CIFAR-10, CIFAR-100, Tiny-ImageNet, and ImageNet-1K show that AT-BPTT achieves state-of-the-art performance, improving accuracy by an average of 6.16% over baseline methods. Moreover, our approach accelerates inner-loop optimization by 3.9x while saving 63% memory cost.

Detailed Aerial Mapping of Photovoltaic Power Plants Through Semantically Significant Keypoints

arXiv:2510.04840v1 Announce Type: new Abstract: An accurate and up-to-date model of a photovoltaic (PV) power plant is essential for its optimal operation and maintenance. However, such a model may not be easily available. This work introduces a novel approach for PV power plant mapping based on aerial overview images. It enables the automation of the mapping process while removing the reliance on third-party data. The presented mapping method takes advantage of the structural layout of the power plants to achieve detailed modeling down to the level of individual PV modules. The approach relies on visual segmentation of PV modules in overview images and the inference of structural information in each image, assigning modules to individual benches, rows, and columns. We identify visual keypoints related to the layout and use these to merge detections from multiple images while maintaining their structural integrity. The presented method was experimentally verified and evaluated on two different power plants. The final fusion of 3D positions and semantic structures results in a compact georeferenced model suitable for power plant maintenance.

From Actions to Kinesics: Extracting Human Psychological States through Bodily Movements

arXiv:2510.04844v1 Announce Type: new Abstract: Understanding the dynamic relationship between humans and the built environment is a key challenge in disciplines ranging from environmental psychology to reinforcement learning (RL). A central obstacle in modeling these interactions is the inability to capture human psychological states in a way that is both generalizable and privacy preserving. Traditional methods rely on theoretical models or questionnaires, which are limited in scope, static, and labor intensive. We present a kinesics recognition framework that infers the communicative functions of human activity -- known as kinesics -- directly from 3D skeleton joint data. Combining a spatial-temporal graph convolutional network (ST-GCN) with a convolutional neural network (CNN), the framework leverages transfer learning to bypass the need for manually defined mappings between physical actions and psychological categories. The approach preserves user anonymity while uncovering latent structures in bodily movements that reflect cognitive and emotional states. Our results on the Dyadic User EngagemenT (DUET) dataset demonstrate that this method enables scalable, accurate, and human-centered modeling of behavior, offering a new pathway for enhancing RL-driven simulations of human-environment interaction.

Read the Room: Inferring Social Context Through Dyadic Interaction Recognition in Cyber-physical-social Infrastructure Systems

arXiv:2510.04854v1 Announce Type: new Abstract: Cyber-physical systems (CPS) integrate sensing, computing, and control to improve infrastructure performance, focusing on economic goals like performance and safety. However, they often neglect potential human-centered (or ''social'') benefits. Cyber-physical-social infrastructure systems (CPSIS) aim to address this by aligning CPS with social objectives. This involves defining social benefits, understanding human interactions with each other and infrastructure, developing privacy-preserving measurement methods, modeling these interactions for prediction, linking them to social benefits, and actuating the physical environment to foster positive social outcomes. This paper delves into recognizing dyadic human interactions using real-world data, which is the backbone to measuring social behavior. This lays a foundation to address the need to enhance understanding of the deeper meanings and mutual responses inherent in human interactions. While RGB cameras are informative for interaction recognition, privacy concerns arise. Depth sensors offer a privacy-conscious alternative by analyzing skeletal movements. This study compares five skeleton-based interaction recognition algorithms on a dataset of 12 dyadic interactions. Unlike single-person datasets, these interactions, categorized into communication types like emblems and affect displays, offer insights into the cultural and emotional aspects of human interactions.

ERDE: Entropy-Regularized Distillation for Early-exit

arXiv:2510.04856v1 Announce Type: new Abstract: Although deep neural networks and in particular Convolutional Neural Networks have demonstrated state-of-the-art performance in image classification with relatively high efficiency, they still exhibit high computational costs, often rendering them impractical for real-time and edge applications. Therefore, a multitude of compression techniques have been developed to reduce these costs while maintaining accuracy. In addition, dynamic architectures have been introduced to modulate the level of compression at execution time, which is a desirable property in many resource-limited application scenarios. The proposed method effectively integrates two well-established optimization techniques: early exits and knowledge distillation, where a reduced student early-exit model is trained from a more complex teacher early-exit model. The primary contribution of this research lies in the approach for training the student early-exit model. In comparison to the conventional Knowledge Distillation loss, our approach incorporates a new entropy-based loss for images where the teacher's classification was incorrect. The proposed method optimizes the trade-off between accuracy and efficiency, thereby achieving significant reductions in computational complexity without compromising classification performance. The validity of this approach is substantiated by experimental results on image classification datasets CIFAR10, CIFAR100 and SVHN, which further opens new research perspectives for Knowledge Distillation in other contexts.

\mu DeepIQA: deep learning-based fast and robust image quality assessment with local predictions for optical microscopy

arXiv:2510.04859v1 Announce Type: new Abstract: Optical microscopy is one of the most widely used techniques in research studies for life sciences and biomedicine. These applications require reliable experimental pipelines to extract valuable knowledge from the measured samples and must be supported by image quality assessment (IQA) to ensure correct processing and analysis of the image data. IQA methods are implemented with variable complexity. However, while most quality metrics have a straightforward implementation, they might be time consuming and computationally expensive when evaluating a large dataset. In addition, quality metrics are often designed for well-defined image features and may be unstable for images out of the ideal domain. To overcome these limitations, recent works have proposed deep learning-based IQA methods, which can provide superior performance, increased generalizability and fast prediction. Our method, named $\mathrm{\mu}$DeepIQA, is inspired by previous studies and applies a deep convolutional neural network designed for IQA on natural images to optical microscopy measurements. We retrained the same architecture to predict individual quality metrics and global quality scores for optical microscopy data. The resulting models provide fast and stable predictions of image quality by generalizing quality estimation even outside the ideal range of standard methods. In addition, $\mathrm{\mu}$DeepIQA provides patch-wise prediction of image quality and can be used to visualize spatially varying quality in a single image. Our study demonstrates that optical microscopy-based studies can benefit from the generalizability of deep learning models due to their stable performance in the presence of outliers, the ability to assess small image patches, and rapid predictions.

In-Field Mapping of Grape Yield and Quality with Illumination-Invariant Deep Learning

arXiv:2510.04864v1 Announce Type: new Abstract: This paper presents an end-to-end, IoT-enabled robotic system for the non-destructive, real-time, and spatially-resolved mapping of grape yield and quality (Brix, Acidity) in vineyards. The system features a comprehensive analytical pipeline that integrates two key modules: a high-performance model for grape bunch detection and weight estimation, and a novel deep learning framework for quality assessment from hyperspectral (HSI) data. A critical barrier to in-field HSI is the ``domain shift" caused by variable illumination. To overcome this, our quality assessment is powered by the Light-Invariant Spectral Autoencoder (LISA), a domain-adversarial framework that learns illumination-invariant features from uncalibrated data. We validated the system's robustness on a purpose-built HSI dataset spanning three distinct illumination domains: controlled artificial lighting (lab), and variable natural sunlight captured in the morning and afternoon. Results show the complete pipeline achieves a recall (0.82) for bunch detection and a $R^2$ (0.76) for weight prediction, while the LISA module improves quality prediction generalization by over 20% compared to the baselines. By combining these robust modules, the system successfully generates high-resolution, georeferenced data of both grape yield and quality, providing actionable, data-driven insights for precision viticulture.

BenthiCat: An opti-acoustic dataset for advancing benthic classification and habitat mapping

arXiv:2510.04876v1 Announce Type: new Abstract: Benthic habitat mapping is fundamental for understanding marine ecosystems, guiding conservation efforts, and supporting sustainable resource management. Yet, the scarcity of large, annotated datasets limits the development and benchmarking of machine learning models in this domain. This paper introduces a thorough multi-modal dataset, comprising about a million side-scan sonar (SSS) tiles collected along the coast of Catalonia (Spain), complemented by bathymetric maps and a set of co-registered optical images from targeted surveys using an autonomous underwater vehicle (AUV). Approximately \num{36000} of the SSS tiles have been manually annotated with segmentation masks to enable supervised fine-tuning of classification models. All the raw sensor data, together with mosaics, are also released to support further exploration and algorithm development. To address challenges in multi-sensor data fusion for AUVs, we spatially associate optical images with corresponding SSS tiles, facilitating self-supervised, cross-modal representation learning. Accompanying open-source preprocessing and annotation tools are provided to enhance accessibility and encourage research. This resource aims to establish a standardized benchmark for underwater habitat mapping, promoting advancements in autonomous seafloor classification and multi-sensor integration.

Comparative Analysis of YOLOv5, Faster R-CNN, SSD, and RetinaNet for Motorbike Detection in Kigali Autonomous Driving Context

arXiv:2510.04912v1 Announce Type: new Abstract: In Kigali, Rwanda, motorcycle taxis are a primary mode of transportation, often navigating unpredictably and disregarding traffic rules, posing significant challenges for autonomous driving systems. This study compares four object detection models--YOLOv5, Faster R-CNN, SSD, and RetinaNet--for motorbike detection using a custom dataset of 198 images collected in Kigali. Implemented in PyTorch with transfer learning, the models were evaluated for accuracy, localization, and inference speed to assess their suitability for real-time navigation in resource-constrained settings. We identify implementation challenges, including dataset limitations and model complexities, and recommend simplified architectures for future work to enhance accessibility for autonomous systems in developing countries like Rwanda.

A Semantics-Aware Hierarchical Self-Supervised Approach to Classification of Remote Sensing Images

arXiv:2510.04916v1 Announce Type: new Abstract: Deep learning has become increasingly important in remote sensing image classification due to its ability to extract semantic information from complex data. Classification tasks often include predefined label hierarchies that represent the semantic relationships among classes. However, these hierarchies are frequently overlooked, and most approaches focus only on fine-grained classification schemes. In this paper, we present a novel Semantics-Aware Hierarchical Consensus (SAHC) method for learning hierarchical features and relationships by integrating hierarchy-specific classification heads within a deep network architecture, each specialized in different degrees of class granularity. The proposed approach employs trainable hierarchy matrices, which guide the network through the learning of the hierarchical structure in a self-supervised manner. Furthermore, we introduce a hierarchical consensus mechanism to ensure consistent probability distributions across different hierarchical levels. This mechanism acts as a weighted ensemble being able to effectively leverage the inherent structure of the hierarchical classification task. The proposed SAHC method is evaluated on three benchmark datasets with different degrees of hierarchical complexity on different tasks, using distinct backbone architectures to effectively emphasize its adaptability. Experimental results show both the effectiveness of the proposed approach in guiding network learning and the robustness of the hierarchical consensus for remote sensing image classification tasks.

REN: Anatomically-Informed Mixture-of-Experts for Interstitial Lung Disease Diagnosis

arXiv:2510.04923v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) architectures have significantly contributed to scalable machine learning by enabling specialized subnetworks to tackle complex tasks efficiently. However, traditional MoE systems lack domain-specific constraints essential for medical imaging, where anatomical structure and regional disease heterogeneity strongly influence pathological patterns. Here, we introduce Regional Expert Networks (REN), the first anatomically-informed MoE framework tailored specifically for medical image classification. REN leverages anatomical priors to train seven specialized experts, each dedicated to distinct lung lobes and bilateral lung combinations, enabling precise modeling of region-specific pathological variations. Multi-modal gating mechanisms dynamically integrate radiomics biomarkers and deep learning (DL) features (CNN, ViT, Mamba) to weight expert contributions optimally. Applied to interstitial lung disease (ILD) classification, REN achieves consistently superior performance: the radiomics-guided ensemble reached an average AUC of 0.8646 +/- 0.0467, a +12.5 percent improvement over the SwinUNETR baseline (AUC 0.7685, p = 0.031). Region-specific experts further revealed that lower-lobe models achieved AUCs of 0.88-0.90, surpassing DL counterparts (CNN: 0.76-0.79) and aligning with known disease progression patterns. Through rigorous patient-level cross-validation, REN demonstrates strong generalizability and clinical interpretability, presenting a scalable, anatomically-guided approach readily extensible to other structured medical imaging applications.

Unsupervised Active Learning via Natural Feature Progressive Framework

arXiv:2510.04939v1 Announce Type: new Abstract: The effectiveness of modern deep learning models is predicated on the availability of large-scale, human-annotated datasets, a process that is notoriously expensive and time-consuming. While Active Learning (AL) offers a strategic solution by labeling only the most informative and representative data, its iterative nature still necessitates significant human involvement. Unsupervised Active Learning (UAL) presents an alternative by shifting the annotation burden to a single, post-selection step. Unfortunately, prevailing UAL methods struggle to achieve state-of-the-art performance. These approaches typically rely on local, gradient-based scoring for sample importance estimation, which not only makes them vulnerable to ambiguous and noisy data but also hinders their capacity to select samples that adequately represent the full data distribution. Moreover, their use of shallow, one-shot linear selection falls short of a true UAL paradigm. In this paper, we propose the Natural Feature Progressive Framework (NFPF), a UAL method that revolutionizes how sample importance is measured. At its core, NFPF employs a Specific Feature Learning Machine (SFLM) to effectively quantify each sample's contribution to model performance. We further utilize the SFLM to define a powerful Reconstruction Difference metric for initial sample selection. Our comprehensive experiments show that NFPF significantly outperforms all established UAL methods and achieves performance on par with supervised AL methods on vision datasets. Detailed ablation studies and qualitative visualizations provide compelling evidence for NFPF's superior performance, enhanced robustness, and improved data distribution coverage.

Bidirectional Mammogram View Translation with Column-Aware and Implicit 3D Conditional Diffusion

arXiv:2510.04947v1 Announce Type: new Abstract: Dual-view mammography, including craniocaudal (CC) and mediolateral oblique (MLO) projections, offers complementary anatomical views crucial for breast cancer diagnosis. However, in real-world clinical workflows, one view may be missing, corrupted, or degraded due to acquisition errors or compression artifacts, limiting the effectiveness of downstream analysis. View-to-view translation can help recover missing views and improve lesion alignment. Unlike natural images, this task in mammography is highly challenging due to large non-rigid deformations and severe tissue overlap in X-ray projections, which obscure pixel-level correspondences. In this paper, we propose Column-Aware and Implicit 3D Diffusion (CA3D-Diff), a novel bidirectional mammogram view translation framework based on conditional diffusion model. To address cross-view structural misalignment, we first design a column-aware cross-attention mechanism that leverages the geometric property that anatomically corresponding regions tend to lie in similar column positions across views. A Gaussian-decayed bias is applied to emphasize local column-wise correlations while suppressing distant mismatches. Furthermore, we introduce an implicit 3D structure reconstruction module that back-projects noisy 2D latents into a coarse 3D feature volume based on breast-view projection geometry. The reconstructed 3D structure is refined and injected into the denoising UNet to guide cross-view generation with enhanced anatomical awareness. Extensive experiments demonstrate that CA3D-Diff achieves superior performance in bidirectional tasks, outperforming state-of-the-art methods in visual fidelity and structural consistency. Furthermore, the synthesized views effectively improve single-view malignancy classification in screening settings, demonstrating the practical value of our method in real-world diagnostics.

SSDD: Single-Step Diffusion Decoder for Efficient Image Tokenization

arXiv:2510.04961v1 Announce Type: new Abstract: Tokenizers are a key component of state-of-the-art generative image models, extracting the most important features from the signal while reducing data dimension and redundancy. Most current tokenizers are based on KL-regularized variational autoencoders (KL-VAE), trained with reconstruction, perceptual and adversarial losses. Diffusion decoders have been proposed as a more principled alternative to model the distribution over images conditioned on the latent. However, matching the performance of KL-VAE still requires adversarial losses, as well as a higher decoding time due to iterative sampling. To address these limitations, we introduce a new pixel diffusion decoder architecture for improved scaling and training stability, benefiting from transformer components and GAN-free training. We use distillation to replicate the performance of the diffusion decoder in an efficient single-step decoder. This makes SSDD the first diffusion decoder optimized for single-step reconstruction trained without adversarial losses, reaching higher reconstruction quality and faster sampling than KL-VAE. In particular, SSDD improves reconstruction FID from $0.87$ to $0.50$ with $1.4\times$ higher throughput and preserve generation quality of DiTs with $3.8\times$ faster sampling. As such, SSDD can be used as a drop-in replacement for KL-VAE, and for building higher-quality and faster generative models.

ActiveMark: on watermarking of visual foundation models via massive activations

arXiv:2510.04966v1 Announce Type: new Abstract: Being trained on large and vast datasets, visual foundation models (VFMs) can be fine-tuned for diverse downstream tasks, achieving remarkable performance and efficiency in various computer vision applications. The high computation cost of data collection and training motivates the owners of some VFMs to distribute them alongside the license to protect their intellectual property rights. However, a dishonest user of the protected model's copy may illegally redistribute it, for example, to make a profit. As a consequence, the development of reliable ownership verification tools is of great importance today, since such methods can be used to differentiate between a redistributed copy of the protected model and an independent model. In this paper, we propose an approach to ownership verification of visual foundation models by fine-tuning a small set of expressive layers of a VFM along with a small encoder-decoder network to embed digital watermarks into an internal representation of a hold-out set of input images. Importantly, the watermarks embedded remain detectable in the functional copies of the protected model, obtained, for example, by fine-tuning the VFM for a particular downstream task. Theoretically and experimentally, we demonstrate that the proposed method yields a low probability of false detection of a non-watermarked model and a low probability of false misdetection of a watermarked model.

Latent Uncertainty Representations for Video-based Driver Action and Intention Recognition

arXiv:2510.05006v1 Announce Type: new Abstract: Deep neural networks (DNNs) are increasingly applied to safety-critical tasks in resource-constrained environments, such as video-based driver action and intention recognition. While last layer probabilistic deep learning (LL-PDL) methods can detect out-of-distribution (OOD) instances, their performance varies. As an alternative to last layer approaches, we propose extending pre-trained DNNs with transformation layers to produce multiple latent representations to estimate the uncertainty. We evaluate our latent uncertainty representation (LUR) and repulsively trained LUR (RLUR) approaches against eight PDL methods across four video-based driver action and intention recognition datasets, comparing classification performance, calibration, and uncertainty-based OOD detection. We also contribute 28,000 frame-level action labels and 1,194 video-level intention labels for the NuScenes dataset. Our results show that LUR and RLUR achieve comparable in-distribution classification performance to other LL-PDL approaches. For uncertainty-based OOD detection, LUR matches top-performing PDL methods while being more efficient to train and easier to tune than approaches that require Markov-Chain Monte Carlo sampling or repulsive training procedures.

Exploring the Efficacy of Modified Transfer Learning in Identifying Parkinson's Disease Through Drawn Image Patterns

arXiv:2510.05015v1 Announce Type: new Abstract: Parkinson's disease (PD) is a progressive neurodegenerative condition characterized by the death of dopaminergic neurons, leading to various movement disorder symptoms. Early diagnosis of PD is crucial to prevent adverse effects, yet traditional diagnostic methods are often cumbersome and costly. In this study, a machine learning-based approach is proposed using hand-drawn spiral and wave images as potential biomarkers for PD detection. Our methodology leverages convolutional neural networks (CNNs), transfer learning, and attention mechanisms to improve model performance and resilience against overfitting. To enhance the diversity and richness of both spiral and wave categories, the training dataset undergoes augmentation to increase the number of images. The proposed architecture comprises three phases: utilizing pre-trained CNNs, incorporating custom convolutional layers, and ensemble voting. Employing hard voting further enhances performance by aggregating predictions from multiple models. Experimental results show promising accuracy rates. For spiral images, weighted average precision, recall, and F1-score are 90%, and for wave images, they are 96.67%. After combining the predictions through ensemble hard voting, the overall accuracy is 93.3%. These findings underscore the potential of machine learning in early PD diagnosis, offering a non-invasive and cost-effective solution to improve patient outcomes.

Video-LMM Post-Training: A Deep Dive into Video Reasoning with Large Multimodal Models

arXiv:2510.05034v1 Announce Type: new Abstract: Video understanding represents the most challenging frontier in computer vision, requiring models to reason about complex spatiotemporal relationships, long-term dependencies, and multimodal evidence. The recent emergence of Video-Large Multimodal Models (Video-LMMs), which integrate visual encoders with powerful decoder-based language models, has demonstrated remarkable capabilities in video understanding tasks. However, the critical phase that transforms these models from basic perception systems into sophisticated reasoning engines, post-training, remains fragmented across the literature. This survey provides the first comprehensive examination of post-training methodologies for Video-LMMs, encompassing three fundamental pillars: supervised fine-tuning (SFT) with chain-of-thought, reinforcement learning (RL) from verifiable objectives, and test-time scaling (TTS) through enhanced inference computation. We present a structured taxonomy that clarifies the roles, interconnections, and video-specific adaptations of these techniques, addressing unique challenges such as temporal localization, spatiotemporal grounding, long video efficiency, and multimodal evidence integration. Through systematic analysis of representative methods, we synthesize key design principles, insights, and evaluation protocols while identifying critical open challenges in reward design, scalability, and cost-performance optimization. We further curate essential benchmarks, datasets, and metrics to facilitate rigorous assessment of post-training effectiveness. This survey aims to provide researchers and practitioners with a unified framework for advancing Video-LMM capabilities. Additional resources and updates are maintained at: https://github.com/yunlong10/Awesome-Video-LMM-Post-Training

SegMASt3R: Geometry Grounded Segment Matching

arXiv:2510.05051v1 Announce Type: new Abstract: Segment matching is an important intermediate task in computer vision that establishes correspondences between semantically or geometrically coherent regions across images. Unlike keypoint matching, which focuses on localized features, segment matching captures structured regions, offering greater robustness to occlusions, lighting variations, and viewpoint changes. In this paper, we leverage the spatial understanding of 3D foundation models to tackle wide-baseline segment matching, a challenging setting involving extreme viewpoint shifts. We propose an architecture that uses the inductive bias of these 3D foundation models to match segments across image pairs with up to 180 degree view-point change. Extensive experiments show that our approach outperforms state-of-the-art methods, including the SAM2 video propagator and local feature matching methods, by upto 30% on the AUPRC metric, on ScanNet++ and Replica datasets. We further demonstrate benefits of the proposed model on relevant downstream tasks, including 3D instance segmentation and image-goal navigation. Project Page: https://segmast3r.github.io/

No-reference Quality Assessment of Contrast-distorted Images using Contrast-enhanced Pseudo Reference

arXiv:2510.05053v1 Announce Type: new Abstract: Contrast change is an important factor that affects the quality of images. During image capturing, unfavorable lighting conditions can cause contrast change and visual quality loss. While various methods have been proposed to assess the quality of images under different distortions such as blur and noise, contrast distortion has been largely overlooked as its visual impact and properties are different from other conventional types of distortions. In this paper, we propose a no-reference image quality assessment (NR-IQA) metric for contrast-distorted images. Using a set of contrast enhancement algorithms, we aim to generate pseudo-reference images that are visually close to the actual reference image, such that the NR problem is transformed to a Full-reference (FR) assessment with higher accuracy. To this end, a large dataset of contrast-enhanced images is produced to train a classification network that can select the most suitable contrast enhancement algorithm based on image content and distortion for pseudo-reference image generation. Finally, the evaluation is performed in the FR manner to assess the quality difference between the contrast-enhanced (pseudoreference) and degraded images. Performance evaluation of the proposed method on three databases containing contrast distortions (CCID2014, TID2013, and CSIQ), indicates the promising performance of the proposed method.

Neuroplastic Modular Framework: Cross-Domain Image Classification of Garbage and Industrial Surfaces

arXiv:2510.05071v1 Announce Type: new Abstract: Efficient and accurate classification of waste and industrial surface defects is essential for ensuring sustainable waste management and maintaining high standards in quality control. This paper introduces the Neuroplastic Modular Classifier, a novel hybrid architecture designed for robust and adaptive image classification in dynamic environments. The model combines a ResNet-50 backbone for localized feature extraction with a Vision Transformer (ViT) to capture global semantic context. Additionally, FAISS-based similarity retrieval is incorporated to provide a memory-like reference to previously encountered data, enriching the model's feature space. A key innovation of our architecture is the neuroplastic modular design composed of expandable, learnable blocks that dynamically grow during training when performance plateaus. Inspired by biological learning systems, this mechanism allows the model to adapt to data complexity over time, improving generalization. Beyond garbage classification, we validate the model on the Kolektor Surface Defect Dataset 2 (KolektorSDD2), which involves industrial defect detection on metal surfaces. Experimental results across domains show that the proposed architecture outperforms traditional static models in both accuracy and adaptability. The Neuroplastic Modular Classifier offers a scalable, high-performance solution for real-world image classification, with strong applicability in both environmental and industrial domains.

Factuality Matters: When Image Generation and Editing Meet Structured Visuals

arXiv:2510.05091v1 Announce Type: new Abstract: While modern visual generation models excel at creating aesthetically pleasing natural images, they struggle with producing or editing structured visuals like charts, diagrams, and mathematical figures, which demand composition planning, text rendering, and multimodal reasoning for factual fidelity. To address this, we present the first comprehensive, systematic investigation of this domain, encompassing data construction, model training, and an evaluation benchmark. First, we construct a large-scale dataset of 1.3 million high-quality structured image pairs derived from executable drawing programs and augmented with chain-of-thought reasoning annotations. Building on it, we train a unified model that integrates a VLM with FLUX.1 Kontext via a lightweight connector for enhanced multimodal understanding. A three-stage training curriculum enables progressive feature alignment, knowledge infusion, and reasoning-augmented generation, further boosted by an external reasoner at inference time. Finally, we introduce StructBench, a novel benchmark for generation and editing with over 1,700 challenging instances, and an accompanying evaluation metric, StructScore, which employs a multi-round Q\&A protocol to assess fine-grained factual accuracy. Evaluations of 15 models reveal that even leading closed-source systems remain far from satisfactory. Our model attains strong editing performance, and inference-time reasoning yields consistent gains across diverse architectures. By releasing the dataset, model, and benchmark, we aim to advance unified multimodal foundations for structured visuals.

Character Mixing for Video Generation

arXiv:2510.05093v1 Announce Type: new Abstract: Imagine Mr. Bean stepping into Tom and Jerry--can we generate videos where characters interact naturally across different worlds? We study inter-character interaction in text-to-video generation, where the key challenge is to preserve each character's identity and behaviors while enabling coherent cross-context interaction. This is difficult because characters may never have coexisted and because mixing styles often causes style delusion, where realistic characters appear cartoonish or vice versa. We introduce a framework that tackles these issues with Cross-Character Embedding (CCE), which learns identity and behavioral logic across multimodal sources, and Cross-Character Augmentation (CCA), which enriches training with synthetic co-existence and mixed-style data. Together, these techniques allow natural interactions between previously uncoexistent characters without losing stylistic fidelity. Experiments on a curated benchmark of cartoons and live-action series with 10 characters show clear improvements in identity preservation, interaction quality, and robustness to style delusion, enabling new forms of generative storytelling.Additional results and videos are available on our project page: https://tingtingliao.github.io/mimix/.

VChain: Chain-of-Visual-Thought for Reasoning in Video Generation

arXiv:2510.05094v1 Announce Type: new Abstract: Recent video generation models can produce smooth and visually appealing clips, but they often struggle to synthesize complex dynamics with a coherent chain of consequences. Accurately modeling visual outcomes and state transitions over time remains a core challenge. In contrast, large language and multimodal models (e.g., GPT-4o) exhibit strong visual state reasoning and future prediction capabilities. To bridge these strengths, we introduce VChain, a novel inference-time chain-of-visual-thought framework that injects visual reasoning signals from multimodal models into video generation. Specifically, VChain contains a dedicated pipeline that leverages large multimodal models to generate a sparse set of critical keyframes as snapshots, which are then used to guide the sparse inference-time tuning of a pre-trained video generator only at these key moments. Our approach is tuning-efficient, introduces minimal overhead and avoids dense supervision. Extensive experiments on complex, multi-step scenarios show that VChain significantly enhances the quality of generated videos.

Paper2Video: Automatic Video Generation from Scientific Papers

arXiv:2510.05096v1 Announce Type: new Abstract: Academic presentation videos have become an essential medium for research communication, yet producing them remains highly labor-intensive, often requiring hours of slide design, recording, and editing for a short 2 to 10 minutes video. Unlike natural video, presentation video generation involves distinctive challenges: inputs from research papers, dense multi-modal information (text, figures, tables), and the need to coordinate multiple aligned channels such as slides, subtitles, speech, and human talker. To address these challenges, we introduce PaperTalker, the first benchmark of 101 research papers paired with author-created presentation videos, slides, and speaker metadata. We further design four tailored evaluation metrics--Meta Similarity, PresentArena, PresentQuiz, and IP Memory--to measure how videos convey the paper's information to the audience. Building on this foundation, we propose PaperTalker, the first multi-agent framework for academic presentation video generation. It integrates slide generation with effective layout refinement by a novel effective tree search visual choice, cursor grounding, subtitling, speech synthesis, and talking-head rendering, while parallelizing slide-wise generation for efficiency. Experiments on Paper2Video demonstrate that the presentation videos produced by our approach are more faithful and informative than existing baselines, establishing a practical step toward automated and ready-to-use academic video generation. Our dataset, agent, and code are available at https://github.com/showlab/Paper2Video.

VIFO: Visual Feature Empowered Multivariate Time Series Forecasting with Cross-Modal Fusion

arXiv:2510.03244v1 Announce Type: cross Abstract: Large time series foundation models often adopt channel-independent architectures to handle varying data dimensions, but this design ignores crucial cross-channel dependencies. Concurrently, existing multimodal approaches have not fully exploited the power of large vision models (LVMs) to interpret spatiotemporal data. Additionally, there remains significant unexplored potential in leveraging the advantages of information extraction from different modalities to enhance time series forecasting performance. To address these gaps, we propose the VIFO, a cross-modal forecasting model. VIFO uniquely renders multivariate time series into image, enabling pre-trained LVM to extract complex cross-channel patterns that are invisible to channel-independent models. These visual features are then aligned and fused with representations from the time series modality. By freezing the LVM and training only 7.45% of its parameters, VIFO achieves competitive performance on multiple benchmarks, offering an efficient and effective solution for capturing cross-variable relationships in

Frequency-Aware Model Parameter Explorer: A new attribution method for improving explainability

arXiv:2510.03245v1 Announce Type: cross Abstract: Ensuring the reliability of deep neural networks (DNNs) in the presence of real world noise and intentional perturbations remains a significant challenge. To address this, attribution methods have been proposed, though their efficacy remains suboptimal and necessitates further refinement. In this paper, we propose a novel category of transferable adversarial attacks, called transferable frequency-aware attacks, enabling frequency-aware exploration via both high-and low-frequency components. Based on this type of attacks, we also propose a novel attribution method, named Frequency-Aware Model Parameter Explorer (FAMPE), which improves the explainability for DNNs. Relative to the current state-of-the-art method AttEXplore, our FAMPE attains an average gain of 13.02% in Insertion Score, thereby outperforming existing approaches. Through detailed ablation studies, we also investigate the role of both high- and low-frequency components in explainability.

Real-Time Brain Biomechanics Prediction with Neural Operators: Toward Clinically Deployable Traumatic Brain Injury Models

arXiv:2510.03248v1 Announce Type: cross Abstract: Traumatic brain injury (TBI) remains a major public health concern, with over 69 million cases annually worldwide. Finite element (FE) models offer high-fidelity predictions of brain deformation but are computationally expensive, requiring hours per simulation and limiting their clinical utility for rapid decision-making. This study benchmarks state-of-the-art neural operator (NO) architectures for rapid, patient-specific prediction of brain displacement fields, aiming to enable real-time TBI modeling in clinical and translational settings. We formulated TBI modeling as an operator learning problem, mapping subject-specific anatomical MRI, magnetic resonance elastography (MRE) stiffness maps, and demographic features to full-field 3D brain displacement predictions. Four architectures - Fourier Neural Operator (FNO), Factorized FNO (F-FNO), Multi-Grid FNO (MG-FNO), and Deep Operator Network (DeepONet) were trained and evaluated on 249 MRE datasets across physiologically relevant frequencies (20 - 90 Hz). MG-FNO achieved the highest accuracy (MSE = 0.0023, 94.3\% spatial fidelity) and preserved fine-scale features, while F-FNO converged 2$\times$ faster than standard FNO. DeepONet offered the fastest inference (14.5 iterations/s) with a 7$\times$ computational speed-up over MG-FNO, suggesting utility for embedded or edge computing applications. All NOs reduced computation time from hours to milliseconds without sacrificing anatomical realism. NOs provide an efficient, resolution-invariant approach for predicting brain deformation, opening the door to real-time, patient-specific TBI risk assessment, clinical triage support, and optimization of protective equipment. These results highlight the potential for NO-based digital twins of the human brain, enabling scalable, on-demand biomechanical modeling in both clinical and population health contexts.

Universal Multi-Domain Translation via Diffusion Routers

arXiv:2510.03252v1 Announce Type: cross Abstract: Multi-domain translation (MDT) aims to learn translations between multiple domains, yet existing approaches either require fully aligned tuples or can only handle domain pairs seen in training, limiting their practicality and excluding many cross-domain mappings. We introduce universal MDT (UMDT), a generalization of MDT that seeks to translate between any pair of $K$ domains using only $K-1$ paired datasets with a central domain. To tackle this problem, we propose Diffusion Router (DR), a unified diffusion-based framework that models all central$\leftrightarrow$non-central translations with a single noise predictor conditioned on the source and target domain labels. DR enables indirect non-central translations by routing through the central domain. We further introduce a novel scalable learning strategy with a variational-bound objective and an efficient Tweedie refinement procedure to support direct non-central mappings. Through evaluation on three large-scale UMDT benchmarks, DR achieves state-of-the-art results for both indirect and direct translations, while lowering sampling cost and unlocking novel tasks such as sketch$\leftrightarrow$segmentation. These results establish DR as a scalable and versatile framework for universal translation across multiple domains.

Rethinking Inter-LoRA Orthogonality in Adapter Merging: Insights from Orthogonal Monte Carlo Dropout

arXiv:2510.03262v1 Announce Type: cross Abstract: We propose Orthogonal Monte Carlo Dropout, a mechanism that enforces strict orthogonality when combining sparse semantic vectors without extra time complexity. LoRA, a popular fine-tuning method for large models, typically trains a module to represent a specific concept such as an object or a style. When multiple LoRAs are merged, for example to generate an object in a particular style, their semantic vectors may interfere with each other. Our method guarantees, at the theoretical and runtime levels, that merged LoRAs remain orthogonal and thus free from direct interference. However, empirical analysis reveals that such orthogonality does not lead to the semantic disentanglement or compositionality highlighted in prior work on compositional adaptation. This finding suggests that inter-LoRA orthogonality alone may be insufficient for achieving true semantic compositionality, prompting a re-examination of its role in adapter merging.

SDQ-LLM: Sigma-Delta Quantization for 1-bit LLMs of any size

arXiv:2510.03275v1 Announce Type: cross Abstract: Large language models (LLMs) face significant computational and memory challenges, making extremely low-bit quantization crucial for their efficient deployment. In this work, we introduce SDQ-LLM: Sigma-Delta Quantization for 1-bit LLMs of any size, a novel framework that enables extremely low-bit quantization of LLMs while preserving their linguistic reasoning capabilities. A distinctive feature of SDQ-LLM is the continuous adjustability of the Over-Sampling Ratio (OSR), enabling dynamic adaptation to memory or VRAM constraints by selecting fractional OSR (e.g. 2.5 times) for an optimal trade-off between model size and accuracy. SDQ-LLM uses upsampling combined with Sigma-Delta Quantizer to binarize or ternarize LLMs weights, encoding high-precision parameters into 1-bit or 1.58-bit representations, replacing the multiplication operations within linear layers with addition. This approach significantly enhances inference efficiency under extremely low-bit quantization. To further reduce the loss of quantization precision, we incorporate Hadamard-based weight smoothing prior to quantization, improving the stability and robustness of the weight representations. Furthermore, to fully leverage the continuity of the OSR and reduce precision loss, recognizing the correlation between quantization sensitivity and weight variance, we propose a fine-grained, layer- and linear-wise OSR allocation strategy, MultiOSR. This strategy distributes OSR both across layers and within each layer, based on weight variance and parameter scale. Finally, extensive experiments on OPT and LLaMA model families demonstrate that SDQ-LLM achieves a more efficient and high-precision performance even under highly aggressive low-OSR settings. Our code is available at https://github.com/Dreamlittlecat/LLM-Quant-Factory.

Revoking Amnesia: RL-based Trajectory Optimization to Resurrect Erased Concepts in Diffusion Models

arXiv:2510.03302v1 Announce Type: cross Abstract: Concept erasure techniques have been widely deployed in T2I diffusion models to prevent inappropriate content generation for safety and copyright considerations. However, as models evolve to next-generation architectures like Flux, established erasure methods (\textit{e.g.}, ESD, UCE, AC) exhibit degraded effectiveness, raising questions about their true mechanisms. Through systematic analysis, we reveal that concept erasure creates only an illusion of ``amnesia": rather than genuine forgetting, these methods bias sampling trajectories away from target concepts, making the erasure fundamentally reversible. This insight motivates the need to distinguish superficial safety from genuine concept removal. In this work, we propose \textbf{RevAm} (\underline{Rev}oking \underline{Am}nesia), an RL-based trajectory optimization framework that resurrects erased concepts by dynamically steering the denoising process without modifying model weights. By adapting Group Relative Policy Optimization (GRPO) to diffusion models, RevAm explores diverse recovery trajectories through trajectory-level rewards, overcoming local optima that limit existing methods. Extensive experiments demonstrate that RevAm achieves superior concept resurrection fidelity while reducing computational time by 10$\times$, exposing critical vulnerabilities in current safety mechanisms and underscoring the need for more robust erasure techniques beyond trajectory manipulation.

Creative synthesis of kinematic mechanisms

arXiv:2510.03308v1 Announce Type: cross Abstract: In this paper, we formulate the problem of kinematic synthesis for planar linkages as a cross-domain image generation task. We develop a planar linkages dataset using RGB image representations, covering a range of mechanisms: from simple types such as crank-rocker and crank-slider to more complex eight-bar linkages like Jansen's mechanism. A shared-latent variational autoencoder (VAE) is employed to explore the potential of image generative models for synthesizing unseen motion curves and simulating novel kinematics. By encoding the drawing speed of trajectory points as color gradients, the same architecture also supports kinematic synthesis conditioned on both trajectory shape and velocity profiles. We validate our method on three datasets of increasing complexity: a standard four-bar linkage set, a mixed set of four-bar and crank-slider mechanisms, and a complex set including multi-loop mechanisms. Preliminary results demonstrate the effectiveness of image-based representations for generative mechanical design, showing that mechanisms with revolute and prismatic joints, and potentially cams and gears, can be represented and synthesized within a unified image generation framework.

Universal Beta Splatting

arXiv:2510.03312v1 Announce Type: cross Abstract: We introduce Universal Beta Splatting (UBS), a unified framework that generalizes 3D Gaussian Splatting to N-dimensional anisotropic Beta kernels for explicit radiance field rendering. Unlike fixed Gaussian primitives, Beta kernels enable controllable dependency modeling across spatial, angular, and temporal dimensions within a single representation. Our unified approach captures complex light transport effects, handles anisotropic view-dependent appearance, and models scene dynamics without requiring auxiliary networks or specific color encodings. UBS maintains backward compatibility by approximating to Gaussian Splatting as a special case, guaranteeing plug-in usability and lower performance bounds. The learned Beta parameters naturally decompose scene properties into interpretable without explicit supervision: spatial (surface vs. texture), angular (diffuse vs. specular), and temporal (static vs. dynamic). Our CUDA-accelerated implementation achieves real-time rendering while consistently outperforming existing methods across static, view-dependent, and dynamic benchmarks, establishing Beta kernels as a scalable universal primitive for radiance field rendering. Our project website is available at https://rongliu-leo.github.io/universal-beta-splatting/.

Real-time nonlinear inversion of magnetic resonance elastography with operator learning

arXiv:2510.03372v1 Announce Type: cross Abstract: $\textbf{Purpose:}$ To develop and evaluate an operator learning framework for nonlinear inversion (NLI) of brain magnetic resonance elastography (MRE) data, which enables real-time inversion of elastograms with comparable spatial accuracy to NLI. $\textbf{Materials and Methods:}$ In this retrospective study, 3D MRE data from 61 individuals (mean age, 37.4 years; 34 female) were used for development of the framework. A predictive deep operator learning framework (oNLI) was trained using 10-fold cross-validation, with the complex curl of the measured displacement field as inputs and NLI-derived reference elastograms as outputs. A structural prior mechanism, analogous to Soft Prior Regularization in the MRE literature, was incorporated to improve spatial accuracy. Subject-level evaluation metrics included Pearson's correlation coefficient, absolute relative error, and structural similarity index measure between predicted and reference elastograms across brain regions of different sizes to understand accuracy. Statistical analyses included paired t-tests comparing the proposed oNLI variants to the convolutional neural network baselines. $\textbf{Results:}$ Whole brain absolute percent error was 8.4 $\pm$ 0.5 ($\mu'$) and 10.0 $\pm$ 0.7 ($\mu''$) for oNLI and 15.8 $\pm$ 0.8 ($\mu'$) and 26.1 $\pm$ 1.1 ($\mu''$) for CNNs. Additionally, oNLI outperformed convolutional architectures as per Pearson's correlation coefficient, $r$, in the whole brain and across all subregions for both the storage modulus and loss modulus (p < 0.05). $\textbf{Conclusion:}$ The oNLI framework enables real-time MRE inversion (30,000x speedup), outperforming CNN-based approaches and maintaining the fine-grained spatial accuracy achievable with NLI in the brain.

Conditional Pseudo-Supervised Contrast for Data-Free Knowledge Distillation

arXiv:2510.03375v1 Announce Type: cross Abstract: Data-free knowledge distillation~(DFKD) is an effective manner to solve model compression and transmission restrictions while retaining privacy protection, which has attracted extensive attention in recent years. Currently, the majority of existing methods utilize a generator to synthesize images to support the distillation. Although the current methods have achieved great success, there are still many issues to be explored. Firstly, the outstanding performance of supervised learning in deep learning drives us to explore a pseudo-supervised paradigm on DFKD. Secondly, current synthesized methods cannot distinguish the distributions of different categories of samples, thus producing ambiguous samples that may lead to an incorrect evaluation by the teacher. Besides, current methods cannot optimize the category-wise diversity samples, which will hinder the student model learning from diverse samples and further achieving better performance. In this paper, to address the above limitations, we propose a novel learning paradigm, i.e., conditional pseudo-supervised contrast for data-free knowledge distillation~(CPSC-DFKD). The primary innovations of CPSC-DFKD are: (1) introducing a conditional generative adversarial network to synthesize category-specific diverse images for pseudo-supervised learning, (2) improving the modules of the generator to distinguish the distributions of different categories, and (3) proposing pseudo-supervised contrastive learning based on teacher and student views to enhance diversity. Comprehensive experiments on three commonly-used datasets validate the performance lift of both the student and generator brought by CPSC-DFKD. The code is available at https://github.com/RoryShao/CPSC-DFKD.git

Efficient Surgical Robotic Instrument Pose Reconstruction in Real World Conditions Using Unified Feature Detection

arXiv:2510.03532v1 Announce Type: cross Abstract: Accurate camera-to-robot calibration is essential for any vision-based robotic control system and especially critical in minimally invasive surgical robots, where instruments conduct precise micro-manipulations. However, MIS robots have long kinematic chains and partial visibility of their degrees of freedom in the camera, which introduces challenges for conventional camera-to-robot calibration methods that assume stiff robots with good visibility. Previous works have investigated both keypoint-based and rendering-based approaches to address this challenge in real-world conditions; however, they often struggle with consistent feature detection or have long inference times, neither of which are ideal for online robot control. In this work, we propose a novel framework that unifies the detection of geometric primitives (keypoints and shaft edges) through a shared encoding, enabling efficient pose estimation via projection geometry. This architecture detects both keypoints and edges in a single inference and is trained on large-scale synthetic data with projective labeling. This method is evaluated across both feature detection and pose estimation, with qualitative and quantitative results demonstrating fast performance and state-of-the-art accuracy in challenging surgical environments.

How We Won BraTS-SSA 2025: Brain Tumor Segmentation in the Sub-Saharan African Population Using Segmentation-Aware Data Augmentation and Model Ensembling

arXiv:2510.03568v1 Announce Type: cross Abstract: Brain tumors, particularly gliomas, pose significant chall-enges due to their complex growth patterns, infiltrative nature, and the variability in brain structure across individuals, which makes accurate diagnosis and monitoring difficult. Deep learning models have been developed to accurately delineate these tumors. However, most of these models were trained on relatively homogenous high-resource datasets, limiting their robustness when deployed in underserved regions. In this study, we performed segmentation-aware offline data augmentation on the BraTS-Africa dataset to increase the data sample size and diversity to enhance generalization. We further constructed an ensemble of three distinct architectures, MedNeXt, SegMamba, and Residual-Encoder U-Net, to leverage their complementary strengths. Our best-performing model, MedNeXt, was trained on 1000 epochs and achieved the highest average lesion-wise dice and normalized surface distance scores of 0.86 and 0.81 respectively. However, the ensemble model trained for 500 epochs produced the most balanced segmentation performance across the tumour subregions. This work demonstrates that a combination of advanced augmentation and model ensembling can improve segmentation accuracy and robustness on diverse and underrepresented datasets. Code available at: https://github.com/SPARK-Academy-2025/SPARK-2025/tree/main/SPARK2025_BraTs_MODELS/SPARK_NeuroAshanti

Longitudinal Flow Matching for Trajectory Modeling

arXiv:2510.03569v1 Announce Type: cross Abstract: Generative models for sequential data often struggle with sparsely sampled and high-dimensional trajectories, typically reducing the learning of dynamics to pairwise transitions. We propose \textit{Interpolative Multi-Marginal Flow Matching} (IMMFM), a framework that learns continuous stochastic dynamics jointly consistent with multiple observed time points. IMMFM employs a piecewise-quadratic interpolation path as a smooth target for flow matching and jointly optimizes drift and a data-driven diffusion coefficient, supported by a theoretical condition for stable learning. This design captures intrinsic stochasticity, handles irregular sparse sampling, and yields subject-specific trajectories. Experiments on synthetic benchmarks and real-world longitudinal neuroimaging datasets show that IMMFM outperforms existing methods in both forecasting accuracy and further downstream tasks.

Efficient Test-Time Scaling for Small Vision-Language Models

arXiv:2510.03574v1 Announce Type: cross Abstract: Small Vision-Language Models (VLMs) provide a computationally efficient alternative to larger models, at the cost of weaker generalization abilities and downstream task performance. These shortcomings could be addressed by test-time scaling techniques, but existing methods are typically computationally demanding, contradicting the resource-efficient design goals of small models. To address these limitations, we propose two novel and efficient test-time scaling strategies that leverage the model-internal features rather than external supervision: (i) Test-Time Augmentation (TTAug), which generates multiple augmented inputs and aggregates outputs at the token level without parameter updates, and (ii) Test-Time Adaptation (TTAdapt), which adapts model parameters during inference using consensus-based pseudolabels from TTAug. Through extensive experiments across nine benchmarks, we demonstrate consistent performance improvements while maintaining computational efficiency suitable for resource-constrained environments. The generality of our approach is demonstrated both within models at different scales and across different VLMs without additional tuning.

Model-Guided Microstimulation Steers Primate Visual Behavior

arXiv:2510.03684v1 Announce Type: cross Abstract: Brain stimulation is a powerful tool for understanding cortical function and holds promise for therapeutic interventions in neuropsychiatric disorders. Initial visual prosthetics apply electric microstimulation to early visual cortex which can evoke percepts of simple symbols such as letters. However, these approaches are fundamentally limited by hardware constraints and the low-level representational properties of this cortical region. In contrast, higher-level visual areas encode more complex object representations and therefore constitute a promising target for stimulation - but determining representational targets that reliably evoke object-level percepts constitutes a major challenge. We here introduce a computational framework to causally model and guide stimulation of high-level cortex, comprising three key components: (1) a perturbation module that translates microstimulation parameters into spatial changes to neural activity, (2) topographic models that capture the spatial organization of cortical neurons and thus enable prototyping of stimulation experiments, and (3) a mapping procedure that links model-optimized stimulation sites back to primate cortex. Applying this framework in two macaque monkeys performing a visual recognition task, model-predicted stimulation experiments produced significant in-vivo changes in perceptual choices. Per-site model predictions and monkey behavior were strongly correlated, underscoring the promise of model-guided stimulation. Image generation further revealed a qualitative similarity between in-silico stimulation of face-selective sites and a patient's report of facephenes. This proof-of-principle establishes a foundation for model-guided microstimulation and points toward next-generation visual prosthetics capable of inducing more complex visual experiences.

EmbodiSwap for Zero-Shot Robot Imitation Learning

arXiv:2510.03706v1 Announce Type: cross Abstract: We introduce EmbodiSwap - a method for producing photorealistic synthetic robot overlays over human video. We employ EmbodiSwap for zero-shot imitation learning, bridging the embodiment gap between in-the-wild ego-centric human video and a target robot embodiment. We train a closed-loop robot manipulation policy over the data produced by EmbodiSwap. We make novel use of V-JEPA as a visual backbone, repurposing V-JEPA from the domain of video understanding to imitation learning over synthetic robot videos. Adoption of V-JEPA outperforms alternative vision backbones more conventionally used within robotics. In real-world tests, our zero-shot trained V-JEPA model achieves an $82\%$ success rate, outperforming a few-shot trained $\pi_0$ network as well as $\pi_0$ trained over data produced by EmbodiSwap. We release (i) code for generating the synthetic robot overlays which takes as input human videos and an arbitrary robot URDF and generates a robot dataset, (ii) the robot dataset we synthesize over EPIC-Kitchens, HOI4D and Ego4D, and (iii) model checkpoints and inference code, to facilitate reproducible research and broader adoption.

Bridging the Gap Between Multimodal Foundation Models and World Models

arXiv:2510.03727v1 Announce Type: cross Abstract: Humans understand the world through the integration of multiple sensory modalities, enabling them to perceive, reason about, and imagine dynamic physical processes. Inspired by this capability, multimodal foundation models (MFMs) have emerged as powerful tools for multimodal understanding and generation. However, today's MFMs fall short of serving as effective world models. They lack the essential ability such as perform counterfactual reasoning, simulate dynamics, understand the spatiotemporal information, control generated visual outcomes, and perform multifaceted reasoning. We investigates what it takes to bridge the gap between multimodal foundation models and world models. We begin by improving the reasoning capabilities of MFMs through discriminative tasks and equipping MFMs with structured reasoning skills, such as causal inference, counterfactual thinking, and spatiotemporal reasoning, enabling them to go beyond surface correlations and understand deeper relationships within visual and textual data. Next, we explore generative capabilities of multimodal foundation models across both image and video modalities, introducing new frameworks for structured and controllable generation. Our approaches incorporate scene graphs, multimodal conditioning, and multimodal alignment strategies to guide the generation process, ensuring consistency with high-level semantics and fine-grained user intent. We further extend these techniques to controllable 4D generation, enabling interactive, editable, and morphable object synthesis over time and space.

Diverse Text-to-Image Generation via Contrastive Noise Optimization

arXiv:2510.03813v1 Announce Type: cross Abstract: Text-to-image (T2I) diffusion models have demonstrated impressive performance in generating high-fidelity images, largely enabled by text-guided inference. However, this advantage often comes with a critical drawback: limited diversity, as outputs tend to collapse into similar modes under strong text guidance. Existing approaches typically optimize intermediate latents or text conditions during inference, but these methods deliver only modest gains or remain sensitive to hyperparameter tuning. In this work, we introduce Contrastive Noise Optimization, a simple yet effective method that addresses the diversity issue from a distinct perspective. Unlike prior techniques that adapt intermediate latents, our approach shapes the initial noise to promote diverse outputs. Specifically, we develop a contrastive loss defined in the Tweedie data space and optimize a batch of noise latents. Our contrastive optimization repels instances within the batch to maximize diversity while keeping them anchored to a reference sample to preserve fidelity. We further provide theoretical insights into the mechanism of this preprocessing to substantiate its effectiveness. Extensive experiments across multiple T2I backbones demonstrate that our approach achieves a superior quality-diversity Pareto frontier while remaining robust to hyperparameter choices.

Towards Robust and Generalizable Continuous Space-Time Video Super-Resolution with Events

arXiv:2510.03833v1 Announce Type: cross Abstract: Continuous space-time video super-resolution (C-STVSR) has garnered increasing interest for its capability to reconstruct high-resolution and high-frame-rate videos at arbitrary spatial and temporal scales. However, prevailing methods often generalize poorly, producing unsatisfactory results when applied to out-of-distribution (OOD) scales. To overcome this limitation, we present EvEnhancer, a novel approach that marries the unique properties of high temporal resolution and high dynamic range encapsulated in event streams to achieve robust and generalizable C-STVSR. Our approach incorporates event-adapted synthesis that capitalizes on the spatiotemporal correlations between frames and events to capture long-term motion trajectories, enabling adaptive interpolation and fusion across space and time. This is then coupled with a local implicit video transformer that integrates local implicit video neural function with cross-scale spatiotemporal attention to learn continuous video representations and generate plausible videos at arbitrary resolutions and frame rates. We further develop EvEnhancerPlus, which builds a controllable switching mechanism that dynamically determines the reconstruction difficulty for each spatiotemporal pixel based on local event statistics. This allows the model to adaptively route reconstruction along the most suitable pathways at a fine-grained pixel level, substantially reducing computational overhead while maintaining excellent performance. Furthermore, we devise a cross-derivative training strategy that stabilizes the convergence of such a multi-pathway framework through staged cross-optimization. Extensive experiments demonstrate that our method achieves state-of-the-art performance on both synthetic and real-world datasets, while maintaining superior generalizability at OOD scales. The code is available at https://github.com/W-Shuoyan/EvEnhancerPlus.

Joint Neural SDF Reconstruction and Semantic Segmentation for CAD Models

arXiv:2510.03837v1 Announce Type: cross Abstract: We propose a simple, data-efficient pipeline that augments an implicit reconstruction network based on neural SDF-based CAD parts with a part-segmentation head trained under PartField-generated supervision. Unlike methods tied to fixed taxonomies, our model accepts meshes with any number of parts and produces coherent, geometry-aligned labels in a single pass. We evaluate on randomly sampled CAD meshes from the ABC dataset with intentionally varied part cardinalities, including over-segmented shapes, and report strong performance across reconstruction (CDL1/CDL2, F1-micro, NC) and segmentation (mIoU, Accuracy), together with a new Segmentation Consistency metric that captures local label smoothness. We attach a lightweight segmentation head to the Flat-CAD SDF trunk; on a paired evaluation it does not alter reconstruction while providing accurate part labels for meshes with any number of parts. Even under degraded reconstructions on thin or intricate geometries, segmentation remains accurate and label-coherent, often preserving the correct part count. Our approach therefore offers a practical route to semantically structured CAD meshes without requiring curated taxonomies or exact palette matches. We discuss limitations in boundary precision, partly due to per-face supervision, and outline paths toward boundary-aware training and higher resolution labels.

AI-Assisted Pleural Effusion Volume Estimation from Contrast-Enhanced CT Images

arXiv:2510.03856v1 Announce Type: cross Abstract: Background: Pleural Effusions (PE) is a common finding in many different clinical conditions, but accurately measuring their volume from CT scans is challenging. Purpose: To improve PE segmentation and quantification for enhanced clinical management, we have developed and trained a semi-supervised deep learning framework on contrast-enhanced CT volumes. Materials and Methods: This retrospective study collected CT Pulmonary Angiogram (CTPA) data from internal and external datasets. A subset of 100 cases was manually annotated for model training, while the remaining cases were used for testing and validation. A novel semi-supervised deep learning framework, Teacher-Teaching Assistant-Student (TTAS), was developed and used to enable efficient training in non-segmented examinations. Segmentation performance was compared to that of state-of-the-art models. Results: 100 patients (mean age, 72 years, 28 [standard deviation]; 55 men) were included in the study. The TTAS model demonstrated superior segmentation performance compared to state-of-the-art models, achieving a mean Dice score of 0.82 (95% CI, 0.79 - 0.84) versus 0.73 for nnU-Net (p < 0.0001, Student's T test). Additionally, TTAS exhibited a four-fold lower mean Absolute Volume Difference (AbVD) of 6.49 mL (95% CI, 4.80 - 8.20) compared to nnU-Net's AbVD of 23.16 mL (p < 0.0001). Conclusion: The developed TTAS framework offered superior PE segmentation, aiding accurate volume determination from CT scans.

NoTVLA: Narrowing of Dense Action Trajectories for Generalizable Robot Manipulation

arXiv:2510.03895v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models represent a pivotal advance in embodied intelligence, yet they confront critical barriers to real-world deployment, most notably catastrophic forgetting. This issue stems from their overreliance on continuous action sequences or action chunks, which inadvertently create isolated data silos that disrupt knowledge retention across tasks. To tackle these challenges, we propose the Narrowing of Trajectory VLA (NoTVLA) framework: a novel approach that narrows its focus to sparse trajectories, thereby avoiding the catastrophic forgetting associated with dense trajectory fine-tuning. A key innovation of NoTVLA lies in its trajectory planning strategy: instead of centering on the target object's trajectory, it leverages temporal compression and spatial reasoning pruning specifically for the robot end effector's trajectory. Furthermore, training is conducted using these sparse trajectories rather than dense action trajectories, an optimization that delivers remarkable practical advantages with better performance in zero-shot. In multi-task evaluation scenarios, NoTVLA achieves superior performance and generalization compared to pi0 while operating under two critical constraints: it uses over an order of magnitude less computing power than pi0 and requires no wrist-mounted camera. This design ensures that NoTVLA's operational accuracy closely approximates that of single-task expert models. Crucially, it also preserves the model's inherent language capabilities, enabling zero-shot generalization in specific scenarios, supporting unified model deployment across multiple robot platforms, and fostering a degree of generalization even when perceiving tasks from novel perspectives.

Sliding Window Attention for Learned Video Compression

arXiv:2510.03926v1 Announce Type: cross Abstract: To manage the complexity of transformers in video compression, local attention mechanisms are a practical necessity. The common approach of partitioning frames into patches, however, creates architectural flaws like irregular receptive fields. When adapted for temporal autoregressive models, this paradigm, exemplified by the Video Compression Transformer (VCT), also necessitates computationally redundant overlapping windows. This work introduces 3D Sliding Window Attention (SWA), a patchless form of local attention. By enabling a decoder-only architecture that unifies spatial and temporal context processing, and by providing a uniform receptive field, our method significantly improves rate-distortion performance, achieving Bj{\o}rntegaard Delta-rate savings of up to 18.6 % against the VCT baseline. Simultaneously, by eliminating the need for overlapping windows, our method reduces overall decoder complexity by a factor of 2.8, while its entropy model is nearly 3.5 times more efficient. We further analyze our model's behavior and show that while it benefits from long-range temporal context, excessive context can degrade performance.

Super-resolution image projection over an extended depth of field using a diffractive decoder

arXiv:2510.03938v1 Announce Type: cross Abstract: Image projection systems must be efficient in data storage, computation and transmission while maintaining a large space-bandwidth-product (SBP) at their output. Here, we introduce a hybrid image projection system that achieves extended depth-of-field (DOF) with improved resolution, combining a convolutional neural network (CNN)-based digital encoder with an all-optical diffractive decoder. A CNN-based encoder compresses input images into compact phase representations, which are subsequently displayed by a low-resolution (LR) projector and processed by an analog diffractive decoder for all-optical image reconstruction. This optical decoder is completely passive, designed to synthesize pixel super-resolved image projections that feature an extended DOF while eliminating the need for additional power consumption for super-resolved image reconstruction. Our pixel super-resolution (PSR) image projection system demonstrates high-fidelity image synthesis over an extended DOF of ~267xW, where W is the illumination wavelength, concurrently offering up to ~16-fold SBP improvement at each lateral plane. The proof of concept of this approach is validated through an experiment conducted in the THz spectrum, and the system is scalable across different parts of the electromagnetic spectrum. This image projection architecture can reduce data storage and transmission requirements for display systems without imposing additional power constraints on the optical decoder. Beyond extended DOF PSR image projection, the underlying principles of this approach can be extended to various applications, including optical metrology and microscopy.

Use of Quadcopter Wakes to Supplement Strawberry Pollination

arXiv:2510.03974v1 Announce Type: cross Abstract: Pollinators are critical to the world's ecosystems and food supply, yet recent studies have found pollination shortfalls in several crops, including strawberry. This is troubling because wild and managed pollinators are currently experiencing declines. One possibility is to try and provide supplemental pollination solutions. These solutions should be affordable and simple for farmers to implement if their use is to be widespread; quadcopters are a great example, already used for monitoring on many farms. This paper investigates a new method for artificial pollination based on wind pollination that bears further investigation. After determining the height where the lateral flow is maximized, we performed field experiments with a quadcopter assisting natural pollinators. Although our results in the field were inconclusive, lab studies show that the idea shows promise and could be adapted for better field results.

Visual Lifelog Retrieval through Captioning-Enhanced Interpretation

arXiv:2510.04010v1 Announce Type: cross Abstract: People often struggle to remember specific details of past experiences, which can lead to the need to revisit these memories. Consequently, lifelog retrieval has emerged as a crucial application. Various studies have explored methods to facilitate rapid access to personal lifelogs for memory recall assistance. In this paper, we propose a Captioning-Integrated Visual Lifelog (CIVIL) Retrieval System for extracting specific images from a user's visual lifelog based on textual queries. Unlike traditional embedding-based methods, our system first generates captions for visual lifelogs and then utilizes a text embedding model to project both the captions and user queries into a shared vector space. Visual lifelogs, captured through wearable cameras, provide a first-person viewpoint, necessitating the interpretation of the activities of the individual behind the camera rather than merely describing the scene. To address this, we introduce three distinct approaches: the single caption method, the collective caption method, and the merged caption method, each designed to interpret the life experiences of lifeloggers. Experimental results show that our method effectively describes first-person visual images, enhancing the outcomes of lifelog retrieval. Furthermore, we construct a textual dataset that converts visual lifelogs into captions, thereby reconstructing personal life experiences.

Using predefined vector systems as latent space configuration for neural network supervised training on data with arbitrarily large number of classes

arXiv:2510.04090v1 Announce Type: cross Abstract: Supervised learning (SL) methods are indispensable for neural network (NN) training used to perform classification tasks. While resulting in very high accuracy, SL training often requires making NN parameter number dependent on the number of classes, limiting their applicability when the number of classes is extremely large or unknown in advance. In this paper we propose a methodology that allows one to train the same NN architecture regardless of the number of classes. This is achieved by using predefined vector systems as the target latent space configuration (LSC) during NN training. We discuss the desired properties of target configurations and choose randomly perturbed vectors of An root system for our experiments. These vectors are used to successfully train encoders and visual transformers (ViT) on Cinic-10 and ImageNet-1K in low- and high-dimensional cases by matching NN predictions with the predefined vectors. Finally, ViT is trained on a dataset with 1.28 million classes illustrating the applicability of the method to training on datasets with extremely large number of classes. In addition, potential applications of LSC in lifelong learning and NN distillation are discussed illustrating versatility of the proposed methodology.

Learning-Based Hashing for ANN Search: Foundations and Early Advances

arXiv:2510.04127v1 Announce Type: cross Abstract: Approximate Nearest Neighbour (ANN) search is a fundamental problem in information retrieval, underpinning large-scale applications in computer vision, natural language processing, and cross-modal search. Hashing-based methods provide an efficient solution by mapping high-dimensional data into compact binary codes that enable fast similarity computations in Hamming space. Over the past two decades, a substantial body of work has explored learning to hash, where projection and quantisation functions are optimised from data rather than chosen at random. This article offers a foundational survey of early learning-based hashing methods, with an emphasis on the core ideas that shaped the field. We review supervised, unsupervised, and semi-supervised approaches, highlighting how projection functions are designed to generate meaningful embeddings and how quantisation strategies convert these embeddings into binary codes. We also examine extensions to multi-bit and multi-threshold models, as well as early advances in cross-modal retrieval. Rather than providing an exhaustive account of the most recent methods, our goal is to introduce the conceptual foundations of learning-based hashing for ANN search. By situating these early models in their historical context, we aim to equip readers with a structured understanding of the principles, trade-offs, and open challenges that continue to inform current research in this area.

MoME: Mixture of Matryoshka Experts for Audio-Visual Speech Recognition

arXiv:2510.04136v1 Announce Type: cross Abstract: Large language models (LLMs) have recently shown strong potential in audio-visual speech recognition (AVSR), but their high computational demands and sensitivity to token granularity limit their practicality in resource-constrained settings. Token compression methods can reduce inference cost, but they require fixing a compression rate in advance and produce a single fixed-length output, offering no flexibility to balance information density and efficiency at inference time. Matryoshka representation learning (MRL) addresses this by enabling a single model to operate across multiple token granularities, allowing compression rates to be adjusted dynamically. However, current MRL-based methods treat each scale independently during training, limiting cross-scale generalization, robustness at high compression, and interpretability. To overcome these limitations, we propose MoME (Mixture of Matryoshka Experts), a novel framework that integrates sparse Mixture-of-Experts (MoE) into MRL-based LLMs for AVSR. MoME augments a frozen LLM with top-k routed and shared experts, allowing dynamic capacity allocation across scales and modalities. A shared router promotes consistent expert activation across granularities, enabling compressed sequences to benefit from representations learned at lower compression. Experiments on LRS2 and LRS3 demonstrate that MoME achieves state-of-the-art performance across AVSR, ASR, and VSR tasks, while requiring significantly fewer parameters and maintaining robustness under noise. MoME unifies the adaptability of MRL with the efficiency of MoE, offering a scalable and interpretable solution for resource-aware speech recognition.

DoRAN: Stabilizing Weight-Decomposed Low-Rank Adaptation via Noise Injection and Auxiliary Networks

arXiv:2510.04331v1 Announce Type: cross Abstract: Parameter-efficient fine-tuning (PEFT) methods have become the standard paradigm for adapting large-scale models. Among these techniques, Weight-Decomposed Low-Rank Adaptation (DoRA) has been shown to improve both the learning capacity and training stability of the vanilla Low-Rank Adaptation (LoRA) method by explicitly decomposing pre-trained weights into magnitude and directional components. In this work, we propose DoRAN, a new variant of DoRA designed to further stabilize training and boost the sample efficiency of DoRA. Our approach includes two key stages: (i) injecting noise into the denominator of DoRA's weight decomposition, which serves as an adaptive regularizer to mitigate instabilities; and (ii) replacing static low-rank matrices with auxiliary networks that generate them dynamically, enabling parameter coupling across layers and yielding better sample efficiency in both theory and practice. Comprehensive experiments on vision and language benchmarks show that DoRAN consistently outperforms LoRA, DoRA, and other PEFT baselines. These results underscore the effectiveness of combining stabilization through noise-based regularization with network-based parameter generation, offering a promising direction for robust and efficient fine-tuning of foundation models.

The method of the approximate inverse for limited-angle CT

arXiv:2510.04369v1 Announce Type: cross Abstract: Limited-angle computerized tomography stands for one of the most difficult challenges in imaging. Although it opens the way to faster data acquisition in industry and less dangerous scans in medicine, standard approaches, such as the filtered backprojection (FBP) algorithm or the widely used total-variation functional, often produce various artefacts that hinder the diagnosis. With the rise of deep learning, many modern techniques have proven themselves successful in removing such artefacts but at the cost of large datasets. In this paper, we propose a new model-driven approach based on the method of the approximate inverse, which could serve as new starting point for learning strategies in the future. In contrast to FBP-type approaches, our reconstruction step consists in evaluating linear functionals on the measured data using reconstruction kernels that are precomputed as solution of an auxiliary problem. With this problem being uniquely solvable, the derived limited-angle reconstruction kernel (LARK) is able to fully reconstruct the object without the well-known streak artefacts, even for large limited angles. However, it inherits severe ill-conditioning which leads to a different kind of artefacts arising from the singular functions of the limited-angle Radon transform. The problem becomes particularly challenging when working on semi-discrete (real or analytical) measurements. We develop a general regularization strategy, named constrained limited-angle reconstruction kernel (CLARK), by combining spectral filter, the method of the approximate inverse and custom edge-preserving denoising in order to stabilize the whole process. We further derive and interpret error estimates for the application on real, i.e. semi-discrete, data and we validate our approach on synthetic and real data.

Adaptive double-phase Rudin--Osher--Fatemi denoising model

arXiv:2510.04382v1 Announce Type: cross Abstract: We propose a new image denoising model based on a variable-growth total variation regularization of double-phase type with adaptive weight. It is designed to reduce staircasing with respect to the classical Rudin--Osher--Fatemi model, while preserving the edges of the image in a similar fashion. We implement the model and test its performance on synthetic and natural images in 1D and 2D over a range of noise levels.

Partial Information Decomposition via Normalizing Flows in Latent Gaussian Distributions

arXiv:2510.04417v1 Announce Type: cross Abstract: The study of multimodality has garnered significant interest in fields where the analysis of interactions among multiple information sources can enhance predictive modeling, data fusion, and interpretability. Partial information decomposition (PID) has emerged as a useful information-theoretic framework to quantify the degree to which individual modalities independently, redundantly, or synergistically convey information about a target variable. However, existing PID methods depend on optimizing over a joint distribution constrained by estimated pairwise probability distributions, which are costly and inaccurate for continuous and high-dimensional modalities. Our first key insight is that the problem can be solved efficiently when the pairwise distributions are multivariate Gaussians, and we refer to this problem as Gaussian PID (GPID). We propose a new gradient-based algorithm that substantially improves the computational efficiency of GPID based on an alternative formulation of the underlying optimization problem. To generalize the applicability to non-Gaussian data, we learn information-preserving encoders to transform random variables of arbitrary input distributions into pairwise Gaussian random variables. Along the way, we resolved an open problem regarding the optimality of joint Gaussian solutions for GPID. Empirical validation in diverse synthetic examples demonstrates that our proposed method provides more accurate and efficient PID estimates than existing baselines. We further evaluate a series of large-scale multimodal benchmarks to show its utility in real-world applications of quantifying PID in multimodal datasets and selecting high-performing models.

Real-time Prediction of Urban Sound Propagation with Conditioned Normalizing Flows

arXiv:2510.04510v1 Announce Type: cross Abstract: Accurate and fast urban noise prediction is pivotal for public health and for regulatory workflows in cities, where the Environmental Noise Directive mandates regular strategic noise maps and action plans, often needed in permission workflows, right-of-way allocation, and construction scheduling. Physics-based solvers are too slow for such time-critical, iterative "what-if" studies. We evaluate conditional Normalizing Flows (Full-Glow) for generating for generating standards-compliant urban sound-pressure maps from 2D urban layouts in real time per 256x256 map on a single RTX 4090), enabling interactive exploration directly on commodity hardware. On datasets covering Baseline, Diffraction, and Reflection regimes, our model accelerates map generation by >2000 times over a reference solver while improving NLoS accuracy by up to 24% versus prior deep models; in Baseline NLoS we reach 0.65 dB MAE with high structural fidelity. The model reproduces diffraction and interference patterns and supports instant recomputation under source or geometry changes, making it a practical engine for urban planning, compliance mapping, and operations (e.g., temporary road closures, night-work variance assessments).

ChartAgent: A Multimodal Agent for Visually Grounded Reasoning in Complex Chart Question Answering

arXiv:2510.04514v1 Announce Type: cross Abstract: Recent multimodal LLMs have shown promise in chart-based visual question answering, but their performance declines sharply on unannotated charts, those requiring precise visual interpretation rather than relying on textual shortcuts. To address this, we introduce ChartAgent, a novel agentic framework that explicitly performs visual reasoning directly within the chart's spatial domain. Unlike textual chain-of-thought reasoning, ChartAgent iteratively decomposes queries into visual subtasks and actively manipulates and interacts with chart images through specialized actions such as drawing annotations, cropping regions (e.g., segmenting pie slices, isolating bars), and localizing axes, using a library of chart-specific vision tools to fulfill each subtask. This iterative reasoning process closely mirrors human cognitive strategies for chart comprehension. ChartAgent achieves state-of-the-art accuracy on the ChartBench and ChartX benchmarks, surpassing prior methods by up to 16.07% absolute gain overall and 17.31% on unannotated, numerically intensive queries. Furthermore, our analyses show that ChartAgent is (a) effective across diverse chart types, (b) achieve the highest scores across varying visual and reasoning complexity levels, and (c) serves as a plug-and-play framework that boosts performance across diverse underlying LLMs. Our work is among the first to demonstrate visually grounded reasoning for chart understanding using tool-augmented multimodal agents.

3Dify: a Framework for Procedural 3D-CG Generation Assisted by LLMs Using MCP and RAG

arXiv:2510.04536v1 Announce Type: cross Abstract: This paper proposes "3Dify," a procedural 3D computer graphics (3D-CG) generation framework utilizing Large Language Models (LLMs). The framework enables users to generate 3D-CG content solely through natural language instructions. 3Dify is built upon Dify, an open-source platform for AI application development, and incorporates several state-of-the-art LLM-related technologies such as the Model Context Protocol (MCP) and Retrieval-Augmented Generation (RAG). For 3D-CG generation support, 3Dify automates the operation of various Digital Content Creation (DCC) tools via MCP. When DCC tools do not support MCP-based interaction, the framework employs the Computer-Using Agent (CUA) method to automate Graphical User Interface (GUI) operations. Moreover, to enhance image generation quality, 3Dify allows users to provide feedback by selecting preferred images from multiple candidates. The LLM then learns variable patterns from these selections and applies them to subsequent generations. Furthermore, 3Dify supports the integration of locally deployed LLMs, enabling users to utilize custom-developed models and to reduce both time and monetary costs associated with external API calls by leveraging their own computational resources.

C3Editor: Achieving Controllable Consistency in 2D Model for 3D Editing

arXiv:2510.04539v1 Announce Type: cross Abstract: Existing 2D-lifting-based 3D editing methods often encounter challenges related to inconsistency, stemming from the lack of view-consistent 2D editing models and the difficulty of ensuring consistent editing across multiple views. To address these issues, we propose C3Editor, a controllable and consistent 2D-lifting-based 3D editing framework. Given an original 3D representation and a text-based editing prompt, our method selectively establishes a view-consistent 2D editing model to achieve superior 3D editing results. The process begins with the controlled selection of a ground truth (GT) view and its corresponding edited image as the optimization target, allowing for user-defined manual edits. Next, we fine-tune the 2D editing model within the GT view and across multiple views to align with the GT-edited image while ensuring multi-view consistency. To meet the distinct requirements of GT view fitting and multi-view consistency, we introduce separate LoRA modules for targeted fine-tuning. Our approach delivers more consistent and controllable 2D and 3D editing results than existing 2D-lifting-based methods, outperforming them in both qualitative and quantitative evaluations.

Post-training quantization of vision encoders needs prefixing registers

arXiv:2510.04547v1 Announce Type: cross Abstract: Transformer-based vision encoders -- such as CLIP -- are central to multimodal intelligence, powering applications from autonomous web agents to robotic control. Since these applications often demand real-time processing of massive visual data, reducing the inference cost of vision encoders is critical. Post-training quantization offers a practical path, but remains challenging even at 8-bit precision due to massive-scale activations (i.e., outliers). In this work, we propose $\textit{RegCache}$, a training-free algorithm to mitigate outliers in vision encoders, enabling quantization with significantly smaller accuracy drops. The proposed RegCache introduces outlier-prone yet semantically meaningless prefix tokens to the target vision encoder, which prevents other tokens from having outliers. Notably, we observe that outliers in vision encoders behave differently from those in language models, motivating two technical innovations: middle-layer prefixing and token deletion. Experiments show that our method consistently improves the accuracy of quantized models across both text-supervised and self-supervised vision encoders.

Fast Witness Persistence for MRI Volumes via Hybrid Landmarking

arXiv:2510.04553v1 Announce Type: cross Abstract: We introduce a scalable witness-based persistent homology pipeline for full-brain MRI volumes that couples density-aware landmark selection with a GPU-ready witness filtration. Candidates are scored by a hybrid metric that balances geometric coverage against inverse kernel density, yielding landmark sets that shrink mean pairwise distances by 30-60% over random or density-only baselines while preserving topological features. Benchmarks on BrainWeb, IXI, and synthetic manifolds execute in under ten seconds on a single NVIDIA RTX 4090 GPU, avoiding the combinatorial blow-up of Cech, Vietoris-Rips, and alpha filtrations. The package is distributed on PyPI as whale-tda (installable via pip); source and issues are hosted at https://github.com/jorgeLRW/whale. The release also exposes a fast preset (mri_deep_dive_fast) for exploratory sweeps, and ships with reproducibility-focused scripts and artifacts for drop-in use in medical imaging workflows.

SONA: Learning Conditional, Unconditional, and Mismatching-Aware Discriminator

arXiv:2510.04576v1 Announce Type: cross Abstract: Deep generative models have made significant advances in generating complex content, yet conditional generation remains a fundamental challenge. Existing conditional generative adversarial networks often struggle to balance the dual objectives of assessing authenticity and conditional alignment of input samples within their conditional discriminators. To address this, we propose a novel discriminator design that integrates three key capabilities: unconditional discrimination, matching-aware supervision to enhance alignment sensitivity, and adaptive weighting to dynamically balance all objectives. Specifically, we introduce Sum of Naturalness and Alignment (SONA), which employs separate projections for naturalness (authenticity) and alignment in the final layer with an inductive bias, supported by dedicated objective functions and an adaptive weighting mechanism. Extensive experiments on class-conditional generation tasks show that \ours achieves superior sample quality and conditional alignment compared to state-of-the-art methods. Furthermore, we demonstrate its effectiveness in text-to-image generation, confirming the versatility and robustness of our approach.

Social Agent: Mastering Dyadic Nonverbal Behavior Generation via Conversational LLM Agents

arXiv:2510.04637v1 Announce Type: cross Abstract: We present Social Agent, a novel framework for synthesizing realistic and contextually appropriate co-speech nonverbal behaviors in dyadic conversations. In this framework, we develop an agentic system driven by a Large Language Model (LLM) to direct the conversation flow and determine appropriate interactive behaviors for both participants. Additionally, we propose a novel dual-person gesture generation model based on an auto-regressive diffusion model, which synthesizes coordinated motions from speech signals. The output of the agentic system is translated into high-level guidance for the gesture generator, resulting in realistic movement at both the behavioral and motion levels. Furthermore, the agentic system periodically examines the movements of interlocutors and infers their intentions, forming a continuous feedback loop that enables dynamic and responsive interactions between the two participants. User studies and quantitative evaluations show that our model significantly improves the quality of dyadic interactions, producing natural, synchronized nonverbal behaviors.

Watch and Learn: Learning to Use Computers from Online Videos

arXiv:2510.04673v1 Announce Type: cross Abstract: Computer use agents (CUAs) need to plan task workflows grounded in diverse, ever-changing applications and environments, but learning is hindered by the scarcity of large-scale, high-quality training data in the target application. Existing datasets are domain-specific, static, and costly to annotate, while current synthetic data generation methods often yield simplistic or misaligned task demonstrations. To address these limitations, we introduce Watch & Learn (W&L), a framework that converts human demonstration videos readily available on the Internet into executable UI trajectories at scale. Instead of directly generating trajectories or relying on ad hoc reasoning heuristics, we cast the problem as an inverse dynamics objective: predicting the user's action from consecutive screen states. This formulation reduces manual engineering, is easier to learn, and generalizes more robustly across applications. Concretely, we develop an inverse dynamics labeling pipeline with task-aware video retrieval, generate over 53k high-quality trajectories from raw web videos, and demonstrate that these trajectories improve CUAs both as in-context demonstrations and as supervised training data. On the challenging OSWorld benchmark, UI trajectories extracted with W&L consistently enhance both general-purpose and state-of-the-art frameworks in-context, and deliver stronger gains for open-source models under supervised training. These results highlight web-scale human demonstration videos as a practical and scalable foundation for advancing CUAs towards real-world deployment.

CLEAR-IR: Clarity-Enhanced Active Reconstruction of Infrared Imagery

arXiv:2510.04883v1 Announce Type: cross Abstract: This paper presents a novel approach for enabling robust robotic perception in dark environments using infrared (IR) stream. IR stream is less susceptible to noise than RGB in low-light conditions. However, it is dominated by active emitter patterns that hinder high-level tasks such as object detection, tracking and localisation. To address this, a U-Net-based architecture is proposed that reconstructs clean IR images from emitter-populated input, improving both image quality and downstream robotic performance. This approach outperforms existing enhancement techniques and enables reliable operation of vision-driven robotic systems across illumination conditions from well-lit to extreme low-light scenes.

On Structured State-Space Duality

arXiv:2510.04944v1 Announce Type: cross Abstract: Structured State-Space Duality (SSD) [Dao & Gu, ICML 2024] is an equivalence between a simple Structured State-Space Model (SSM) and a masked attention mechanism. In particular, a state-space model with a scalar-times-identity state matrix is equivalent to a masked self-attention with a $1$-semiseparable causal mask. Consequently, the same sequence transformation (model) has two algorithmic realizations: as a linear-time $O(T)$ recurrence or as a quadratic-time $O(T^2)$ attention. In this note, we formalize and generalize this duality: (i) we extend SSD from the scalar-identity case to general diagonal SSMs (diagonal state matrices); (ii) we show that these diagonal SSMs match the scalar case's training complexity lower bounds while supporting richer dynamics; (iii) we establish a necessary and sufficient condition under which an SSM is equivalent to $1$-semiseparable masked attention; and (iv) we show that such duality fails to extend to standard softmax attention due to rank explosion. Together, these results tighten bridge between recurrent SSMs and Transformers, and widen the design space for expressive yet efficient sequence models.

Bridging Text and Video Generation: A Survey

arXiv:2510.04999v1 Announce Type: cross Abstract: Text-to-video (T2V) generation technology holds potential to transform multiple domains such as education, marketing, entertainment, and assistive technologies for individuals with visual or reading comprehension challenges, by creating coherent visual content from natural language prompts. From its inception, the field has advanced from adversarial models to diffusion-based models, yielding higher-fidelity, temporally consistent outputs. Yet challenges persist, such as alignment, long-range coherence, and computational efficiency. Addressing this evolving landscape, we present a comprehensive survey of text-to-video generative models, tracing their development from early GANs and VAEs to hybrid Diffusion-Transformer (DiT) architectures, detailing how these models work, what limitations they addressed in their predecessors, and why shifts toward new architectural paradigms were necessary to overcome challenges in quality, coherence, and control. We provide a systematic account of the datasets, which the surveyed text-to-video models were trained and evaluated on, and, to support reproducibility and assess the accessibility of training such models, we detail their training configurations, including their hardware specifications, GPU counts, batch sizes, learning rates, optimizers, epochs, and other key hyperparameters. Further, we outline the evaluation metrics commonly used for evaluating such models and present their performance across standard benchmarks, while also discussing the limitations of these metrics and the emerging shift toward more holistic, perception-aligned evaluation strategies. Finally, drawing from our analysis, we outline the current open challenges and propose a few promising future directions, laying out a perspective for future researchers to explore and build upon in advancing T2V research and applications.

StaMo: Unsupervised Learning of Generalizable Robot Motion from Compact State Representation

arXiv:2510.05057v1 Announce Type: cross Abstract: A fundamental challenge in embodied intelligence is developing expressive and compact state representations for efficient world modeling and decision making. However, existing methods often fail to achieve this balance, yielding representations that are either overly redundant or lacking in task-critical information. We propose an unsupervised approach that learns a highly compressed two-token state representation using a lightweight encoder and a pre-trained Diffusion Transformer (DiT) decoder, capitalizing on its strong generative prior. Our representation is efficient, interpretable, and integrates seamlessly into existing VLA-based models, improving performance by 14.3% on LIBERO and 30% in real-world task success with minimal inference overhead. More importantly, we find that the difference between these tokens, obtained via latent interpolation, naturally serves as a highly effective latent action, which can be further decoded into executable robot actions. This emergent capability reveals that our representation captures structured dynamics without explicit supervision. We name our method StaMo for its ability to learn generalizable robotic Motion from compact State representation, which is encoded from static images, challenging the prevalent dependence to learning latent action on complex architectures and video data. The resulting latent actions also enhance policy co-training, outperforming prior methods by 10.4% with improved interpretability. Moreover, our approach scales effectively across diverse data sources, including real-world robot data, simulation, and human egocentric video.

SAEdit: Token-level control for continuous image editing via Sparse AutoEncoder

arXiv:2510.05081v1 Announce Type: cross Abstract: Large-scale text-to-image diffusion models have become the backbone of modern image editing, yet text prompts alone do not offer adequate control over the editing process. Two properties are especially desirable: disentanglement, where changing one attribute does not unintentionally alter others, and continuous control, where the strength of an edit can be smoothly adjusted. We introduce a method for disentangled and continuous editing through token-level manipulation of text embeddings. The edits are applied by manipulating the embeddings along carefully chosen directions, which control the strength of the target attribute. To identify such directions, we employ a Sparse Autoencoder (SAE), whose sparse latent space exposes semantically isolated dimensions. Our method operates directly on text embeddings without modifying the diffusion process, making it model agnostic and broadly applicable to various image synthesis backbones. Experiments show that it enables intuitive and efficient manipulations with continuous control across diverse attributes and domains.

Pulp Motion: Framing-aware multimodal camera and human motion generation

arXiv:2510.05097v1 Announce Type: cross Abstract: Treating human motion and camera trajectory generation separately overlooks a core principle of cinematography: the tight interplay between actor performance and camera work in the screen space. In this paper, we are the first to cast this task as a text-conditioned joint generation, aiming to maintain consistent on-screen framing while producing two heterogeneous, yet intrinsically linked, modalities: human motion and camera trajectories. We propose a simple, model-agnostic framework that enforces multimodal coherence via an auxiliary modality: the on-screen framing induced by projecting human joints onto the camera. This on-screen framing provides a natural and effective bridge between modalities, promoting consistency and leading to more precise joint distribution. We first design a joint autoencoder that learns a shared latent space, together with a lightweight linear transform from the human and camera latents to a framing latent. We then introduce auxiliary sampling, which exploits this linear transform to steer generation toward a coherent framing modality. To support this task, we also introduce the PulpMotion dataset, a human-motion and camera-trajectory dataset with rich captions, and high-quality human motions. Extensive experiments across DiT- and MAR-based architectures show the generality and effectiveness of our method in generating on-frame coherent human-camera motions, while also achieving gains on textual alignment for both modalities. Our qualitative results yield more cinematographically meaningful framings setting the new state of the art for this task. Code, models and data are available in our \href{https://www.lix.polytechnique.fr/vista/projects/2025_pulpmotion_courant/}{project page}.

QGFace: Quality-Guided Joint Training For Mixed-Quality Face Recognition

arXiv:2312.17494v2 Announce Type: replace Abstract: The quality of a face crop in an image is decided by many factors such as camera resolution, distance, and illumination condition. This makes the discrimination of face images with different qualities a challenging problem in realistic applications. However, most existing approaches are designed specifically for high-quality (HQ) or low-quality (LQ) images, and the performances would degrade for the mixed-quality images. Besides, many methods ask for pre-trained feature extractors or other auxiliary structures to support the training and the evaluation. In this paper, we point out that the key to better understand both the HQ and the LQ images simultaneously is to apply different learning methods according to their qualities. We propose a novel quality-guided joint training approach for mixed-quality face recognition, which could simultaneously learn the images of different qualities with a single encoder. Based on quality partition, classification-based method is employed for HQ data learning. Meanwhile, for the LQ images which lack identity information, we learn them with self-supervised image-image contrastive learning. To effectively catch up the model update and improve the discriminability of contrastive learning in our joint training scenario, we further propose a proxy-updated real-time queue to compose the contrastive pairs with features from the genuine encoder. Experiments on the low-quality datasets SCface and Tinyface, the mixed-quality dataset IJB-B, and five high-quality datasets demonstrate the effectiveness of our proposed approach in recognizing face images of different qualities.

Foveated Retinotopy Improves Classification and Localization in CNNs

arXiv:2402.15480v4 Announce Type: replace Abstract: From a falcon detecting prey to humans recognizing faces, many species exhibit extraordinary abilities in rapid visual localization and classification. These are made possible by a specialized retinal region called the fovea, which provides high acuity at the center of vision while maintaining lower resolution in the periphery. This distinctive spatial organization, preserved along the early visual pathway through retinotopic mapping, is fundamental to biological vision, yet remains largely unexplored in machine learning. Our study investigates how incorporating foveated retinotopy may benefit deep convolutional neural networks (CNNs) in image classification tasks. By implementing a foveated retinotopic transformation in the input layer of standard ResNet models and re-training them, we maintain comparable classification accuracy while enhancing the network's robustness to scale and rotational perturbations. Although this architectural modification introduces increased sensitivity to fixation point shifts, we demonstrate how this apparent limitation becomes advantageous: variations in classification probabilities across different gaze positions serve as effective indicators for object localization. Our findings suggest that foveated retinotopic mapping encodes implicit knowledge about visual object geometry, offering an efficient solution to the visual search problem - a capability crucial for many living species.

Interactive Test-Time Adaptation with Reliable Spatial-Temporal Voxels for Multi-Modal Segmentation

arXiv:2403.06461v5 Announce Type: replace Abstract: Multi-modal test-time adaptation (MM-TTA) adapts models to an unlabeled target domain by leveraging the complementary multi-modal inputs in an online manner. While previous MM-TTA methods for 3D segmentation offer a promising solution by leveraging self-refinement per frame, they suffer from two major limitations: 1) unstable frame-wise predictions caused by temporal inconsistency, and 2) consistently incorrect predictions that violate the assumption of reliable modality guidance. To address these limitations, this work introduces a comprehensive two-fold framework. Firstly, building upon our previous work ReLiable Spatial-temporal Voxels (Latte), we propose Latte++ that better suppresses the unstable frame-wise predictions with more informative geometric correspondences. Instead of utilizing a universal sliding window, Latte++ employs multi-window aggregation to capture more reliable correspondences to better evaluate the local prediction consistency of different semantic categories. Secondly, to tackle the consistently incorrect predictions, we propose Interactive Test-Time Adaptation (ITTA), a flexible add-on to empower effortless human feedback with existing MM-TTA methods. ITTA introduces a novel human-in-the-loop approach that efficiently integrates minimal human feedback through interactive segmentation, requiring only simple point clicks and bounding box annotations. Instead of using independent interactive networks, ITTA employs a lightweight promptable branch with a momentum gradient module to capture and reuse knowledge from scarce human feedback during online inference. Extensive experiments across five MM-TTA benchmarks demonstrate that ITTA achieves consistent and notable improvements with robust performance gains for target classes of interest in challenging imbalanced scenarios, while Latte++ provides complementary benefits for temporal stability.

Evaluating Perceptual Distance Models by Fitting Binomial Distributions to Two-Alternative Forced Choice Data

arXiv:2403.10390v4 Announce Type: replace Abstract: The Two Alternative Forced Choice (2AFC) paradigm offers advantages over the Mean Opinion Score (MOS) paradigm in psychophysics (PF), such as simplicity and robustness. However, when evaluating perceptual distance models, MOS enables direct correlation between model predictions and PF data. In contrast, 2AFC only allows pairwise comparisons to be converted into a quality ranking similar to MOS when comparisons include shared images. In large datasets, like BAPPS, where image patches and distortions are combined randomly, deriving rankings from 2AFC PF data becomes infeasible, as distorted images included in each comparisons are independent. To address this, instead of relying on MOS correlation, researchers have trained ad-hoc neural networks to reproduce 2AFC PF data based on pairs of model distances - a black-box approach with conceptual and operational limitations. This paper introduces a more robust distance-model evaluation method using a pure probabilistic approach, applying maximum likelihood estimation to a binomial decision model. Our method demonstrates superior simplicity, interpretability, flexibility, and computational efficiency, as shown through evaluations of various visual distance models on two 2AFC PF datasets.

Reconstructing Topology-Consistent Face Mesh by Volume Rendering from Multi-View Images

arXiv:2404.05606v2 Announce Type: replace Abstract: Industrial 3D face assets creation typically reconstructs topology-consistent face meshes from multi-view images for downstream production. However, high-quality reconstruction usually requires manual processing or specific capture settings. Recently NeRF has shown great advantages in 3D reconstruction, by representing scenes as density and radiance fields and utilizing neural volume rendering for novel view synthesis. Inspired by this, we introduce a novel method which combines explicit mesh with neural volume rendering to optimize geometry of an artist-made template face mesh from multi-view images while keeping the topology unchanged. Our method derives density fields from meshes using distance fields as an intermediary and encodes radiance field in compact tri-planes. To improve convergence, several adaptions tailored for meshes are introduced to the volume rendering. Experiments demonstrate that our method achieves superior reconstruction quality compared to previous approaches, validating the feasibility of integrating mesh and neural volume rendering.

Capsule Network Projectors are Equivariant and Invariant Learners

arXiv:2405.14386v4 Announce Type: replace Abstract: Learning invariant representations has been the long-standing approach to self-supervised learning. However, recently progress has been made in preserving equivariant properties in representations, yet do so with highly prescribed architectures. In this work, we propose an invariant-equivariant self-supervised architecture that employs Capsule Networks (CapsNets), which have been shown to capture equivariance with respect to novel viewpoints. We demonstrate that the use of CapsNets in equivariant self-supervised architectures achieves improved downstream performance on equivariant tasks with higher efficiency and fewer network parameters. To accommodate the architectural changes of CapsNets, we introduce a new objective function based on entropy minimisation. This approach, which we name CapsIE (Capsule Invariant Equivariant Network), achieves state-of-the-art performance on the equivariant rotation tasks on the 3DIEBench dataset compared to prior equivariant SSL methods, while performing competitively against supervised counterparts. Our results demonstrate the ability of CapsNets to learn complex and generalised representations for large-scale, multi-task datasets compared to previous CapsNet benchmarks. Code is available at https://github.com/AberdeenML/CapsIE.

Towards Cross-modal Backward-compatible Representation Learning for Vision-Language Models

arXiv:2405.14715v3 Announce Type: replace Abstract: Modern retrieval systems often struggle with upgrading to new and more powerful models due to the incompatibility of embeddings between the old and new models. This necessitates a costly process known as backfilling, which involves re-computing the embeddings for a large number of data samples. In vision, Backward-compatible Training (BT) has been proposed to ensure that the new model aligns with the old model's embeddings. This paper extends the concept of vision-only BT to the field of cross-modal retrieval, marking the first attempt to address Cross-modal BT (XBT). Our goal is to achieve backward-compatibility between Vision-Language Pretraining (VLP) models, such as CLIP, for the cross-modal retrieval task. To address XBT challenges, we propose an efficient solution: a projection module that maps the new model's embeddings to those of the old model. This module, pretrained solely with text data, significantly reduces the number of image-text pairs required for XBT learning, and, once it is pretrained, it avoids using the old model during training. Furthermore, we utilize parameter-efficient training strategies that improve efficiency and preserve the off-the-shelf new model's knowledge by avoiding any modifications. Experimental results on cross-modal retrieval datasets demonstrate the effectiveness of XBT and its potential to enable backfill-free upgrades when a new VLP model emerges.

How Far are AI-generated Videos from Simulating the 3D Visual World: A Learned 3D Evaluation Approach

arXiv:2406.19568v2 Announce Type: replace Abstract: Recent advancements in video diffusion models enable the generation of photorealistic videos with impressive 3D consistency and temporal coherence. However, the extent to which these AI-generated videos simulate the 3D visual world remains underexplored. In this paper, we introduce Learned 3D Evaluation (L3DE), an objective, quantifiable, and interpretable method for assessing AI-generated videos' ability to simulate the real world in terms of 3D visual qualities and consistencies, without requiring manually labeled defects or quality annotations. Instead of relying on 3D reconstruction, which is prone to failure with in-the-wild videos, L3DE employs a 3D convolutional network, trained on monocular 3D cues of motion, depth, and appearance, to distinguish real from synthetic videos. Confidence scores from L3DE quantify the gap between real and synthetic videos in terms of 3D visual coherence, while a gradient-based visualization pinpoints unrealistic regions, improving interpretability. We validate L3DE through extensive experiments, demonstrating strong alignment with 3D reconstruction quality and human judgments. Our evaluations on leading generative models (e.g., Kling, Sora, and MiniMax) reveal persistent simulation gaps and subtle inconsistencies. Beyond generative video assessment, L3DE extends to broader applications: benchmarking video generation models, serving as a deepfake detector, and enhancing video synthesis by inpainting flagged inconsistencies. Project page: https://justin-crchang.github.io/l3de-project-page/

PanDORA: Casual HDR Radiance Acquisition for Indoor Scenes

arXiv:2407.06150v2 Announce Type: replace Abstract: Most novel view synthesis methods-including Neural Radiance Fields (NeRF)-struggle to capture the true high dynamic range (HDR) radiance of scenes. This is primarily due to their dependence on low dynamic range (LDR) images from conventional cameras. Exposure bracketing techniques aim to address this challenge, but they introduce a considerable time burden during the acquisition process. In this work, we introduce PanDORA: PANoramic Dual-Observer Radiance Acquisition, a system designed for the casual, high quality HDR capture of indoor environments. Our approach uses two 360{\deg} cameras mounted on a portable monopod to simultaneously record two panoramic 360{\deg} videos: one with standard exposure and another at fast shutter speed. The resulting video data is processed by a proposed two-stage NeRF-based algorithm, including an algorithm for the fine alignment of the fast- and well-exposed frames, generating non-saturated HDR radiance maps. Compared to existing methods on a novel dataset of real indoor scenes captured with our apparatus and including HDR ground truth lighting, PanDORA achieves superior visual fidelity and provides a scalable solution for capturing real environments in HDR.

A Survey of Defenses Against AI-Generated Visual Media: Detection,Disruption, and Authentication

arXiv:2407.10575v3 Announce Type: replace Abstract: Deep generative models have demonstrated impressive performance in various computer vision applications, including image synthesis, video generation, and medical analysis. Despite their significant advancements, these models may be used for malicious purposes, such as misinformation, deception, and copyright violation. In this paper, we provide a systematic and timely review of research efforts on defenses against AI-generated visual media, covering detection, disruption, and authentication. We review existing methods and summarize the mainstream defense-related tasks within a unified passive and proactive framework. Moreover, we survey the derivative tasks concerning the trustworthiness of defenses, such as their robustness and fairness. For each defense strategy, we formulate its general pipeline and propose a multidimensional taxonomy applicable across defense tasks, based on methodological strategies. Additionally, we summarize the commonly used evaluation datasets, criteria, and metrics. Finally, by analyzing the reviewed studies, we provide insights into current research challenges and suggest possible directions for future research.

ShapeICP: Iterative Category-level Object Pose and Shape Estimation from Depth

arXiv:2408.13147v3 Announce Type: replace Abstract: Category-level object pose and shape estimation from a single depth image has recently drawn research attention due to its potential utility for tasks such as robotics manipulation. The task is particularly challenging because the three unknowns, object pose, object shape, and model-to-measurement correspondences, are compounded together, but only a single view of depth measurements is provided. Most of the prior work heavily relies on data-driven approaches to obtain solutions to at least one of the unknowns, and typically two, risking generalization failures if not designed and trained carefully. The shape representations used in the prior work also mainly focus on point clouds and signed distance fields (SDFs). In stark contrast to the prior work, we approach the problem using an iterative estimation method that does not require learning from pose-annotated data. Moreover, we construct and adopt a novel mesh-based object active shape model (ASM), which additionally maintains vertex connectivity compared to the commonly used point-based object ASM. Our algorithm, ShapeICP, is based on the iterative closest point (ICP) algorithm but is equipped with additional features for the category-level pose and shape estimation task. Although not using pose-annotated data, ShapeICP surpasses many data-driven approaches that rely on pose data for training, opening up a new solution space for researchers to consider.

Law of Vision Representation in MLLMs

arXiv:2408.16357v3 Announce Type: replace Abstract: We present the "Law of Vision Representation" in multimodal large language models (MLLMs). It reveals a strong correlation between the combination of cross-modal alignment, correspondence in vision representation, and MLLM performance. We quantify the two factors using the cross-modal Alignment and Correspondence score (AC score). Through extensive experiments involving thirteen different vision representation settings and evaluations across eight benchmarks, we find that the AC score is linearly correlated to model performance. By leveraging this relationship, we are able to identify and train the optimal vision representation only, which does not require finetuning the language model every time, resulting in a 99.7% reduction in computational cost.

SteerDiff: Steering towards Safe Text-to-Image Diffusion Models

arXiv:2410.02710v2 Announce Type: replace Abstract: Text-to-image (T2I) diffusion models have drawn attention for their ability to generate high-quality images with precise text alignment. However, these models can also be misused to produce inappropriate content. Existing safety measures, which typically rely on text classifiers or ControlNet-like approaches, are often insufficient. Traditional text classifiers rely on large-scale labeled datasets and can be easily bypassed by rephrasing. As diffusion models continue to scale, fine-tuning these safeguards becomes increasingly challenging and lacks flexibility. Recent red-teaming attack researches further underscore the need for a new paradigm to prevent the generation of inappropriate content. In this paper, we introduce SteerDiff, a lightweight adaptor module designed to act as an intermediary between user input and the diffusion model, ensuring that generated images adhere to ethical and safety standards with little to no impact on usability. SteerDiff identifies and manipulates inappropriate concepts within the text embedding space to guide the model away from harmful outputs. We conduct extensive experiments across various concept unlearning tasks to evaluate the effectiveness of our approach. Furthermore, we benchmark SteerDiff against multiple red-teaming strategies to assess its robustness. Finally, we explore the potential of SteerDiff for concept forgetting tasks, demonstrating its versatility in text-conditioned image generation.

Fast constrained sampling in pre-trained diffusion models

arXiv:2410.18804v3 Announce Type: replace Abstract: Large denoising diffusion models, such as Stable Diffusion, have been trained on billions of image-caption pairs to perform text-conditioned image generation. As a byproduct of this training, these models have acquired general knowledge about image statistics, which can be useful for other inference tasks. However, when confronted with sampling an image under new constraints, e.g. generating the missing parts of an image, using large pre-trained text-to-image diffusion models is inefficient and often unreliable. Previous approaches either utilized backpropagation through the denoiser network, making them significantly slower and more memory-demanding than simple text-to-image generation, or only enforced the constraint locally, failing to capture critical long-range correlations in the sampled image. In this work, we propose an algorithm that enables fast, high-quality generation under arbitrary constraints. We show that in denoising diffusion models, we can employ an approximation to Newton's optimization method that allows us to speed up inference and avoid the expensive backpropagation operations. Our approach produces results that rival or surpass the state-of-the-art training-free inference methods while requiring a fraction of the time. We demonstrate the effectiveness of our algorithm under both linear (inpainting, super-resolution) and non-linear (style-guided generation) constraints. An implementation is provided at https://github.com/cvlab-stonybrook/fast-constrained-sampling.

Probabilistic Language-Image Pre-Training

arXiv:2410.18857v4 Announce Type: replace Abstract: Vision-language models (VLMs) embed aligned image-text pairs into a joint space but often rely on deterministic embeddings, assuming a one-to-one correspondence between images and texts. This oversimplifies real-world relationships, which are inherently many-to-many, with multiple captions describing a single image and vice versa. We introduce Probabilistic Language-Image Pre-training (ProLIP), the first probabilistic VLM pre-trained on a billion-scale image-text dataset using only probabilistic objectives, achieving a strong zero-shot capability (e.g., 74.6% ImageNet zero-shot accuracy with ViT-B/16). ProLIP efficiently estimates uncertainty by an "uncertainty token" without extra parameters. We also introduce a novel inclusion loss that enforces distributional inclusion relationships between image-text pairs and between original and masked inputs. Experiments demonstrate that, by leveraging uncertainty estimates, ProLIP benefits downstream tasks and aligns with intuitive notions of uncertainty, e.g., shorter texts being more uncertain and more general inputs including specific ones. Utilizing text uncertainties, we further improve ImageNet accuracy from 74.6% to 75.8% (under a few-shot setting), supporting the practical advantages of our probabilistic approach. The code is available at https://github.com/naver-ai/prolip

CoralSCOP-LAT: Labeling and Analyzing Tool for Coral Reef Images with Dense Mask

arXiv:2410.20436v2 Announce Type: replace Abstract: Coral reef imagery offers critical data for monitoring ecosystem health, in particular as the ease of image datasets continues to rapidly expand. Whilst semi-automated analytical platforms for reef imagery are becoming more available, the dominant approaches face fundamental limitations. To address these challenges, we propose CoralSCOP-LAT, a coral reef image analysis and labeling tool that automatically segments and analyzes coral regions. By leveraging advanced machine learning models tailored for coral reef segmentation, CoralSCOP-LAT enables users to generate dense segmentation masks with minimal manual effort, significantly enhancing both the labeling efficiency and precision of coral reef analysis. Our extensive evaluations demonstrate that CoralSCOP-LAT surpasses existing coral reef analysis tools in terms of time efficiency, accuracy, precision, and flexibility. CoralSCOP-LAT, therefore, not only accelerates the coral reef annotation process but also assists users in obtaining high-quality coral reef segmentation and analysis outcomes. Github Page: https://github.com/ykwongaq/CoralSCOP-LAT.

SEE-DPO: Self Entropy Enhanced Direct Preference Optimization

arXiv:2411.04712v2 Announce Type: replace Abstract: Direct Preference Optimization (DPO) has been successfully used to align large language models (LLMs) according to human preferences, and more recently it has also been applied to improving the quality of text-to-image diffusion models. However, DPO-based methods such as SPO, Diffusion-DPO, and D3PO are highly susceptible to overfitting and reward hacking, especially when the generative model is optimized to fit out-of-distribution during prolonged training. To overcome these challenges and stabilize the training of diffusion models, we introduce a self-entropy regularization mechanism in reinforcement learning from human feedback. This enhancement improves DPO training by encouraging broader exploration and greater robustness. Our regularization technique effectively mitigates reward hacking, leading to improved stability and enhanced image quality across the latent space. Extensive experiments demonstrate that integrating human feedback with self-entropy regularization can significantly boost image diversity and specificity, achieving state-of-the-art results on key image generation metrics.

TimeFormer: Capturing Temporal Relationships of Deformable 3D Gaussians for Robust Reconstruction

arXiv:2411.11941v2 Announce Type: replace Abstract: Dynamic scene reconstruction is a long-term challenge in 3D vision. Recent methods extend 3D Gaussian Splatting to dynamic scenes via additional deformation fields and apply explicit constraints like motion flow to guide the deformation. However, they learn motion changes from individual timestamps independently, making it challenging to reconstruct complex scenes, particularly when dealing with violent movement, extreme-shaped geometries, or reflective surfaces. To address the above issue, we design a plug-and-play module called TimeFormer to enable existing deformable 3D Gaussians reconstruction methods with the ability to implicitly model motion patterns from a learning perspective. Specifically, TimeFormer includes a Cross-Temporal Transformer Encoder, which adaptively learns the temporal relationships of deformable 3D Gaussians. Furthermore, we propose a two-stream optimization strategy that transfers the motion knowledge learned from TimeFormer to the base stream during the training phase. This allows us to remove TimeFormer during inference, thereby preserving the original rendering speed. Extensive experiments in the multi-view and monocular dynamic scenes validate qualitative and quantitative improvement brought by TimeFormer. Project Page: https://patrickddj.github.io/TimeFormer/

Grounding-IQA: Grounding Multimodal Language Model for Image Quality Assessment

arXiv:2411.17237v3 Announce Type: replace Abstract: The development of multimodal large language models (MLLMs) enables the evaluation of image quality through natural language descriptions. This advancement allows for more detailed assessments. However, these MLLM-based IQA methods primarily rely on general contextual descriptions, sometimes limiting fine-grained quality assessment. To address this limitation, we introduce a new image quality assessment (IQA) task paradigm, grounding-IQA. This paradigm integrates multimodal referring and grounding with IQA to realize more fine-grained quality perception, thereby extending existing IQA. Specifically, grounding-IQA comprises two subtasks: grounding-IQA-description (GIQA-DES) and visual question answering (GIQA-VQA). GIQA-DES involves detailed descriptions with precise locations (e.g., bounding boxes), while GIQA-VQA focuses on quality QA for local regions. To realize grounding-IQA, we construct a corresponding dataset, GIQA-160K, through our proposed automated annotation pipeline. Furthermore, we develop a well-designed benchmark, GIQA-Bench. The benchmark evaluates the grounding-IQA performance from three perspectives: description quality, VQA accuracy, and grounding precision. Experiments demonstrate that our proposed method facilitates the more fine-grained IQA application. Code: https://github.com/zhengchen1999/Grounding-IQA.

Autonomous Imagination: Closed-Loop Decomposition of Visual-to-Textual Conversion in Visual Reasoning for Multimodal Large Language Models

arXiv:2411.18142v4 Announce Type: replace Abstract: Under pure textual modality, Large Language Models (LLMs) have demonstrated remarkable success in complex reasoning tasks by decomposing them into simpler sub-problems. However, Multimodal Large Language Models (MLLMs) still struggle with some seemingly straightforward visual tasks, such as counting and solving jigsaw puzzles. We argue that these tasks challenge the ability of visual-to-textual conversion, where MLLMs convert visual information perceived from the input scene, to textual information for further reasoning and generating the answer. If the complexity of the visual input is beyond the perceptual capability of the MLLMs, without decomposing this conversion process, simply scaling inference-time reasoning cannot solve the task because it repeatedly encounters the same perceptual bottleneck. We propose an approach, autonomous imagination, to enable MLLMs to iteratively modify visual inputs (e.g. isolating objects, rearranging puzzle pieces) into intermediate visual states, decomposing visual-to-textual conversion into closed-loop visual modification steps. We show that, without any retraining, MLLMs can now solve tasks initially beyond their perceptual capability, highlighting that closed-loop visual modification can be an effective way of decomposing the visual reasoning task into solvable substeps. Our code and data are released at https://future-item.github.io/autoimagine-site/.

STIV: Scalable Text and Image Conditioned Video Generation

arXiv:2412.07730v2 Announce Type: replace Abstract: The field of video generation has made remarkable advancements, yet there remains a pressing need for a clear, systematic recipe that can guide the development of robust and scalable models. In this work, we present a comprehensive study that systematically explores the interplay of model architectures, training recipes, and data curation strategies, culminating in a simple and scalable text-image-conditioned video generation method, named STIV. Our framework integrates image condition into a Diffusion Transformer (DiT) through frame replacement, while incorporating text conditioning via a joint image-text conditional classifier-free guidance. This design enables STIV to perform both text-to-video (T2V) and text-image-to-video (TI2V) tasks simultaneously. Additionally, STIV can be easily extended to various applications, such as video prediction, frame interpolation, multi-view generation, and long video generation, etc. With comprehensive ablation studies on T2I, T2V, and TI2V, STIV demonstrate strong performance, despite its simple design. An 8.7B model with 512 resolution achieves 83.1 on VBench T2V, surpassing both leading open and closed-source models like CogVideoX-5B, Pika, Kling, and Gen-3. The same-sized model also achieves a state-of-the-art result of 90.1 on VBench I2V task at 512 resolution. By providing a transparent and extensible recipe for building cutting-edge video generation models, we aim to empower future research and accelerate progress toward more versatile and reliable video generation solutions.

RowDetr: End-to-End Crop Row Detection Using Polynomials

arXiv:2412.10525v3 Announce Type: replace Abstract: Crop row detection enables autonomous robots to navigate in gps denied environments. Vision based strategies often struggle in the environments due to gaps, curved crop rows and require post-processing steps. Furthermore, labeling crop rows in under the canopy environments accurately is very difficult due to occlusions. This study introduces RowDetr, an efficient end-to-end transformer-based neural network for crop row detection in precision agriculture. RowDetr leverages a lightweight backbone and a hybrid encoder to model straight, curved, or occluded crop rows with high precision. Central to the architecture is a novel polynomial representation that enables direct parameterization of crop rows, eliminating computationally expensive post-processing. Key innovations include a PolySampler module and multi-scale deformable attention, which work together with PolyOptLoss, an energy-based loss function designed to optimize geometric alignment between predicted and the annotated crop rows, while also enhancing robustness against labeling noise. RowDetr was evaluated against other state-of-the-art end-to-end crop row detection methods like AgroNav and RolColAttention on a diverse dataset of 6,962 high-resolution images, used for training, validation, and testing across multiple crop types with annotated crop rows. The system demonstrated superior performance, achieved an F1 score up to 0.74 and a lane position deviation as low as 0.405. Furthermore, RowDetr achieves a real-time inference latency of 6.7ms, which was optimized to 3.5ms with INT8 quantization on an NVIDIA Jetson Orin AGX. This work highlighted the critical efficiency of polynomial parameterization, making RowDetr particularly suitable for deployment on edge computing devices in agricultural robotics and autonomous farming equipment. Index terms > Crop Row Detection, Under Canopy Navigation, Transformers, RT-DETR, RT-DETRv2

DynamicScaler: Seamless and Scalable Video Generation for Panoramic Scenes

arXiv:2412.11100v2 Announce Type: replace Abstract: The increasing demand for immersive AR/VR applications and spatial intelligence has heightened the need to generate high-quality scene-level and 360${\deg}$ panoramic video. However, most video diffusion models are constrained by limited resolution and aspect ratio, which restricts their applicability to scene-level dynamic content synthesis. In this work, we propose $\textbf{DynamicScaler}$, addressing these challenges by enabling spatially scalable and panoramic dynamic scene synthesis that preserves coherence across panoramic scenes of arbitrary size. Specifically, we introduce a Offset Shifting Denoiser, facilitating efficient, synchronous, and coherent denoising panoramic dynamic scenes via a diffusion model with fixed resolution through a seamless rotating Window, which ensures seamless boundary transitions and consistency across the entire panoramic space, accommodating varying resolutions and aspect ratios. Additionally, we employ a Global Motion Guidance mechanism to ensure both local detail fidelity and global motion continuity. Extensive experiments demonstrate our method achieves superior content and motion quality in panoramic scene-level video generation, offering a training-free, efficient, and scalable solution for immersive dynamic scene creation with constant VRAM consumption regardless of the output video resolution. Project page is available at $\href{https://dynamic-scaler.pages.dev/new}{https://dynamic-scaler.pages.dev/new}$.

Leveraging Confident Image Regions for Source-Free Domain-Adaptive Object Detection

arXiv:2501.10081v2 Announce Type: replace Abstract: Source-free domain-adaptive object detection is an interesting but scarcely addressed topic. It aims at adapting a source-pretrained detector to a distinct target domain without resorting to source data during adaptation. So far, there is no data augmentation scheme tailored to source-free domain-adaptive object detection. To this end, this paper presents a novel data augmentation approach that cuts out target image regions where the detector is confident, augments them along with their respective pseudo-labels, and joins them into a challenging target image to adapt the detector. As the source data is out of reach during adaptation, we implement our approach within a teacher-student learning paradigm to ensure that the model does not collapse during the adaptation procedure. We evaluated our approach on three adaptation benchmarks of traffic scenes, scoring new state-of-the-art on two of them.

Fast-RF-Shimming: Accelerate RF Shimming in 7T MRI using Deep Learning

arXiv:2501.12157v2 Announce Type: replace Abstract: Ultrahigh field (UHF) Magnetic Resonance Imaging (MRI) offers an elevated signal-to-noise ratio (SNR), enabling exceptionally high spatial resolution that benefits both clinical diagnostics and advanced research. However, the jump to higher fields introduces complications, particularly transmit radiofrequency (RF) field ($B_{1}^{+}$) inhomogeneities, manifesting as uneven flip angles and image intensity irregularities. These artifacts can degrade image quality and impede broader clinical adoption. Traditional RF shimming methods, such as Magnitude Least Squares (MLS) optimization, effectively mitigate $B_{1}^{+}$ inhomogeneity, but remain time-consuming. Recent machine learning approaches, including RF Shim Prediction by Iteratively Projected Ridge Regression and other deep learning architectures, suggest alternative pathways. Although these approaches show promise, challenges such as extensive training periods, limited network complexity, and practical data requirements persist. In this paper, we introduce a holistic learning-based framework called Fast-RF-Shimming, which achieves a 5000x speed-up compared to the traditional MLS method. In the initial phase, we employ random-initialized Adaptive Moment Estimation (Adam) to derive the desired reference shimming weights from multi-channel $B_{1}^{+}$ fields. Next, we train a Residual Network (ResNet) to map $B_{1}^{+}$ fields directly to the ultimate RF shimming outputs, incorporating the confidence parameter into its loss function. Finally, we design Non-uniformity Field Detector (NFD), an optional post-processing step, to ensure the extreme non-uniform outcomes are identified. Comparative evaluations with standard MLS optimization underscore notable gains in both processing speed and predictive accuracy, which indicates that our technique shows a promising solution for addressing persistent inhomogeneity challenges.

CBVLM: Training-free Explainable Concept-based Large Vision Language Models for Medical Image Classification

arXiv:2501.12266v3 Announce Type: replace Abstract: The main challenges limiting the adoption of deep learning-based solutions in medical workflows are the availability of annotated data and the lack of interpretability of such systems. Concept Bottleneck Models (CBMs) tackle the latter by constraining the model output on a set of predefined and human-interpretable concepts. However, the increased interpretability achieved through these concept-based explanations implies a higher annotation burden. Moreover, if a new concept needs to be added, the whole system needs to be retrained. Inspired by the remarkable performance shown by Large Vision-Language Models (LVLMs) in few-shot settings, we propose a simple, yet effective, methodology, CBVLM, which tackles both of the aforementioned challenges. First, for each concept, we prompt the LVLM to answer if the concept is present in the input image. Then, we ask the LVLM to classify the image based on the previous concept predictions. Moreover, in both stages, we incorporate a retrieval module responsible for selecting the best examples for in-context learning. By grounding the final diagnosis on the predicted concepts, we ensure explainability, and by leveraging the few-shot capabilities of LVLMs, we drastically lower the annotation cost. We validate our approach with extensive experiments across four medical datasets and twelve LVLMs (both generic and medical) and show that CBVLM consistently outperforms CBMs and task-specific supervised methods without requiring any training and using just a few annotated examples. More information on our project page: https://cristianopatricio.github.io/CBVLM/.

UniUIR: Considering Underwater Image Restoration as An All-in-One Learner

arXiv:2501.12981v2 Announce Type: replace Abstract: Existing underwater image restoration (UIR) methods generally only handle color distortion or jointly address color and haze issues, but they often overlook the more complex degradations that can occur in underwater scenes. To address this limitation, we propose a Universal Underwater Image Restoration method, termed as UniUIR, considering the complex scenario of real-world underwater mixed distortions as an all-in-one manner. To decouple degradation-specific issues and explore the inter-correlations among various degradations in UIR task, we designed the Mamba Mixture-of-Experts module. This module enables each expert to identify distinct types of degradation and collaboratively extract task-specific priors while maintaining global feature representation based on linear complexity. Building upon this foundation, to enhance degradation representation and address the task conflicts that arise when handling multiple types of degradation, we introduce the spatial-frequency prior generator. This module extracts degradation prior information in both spatial and frequency domains, and adaptively selects the most appropriate task-specific prompts based on image content, thereby improving the accuracy of image restoration. Finally, to more effectively address complex, region-dependent distortions in UIR task, we incorporate depth information derived from a large-scale pre-trained depth prediction model, thereby enabling the network to perceive and leverage depth variations across different image regions to handle localized degradation. Extensive experiments demonstrate that UniUIR can produce more attractive results across qualitative and quantitative comparisons, and shows strong generalization than state-of-the-art methods.

WonderHuman: Hallucinating Unseen Parts in Dynamic 3D Human Reconstruction

arXiv:2502.01045v2 Announce Type: replace Abstract: In this paper, we present WonderHuman to reconstruct dynamic human avatars from a monocular video for high-fidelity novel view synthesis. Previous dynamic human avatar reconstruction methods typically require the input video to have full coverage of the observed human body. However, in daily practice, one typically has access to limited viewpoints, such as monocular front-view videos, making it a cumbersome task for previous methods to reconstruct the unseen parts of the human avatar. To tackle the issue, we present WonderHuman, which leverages 2D generative diffusion model priors to achieve high-quality, photorealistic reconstructions of dynamic human avatars from monocular videos, including accurate rendering of unseen body parts. Our approach introduces a Dual-Space Optimization technique, applying Score Distillation Sampling (SDS) in both canonical and observation spaces to ensure visual consistency and enhance realism in dynamic human reconstruction. Additionally, we present a View Selection strategy and Pose Feature Injection to enforce the consistency between SDS predictions and observed data, ensuring pose-dependent effects and higher fidelity in the reconstructed avatar. In the experiments, our method achieves SOTA performance in producing photorealistic renderings from the given monocular video, particularly for those challenging unseen parts. The project page and source code can be found at https://wyiguanw.github.io/WonderHuman/.

Controllable Video Generation with Provable Disentanglement

arXiv:2502.02690v3 Announce Type: replace Abstract: Controllable video generation remains a significant challenge, despite recent advances in generating high-quality and consistent videos. Most existing methods for controlling video generation treat the video as a whole, neglecting intricate fine-grained spatiotemporal relationships, which limits both control precision and efficiency. In this paper, we propose Controllable Video Generative Adversarial Networks (CoVoGAN) to disentangle the video concepts, thus facilitating efficient and independent control over individual concepts. Specifically, following the minimal change principle, we first disentangle static and dynamic latent variables. We then leverage the sufficient change property to achieve component-wise identifiability of dynamic latent variables, enabling disentangled control of video generation. To establish the theoretical foundation, we provide a rigorous analysis demonstrating the identifiability of our approach. Building on these theoretical insights, we design a Temporal Transition Module to disentangle latent dynamics. To enforce the minimal change principle and sufficient change property, we minimize the dimensionality of latent dynamic variables and impose temporal conditional independence. To validate our approach, we integrate this module as a plug-in for GANs. Extensive qualitative and quantitative experiments on various video generation benchmarks demonstrate that our method significantly improves generation quality and controllability across diverse real-world scenarios.

Medical Image Classification with KAN-Integrated Transformers and Dilated Neighborhood Attention

arXiv:2502.13693v3 Announce Type: replace Abstract: Convolutional networks, transformers, hybrid models, and Mamba-based architectures have demonstrated strong performance across various medical image classification tasks. However, these methods were primarily designed to classify clean images using labeled data. In contrast, real-world clinical data often involve image corruptions that are unique to multi-center studies and stem from variations in imaging equipment across manufacturers. In this paper, we introduce the Medical Vision Transformer (MedViTV2), a novel architecture incorporating Kolmogorov-Arnold Network (KAN) layers into the transformer architecture for the first time, aiming for generalized medical image classification. We have developed an efficient KAN block to reduce computational load while enhancing the accuracy of the original MedViT. Additionally, to counteract the fragility of our MedViT when scaled up, we propose an enhanced Dilated Neighborhood Attention (DiNA), an adaptation of the efficient fused dot-product attention kernel capable of capturing global context and expanding receptive fields to scale the model effectively and addressing feature collapse issues. Moreover, a hierarchical hybrid strategy is introduced to stack our Local Feature Perception and Global Feature Perception blocks in an efficient manner, which balances local and global feature perceptions to boost performance. Extensive experiments on 17 medical image classification datasets and 12 corrupted medical image datasets demonstrate that MedViTV2 achieved state-of-the-art results in 27 out of 29 experiments with reduced computational complexity. MedViTV2 is 44\% more computationally efficient than the previous version and significantly enhances accuracy, achieving improvements of 4.6\% on MedMNIST, 5.8\% on NonMNIST, and 13.4\% on the MedMNIST-C benchmark.

Generative Human Geometry Distribution

arXiv:2503.01448v3 Announce Type: replace Abstract: Realistic human geometry generation is an important yet challenging task, requiring both the preservation of fine clothing details and the accurate modeling of clothing-body interactions. To tackle this challenge, we build upon Geometry distributions, a recently proposed representation that can model a single human geometry with high fidelity using a flow matching model. However, extending a single-geometry distribution to a dataset is non-trivial and inefficient for large-scale learning. To address this, we propose a new geometry distribution model by two key techniques: (1) encoding distributions as 2D feature maps rather than network parameters, and (2) using SMPL models as the domain instead of Gaussian and refining the associated flow velocity field. We then design a generative framework adopting a two staged training paradigm analogous to state-of-the-art image and 3D generative models. In the first stage, we compress geometry distributions into a latent space using a diffusion flow model; the second stage trains another flow model on this latent space. We validate our approach on two key tasks: pose-conditioned random avatar generation and avatar-consistent novel pose synthesis. Experimental results demonstrate that our method outperforms existing state-of-the-art methods, achieving a 57% improvement in geometry quality.

Novel Object 6D Pose Estimation with a Single Reference View

arXiv:2503.05578v3 Announce Type: replace Abstract: Existing novel object 6D pose estimation methods typically rely on CAD models or dense reference views, which are both difficult to acquire. Using only a single reference view is more scalable, but challenging due to large pose discrepancies and limited geometric and spatial information. To address these issues, we propose a Single-Reference-based novel object 6D (SinRef-6D) pose estimation method. Our key idea is to iteratively establish point-wise alignment in a common coordinate system based on state space models (SSMs). Specifically, iterative object-space point-wise alignment can effectively handle large pose discrepancies, while our proposed RGB and Points SSMs can capture long-range dependencies and spatial information from a single view, offering linear complexity and superior spatial modeling capability. Once pre-trained on synthetic data, SinRef-6D can estimate the 6D pose of a novel object using only a single reference view, without requiring retraining or a CAD model. Extensive experiments on six popular datasets and real-world robotic scenes demonstrate that we achieve on-par performance with CAD-based and dense reference view-based methods, despite operating in the more challenging single reference setting. Code will be released at https://github.com/CNJianLiu/SinRef-6D.

PRO-VPT: Distribution-Adaptive Visual Prompt Tuning via Prompt Relocation

arXiv:2503.06901v2 Announce Type: replace Abstract: Visual prompt tuning (VPT), i.e., fine-tuning some lightweight prompt tokens, provides an efficient and effective approach for adapting pre-trained models to various downstream tasks. However, most prior art indiscriminately uses a fixed prompt distribution across different tasks, neglecting the importance of each block varying depending on the task. In this paper, we introduce adaptive distribution optimization (ADO) by tackling two key questions: (1) How to appropriately and formally define ADO, and (2) How to design an adaptive distribution strategy guided by this definition? Through empirical analysis, we first confirm that properly adjusting the distribution significantly improves VPT performance, and further uncover a key insight that a nested relationship exists between ADO and VPT. Based on these findings, we propose a new VPT framework, termed PRO-VPT (iterative Prompt RelOcation-based VPT), which adaptively adjusts the distribution built upon a nested optimization formulation. Specifically, we develop a prompt relocation strategy derived from this formulation, comprising two steps: pruning idle prompts from prompt-saturated blocks, followed by allocating these prompts to the most prompt-needed blocks. By iteratively performing prompt relocation and VPT, our proposal can adaptively learn the optimal prompt distribution in a nested optimization-based manner, thereby unlocking the full potential of VPT. Extensive experiments demonstrate that our proposal significantly outperforms advanced VPT methods, e.g., PRO-VPT surpasses VPT by 1.6 pp and 2.0 pp average accuracy, leading prompt-based methods to state-of-the-art performance on VTAB-1k and FGVC benchmarks. The code is available at https://github.com/ckshang/PRO-VPT.

FEB-Cache: Frequency-Guided Exposure Bias Reduction for Enhancing Diffusion Transformer Caching

arXiv:2503.07120v3 Announce Type: replace Abstract: Diffusion Transformer (DiT) has exhibited impressive generation capabilities but faces great challenges due to its high computational complexity. To address this issue, various methods, notably feature caching, have been introduced. However, these approaches focus on aligning non-cache diffusion without analyzing why caching damage the generation processes. In this paper, we first confirm that the cache greatly amplifies the exposure bias, resulting in a decline in the generation quality. However, directly applying noise scaling is challenging for this issue due to the non-smoothness of exposure bias. We found that this phenomenon stems from the mismatch between its frequency response characteristics and the simple cache of Attention and MLP. Since these two components exhibit unique preferences for frequency signals, which provides us with a caching strategy to separate Attention and MLP to achieve an enhanced fit of exposure bias and reduce it. Based on this, we introduced FEB-Cache, a joint caching strategy that aligns with the non-exposed bias diffusion process (which gives us a higher performance cap) of caching Attention and MLP based on the frequency-guided cache table. Our approach combines a comprehensive understanding of the caching mechanism and offers a new perspective on leveraging caching to accelerate the diffusion process. Empirical results indicate that FEB-Cache optimizes model performance while concurrently facilitating acceleration. Code is available at https://github.com/aSleepyTree/EB-Cache.

TRCE: Towards Reliable Malicious Concept Erasure in Text-to-Image Diffusion Models

arXiv:2503.07389v2 Announce Type: replace Abstract: Recent advances in text-to-image diffusion models enable photorealistic image generation, but they also risk producing malicious content, such as NSFW images. To mitigate risk, concept erasure methods are studied to facilitate the model to unlearn specific concepts. However, current studies struggle to fully erase malicious concepts implicitly embedded in prompts (e.g., metaphorical expressions or adversarial prompts) while preserving the model's normal generation capability. To address this challenge, our study proposes TRCE, using a two-stage concept erasure strategy to achieve an effective trade-off between reliable erasure and knowledge preservation. Firstly, TRCE starts by erasing the malicious semantics implicitly embedded in textual prompts. By identifying a critical mapping objective(i.e., the [EoT] embedding), we optimize the cross-attention layers to map malicious prompts to contextually similar prompts but with safe concepts. This step prevents the model from being overly influenced by malicious semantics during the denoising process. Following this, considering the deterministic properties of the sampling trajectory of the diffusion model, TRCE further steers the early denoising prediction toward the safe direction and away from the unsafe one through contrastive learning, thus further avoiding the generation of malicious content. Finally, we conduct comprehensive evaluations of TRCE on multiple malicious concept erasure benchmarks, and the results demonstrate its effectiveness in erasing malicious concepts while better preserving the model's original generation ability. The code is available at: http://github.com/ddgoodgood/TRCE. CAUTION: This paper includes model-generated content that may contain offensive material.

Exploring Representation Invariance in Finetuning

arXiv:2503.07399v3 Announce Type: replace Abstract: Foundation models pretrained on large-scale natural images are widely adapted to various cross-domain low-resource downstream tasks, benefiting from generalizable and transferable patterns captured by their representations. However, these representations are later found to gradually vanish during finetuning, accompanied by a degradation of model's original generalizability. In this paper, we argue that such tasks can be effectively adapted without sacrificing the benefits of pretrained representations. We approach this by introducing \textit{Representation Invariance FineTuning (RIFT)}, a regularization that maximizes the representation similarity between pretrained and finetuned models by leveraging orthogonal invariance of manifolds in a computationally efficient way. Experiments demonstrate that our method is compatible with mainstream finetuning methods, offering competitive or even enhanced performance and better preservation of the generalizability.

Explaining Human Preferences via Metrics for Structured 3D Reconstruction

arXiv:2503.08208v2 Announce Type: replace Abstract: "What cannot be measured cannot be improved" while likely never uttered by Lord Kelvin, summarizes effectively the driving force behind this work. This paper presents a detailed discussion of automated metrics for evaluating structured 3D reconstructions. Pitfalls of each metric are discussed, and an analysis through the lens of expert 3D modelers' preferences is presented. A set of systematic "unit tests" are proposed to empirically verify desirable properties, and context aware recommendations regarding which metric to use depending on application are provided. Finally, a learned metric distilled from human expert judgments is proposed and analyzed. The source code is available at https://github.com/s23dr/wireframe-metrics-iccv2025

Motion Blender Gaussian Splatting for Dynamic Scene Reconstruction

arXiv:2503.09040v4 Announce Type: replace Abstract: Gaussian splatting has emerged as a powerful tool for high-fidelity reconstruction of dynamic scenes. However, existing methods primarily rely on implicit motion representations, such as encoding motions into neural networks or per-Gaussian parameters, which makes it difficult to further manipulate the reconstructed motions. This lack of explicit controllability limits existing methods to replaying recorded motions only, which hinders a wider application in robotics. To address this, we propose Motion Blender Gaussian Splatting (MBGS), a novel framework that uses motion graphs as an explicit and sparse motion representation. The motion of a graph's links is propagated to individual Gaussians via dual quaternion skinning, with learnable weight painting functions that determine the influence of each link. The motion graphs and 3D Gaussians are jointly optimized from input videos via differentiable rendering. Experiments show that MBGS achieves state-of-the-art performance on the highly challenging iPhone dataset while being competitive on HyperNeRF. We demonstrate the application potential of our method in animating novel object poses, synthesizing real robot demonstrations, and predicting robot actions through visual planning. The source code, models, video demonstrations can be found at http://mlzxy.github.io/motion-blender-gs.

DecompDreamer: A Composition-Aware Curriculum for Structured 3D Asset Generation

arXiv:2503.11981v2 Announce Type: replace Abstract: Current text-to-3D methods excel at generating single objects but falter on compositional prompts. We argue this failure is fundamental to their optimization schedules, as simultaneous or iterative heuristics predictably collapse under a combinatorial explosion of conflicting gradients, leading to entangled geometry or catastrophic divergence. In this paper, we reframe the core challenge of compositional generation as one of optimization scheduling. We introduce DecompDreamer, a framework built on a novel staged optimization strategy that functions as an implicit curriculum. Our method first establishes a coherent structural scaffold by prioritizing inter-object relationships before shifting to the high-fidelity refinement of individual components. This temporal decoupling of competing objectives provides a robust solution to gradient conflict. Qualitative and quantitative evaluations on diverse compositional prompts demonstrate that DecompDreamer outperforms state-of-the-art methods in fidelity, disentanglement, and spatial coherence.

LIAM: Multimodal Transformer for Language Instructions, Images, Actions and Semantic Maps

arXiv:2503.12230v2 Announce Type: replace Abstract: The availability of large language models and open-vocabulary object perception methods enables more flexibility for domestic service robots. The large variability of domestic tasks can be addressed without implementing each task individually by providing the robot with a task description along with appropriate environment information. In this work, we propose LIAM - an end-to-end model that predicts action transcripts based on language, image, action, and map inputs. Language and image inputs are encoded with a CLIP backbone, for which we designed two pre-training tasks to fine-tune its weights and pre-align the latent spaces. We evaluate our method on the ALFRED dataset, a simulator-generated benchmark for domestic tasks. Our results demonstrate the importance of pre-aligning embedding spaces from different modalities and the efficacy of incorporating semantic maps.

AutoDrive-QA: A Multiple-Choice Benchmark for Vision-Language Evaluation in Urban Autonomous Driving

arXiv:2503.15778v2 Announce Type: replace Abstract: Evaluating vision-language models (VLMs) in urban driving contexts remains challenging, as existing benchmarks rely on open-ended responses that are ambiguous, annotation-intensive, and inconsistent to score. This lack of standardized evaluation slows progress toward safe and reliable AI for urban mobility. We introduce AutoDrive-QA, the first benchmark that systematically converts open-ended driving QA datasets (DriveLM, NuScenes-QA, LingoQA) into structured multiple-choice questions (MCQs) with distractors grounded in five realistic error categories: Driving Domain Misconceptions, Logical Inconsistencies, Misinterpreted Sensor Inputs, Computational Oversights, and Question Ambiguity. This framework enables reproducible and interpretable evaluation of VLMs across perception, prediction, and planning tasks in complex urban scenes. Experiments show that fine-tuning LLaVA-1.5-7B improves accuracy by about six percentage points across tasks, GPT-4V achieves the strongest zero-shot performance with up to 69.8% accuracy, and Qwen2-VL models also perform competitively, particularly in multi-view settings. Moreover, traditional metrics such as BLEU and CIDEr fail to distinguish strong from weak models. By providing an objective, domain-grounded evaluation protocol, AutoDrive-QA contributes to more transparent benchmarking of urban AI systems, supporting the development of safer and more trustworthy autonomous driving technologies for smart cities.

Jasmine: Harnessing Diffusion Prior for Self-supervised Depth Estimation

arXiv:2503.15905v2 Announce Type: replace Abstract: In this paper, we propose Jasmine, the first Stable Diffusion (SD)-based self-supervised framework for monocular depth estimation, which effectively harnesses SD's visual priors to enhance the sharpness and generalization of unsupervised prediction. Previous SD-based methods are all supervised since adapting diffusion models for dense prediction requires high-precision supervision. In contrast, self-supervised reprojection suffers from inherent challenges (e.g., occlusions, texture-less regions, illumination variance), and the predictions exhibit blurs and artifacts that severely compromise SD's latent priors. To resolve this, we construct a novel surrogate task of hybrid image reconstruction. Without any additional supervision, it preserves the detail priors of SD models by reconstructing the images themselves while preventing depth estimation from degradation. Furthermore, to address the inherent misalignment between SD's scale and shift invariant estimation and self-supervised scale-invariant depth estimation, we build the Scale-Shift GRU. It not only bridges this distribution gap but also isolates the fine-grained texture of SD output against the interference of reprojection loss. Extensive experiments demonstrate that Jasmine achieves SoTA performance on the KITTI benchmark and exhibits superior zero-shot generalization across multiple datasets.

MetaSpatial: Reinforcing 3D Spatial Reasoning in VLMs for the Metaverse

arXiv:2503.18470v2 Announce Type: replace Abstract: We present MetaSpatial, the first reinforcement learning (RL)-based framework designed to enhance 3D spatial reasoning in vision-language models (VLMs), enabling real-time 3D scene generation without the need for hard-coded optimizations. MetaSpatial addresses two core challenges: (i) the lack of internalized 3D spatial reasoning in VLMs, which limits their ability to generate realistic layouts, and (ii) the inefficiency of traditional supervised fine-tuning (SFT) for layout generation tasks, as perfect ground truth annotations are unavailable. Our key innovation is a multi-turn RL-based optimization mechanism that integrates physics-aware constraints and rendered image evaluations, ensuring generated 3D layouts are coherent, physically plausible, and aesthetically consistent. Methodologically, MetaSpatial introduces an adaptive, iterative reasoning process, where the VLM refines spatial arrangements over multiple turns by analyzing rendered outputs, improving scene coherence progressively. Empirical evaluations demonstrate that MetaSpatial significantly enhances the spatial consistency and formatting stability of various scale models. Post-training, object placements are more realistic, aligned, and functionally coherent, validating the effectiveness of RL for 3D spatial reasoning in metaverse, AR/VR, digital twins, and game development applications. Our code, data, and training pipeline are publicly available at https://github.com/PzySeere/MetaSpatial.

Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning of Vision Language Models

arXiv:2503.20752v3 Announce Type: replace Abstract: Visual reasoning abilities play a crucial role in understanding complex multimodal data, advancing both domain-specific applications and artificial general intelligence (AGI). Existing methods enhance Vision-Language Models (VLMs) through Chain-of-Thought (CoT) supervised fine-tuning using meticulously annotated data. However, this approach may lead to overfitting and cognitive rigidity, limiting the model's generalization ability under domain shifts and reducing real-world applicability. To overcome these limitations, we propose Reason-RFT, a two-stage reinforcement fine-tuning framework for visual reasoning. First, Supervised Fine-Tuning (SFT) with curated CoT data activates the reasoning potential of VLMs. This is followed by reinforcement learning based on Group Relative Policy Optimization (GRPO), which generates multiple reasoning-response pairs to enhance adaptability to domain shifts. To evaluate Reason-RFT, we reconstructed a comprehensive dataset covering visual counting, structural perception, and spatial transformation, serving as a benchmark for systematic assessment across three key dimensions. Experimental results highlight three advantages: (1) performance enhancement, with Reason-RFT achieving state-of-the-art results and outperforming both open-source and proprietary models; (2) generalization superiority, maintaining robust performance under domain shifts across various tasks; and (3) data efficiency, excelling in few-shot learning scenarios and surpassing full-dataset SFT baselines. Reason-RFT introduces a novel training paradigm for visual reasoning and marks a significant step forward in multimodal research. Project website: https://tanhuajie.github.io/ReasonRFT

FVQ: A Large-Scale Dataset and an LMM-based Method for Face Video Quality Assessment

arXiv:2504.09255v2 Announce Type: replace Abstract: Face video quality assessment (FVQA) deserves to be explored in addition to general video quality assessment (VQA), as face videos are the primary content on social media platforms and human visual system (HVS) is particularly sensitive to human faces. However, FVQA is rarely explored due to the lack of large-scale FVQA datasets. To fill this gap, we present the first large-scale in-the-wild FVQA dataset, FVQ-20K, which contains 20,000 in-the-wild face videos together with corresponding mean opinion score (MOS) annotations. Along with the FVQ-20K dataset, we further propose a specialized FVQA method named FVQ-Rater to achieve human-like rating and scoring for face video, which is the first attempt to explore the potential of large multimodal models (LMMs) for the FVQA task. Concretely, we elaborately extract multi-dimensional features including spatial features, temporal features, and face-specific features (i.e., portrait features and face embeddings) to provide comprehensive visual information, and take advantage of the LoRA-based instruction tuning technique to achieve quality-specific fine-tuning, which shows superior performance on both FVQ-20K and CFVQA datasets. Extensive experiments and comprehensive analysis demonstrate the significant potential of the FVQ-20K dataset and FVQ-Rater method in promoting the development of FVQA.

From Gaze to Insight: Bridging Human Visual Attention and Vision Language Model Explanation for Weakly-Supervised Medical Image Segmentation

arXiv:2504.11368v2 Announce Type: replace Abstract: Medical image segmentation remains challenging due to the high cost of pixel-level annotations for training. In the context of weak supervision, clinician gaze data captures regions of diagnostic interest; however, its sparsity limits its use for segmentation. In contrast, vision-language models (VLMs) provide semantic context through textual descriptions but lack the explanation precision required. Recognizing that neither source alone suffices, we propose a teacher-student framework that integrates both gaze and language supervision, leveraging their complementary strengths. Our key insight is that gaze data indicates where clinicians focus during diagnosis, while VLMs explain why those regions are significant. To implement this, the teacher model first learns from gaze points enhanced by VLM-generated descriptions of lesion morphology, establishing a foundation for guiding the student model. The teacher then directs the student through three strategies: (1) Multi-scale feature alignment to fuse visual cues with textual semantics; (2) Confidence-weighted consistency constraints to focus on reliable predictions; (3) Adaptive masking to limit error propagation in uncertain areas. Experiments on the Kvasir-SEG, NCI-ISBI, and ISIC datasets show that our method achieves Dice scores of 80.78%, 80.53%, and 84.22%, respectively-improving 3-5% over gaze baselines without increasing the annotation burden. By preserving correlations among predictions, gaze data, and lesion descriptions, our framework also maintains clinical interpretability. This work illustrates how integrating human visual attention with AI-generated semantic context can effectively overcome the limitations of individual weak supervision signals, thereby advancing the development of deployable, annotation-efficient medical AI systems. Code is available at: https://github.com/jingkunchen/FGI.

MambaMoE: Mixture-of-Spectral-Spatial-Experts State Space Model for Hyperspectral Image Classification

arXiv:2504.20509v2 Announce Type: replace Abstract: Mamba-based models have recently demonstrated significant potential in hyperspectral image (HSI) classification, primarily due to their ability to perform contextual modeling with linear computational complexity. However, existing Mamba-based approaches often overlook the directional modeling heterogeneity across different land-cover types, leading to limited classification performance. To address these limitations, we propose MambaMoE, a novel spectral-spatial Mixture-of-Experts (MoE) framework, which represents the first MoE-based approach in the HSI classification domain. Specifically, we design a Mixture of Mamba Expert Block (MoMEB) that performs adaptive spectral-spatial feature modeling via a sparse expert activation mechanism. Additionally, we introduce an uncertainty-guided corrective learning (UGCL) strategy that encourages the model to focus on complex regions prone to prediction ambiguity. This strategy dynamically samples supervision signals from regions with high predictive uncertainty, guiding the model to adaptively refine feature representations and thereby enhancing its focus on challenging areas. Extensive experiments conducted on multiple public HSI benchmark datasets show that MambaMoE achieves state-of-the-art performance in both classification accuracy and computational efficiency compared to existing advanced methods, particularly Mamba-based ones. The code will be available online at https://github.com/YichuXu/MambaMoE.

FreeInsert: Disentangled Text-Guided Object Insertion in 3D Gaussian Scene without Spatial Priors

arXiv:2505.01322v3 Announce Type: replace Abstract: Text-driven object insertion in 3D scenes is an emerging task that enables intuitive scene editing through natural language. However, existing 2D editing-based methods often rely on spatial priors such as 2D masks or 3D bounding boxes, and they struggle to ensure consistency of the inserted object. These limitations hinder flexibility and scalability in real-world applications. In this paper, we propose FreeInsert, a novel framework that leverages foundation models including MLLMs, LGMs, and diffusion models to disentangle object generation from spatial placement. This enables unsupervised and flexible object insertion in 3D scenes without spatial priors. FreeInsert starts with an MLLM-based parser that extracts structured semantics, including object types, spatial relationships, and attachment regions, from user instructions. These semantics guide both the reconstruction of the inserted object for 3D consistency and the learning of its degrees of freedom. We leverage the spatial reasoning capabilities of MLLMs to initialize object pose and scale. A hierarchical, spatially aware refinement stage further integrates spatial semantics and MLLM-inferred priors to enhance placement. Finally, the appearance of the object is improved using the inserted-object image to enhance visual fidelity. Experimental results demonstrate that FreeInsert achieves semantically coherent, spatially precise, and visually realistic 3D insertions without relying on spatial priors, offering a user-friendly and flexible editing experience.

MMLongBench: Benchmarking Long-Context Vision-Language Models Effectively and Thoroughly

arXiv:2505.10610v3 Announce Type: replace Abstract: The rapid extension of context windows in large vision-language models has given rise to long-context vision-language models (LCVLMs), which are capable of handling hundreds of images with interleaved text tokens in a single forward pass. In this work, we introduce MMLongBench, the first benchmark covering a diverse set of long-context vision-language tasks, to evaluate LCVLMs effectively and thoroughly. MMLongBench is composed of 13,331 examples spanning five different categories of downstream tasks, such as Visual RAG and Many-Shot ICL. It also provides broad coverage of image types, including various natural and synthetic images. To assess the robustness of the models to different input lengths, all examples are delivered at five standardized input lengths (8K-128K tokens) via a cross-modal tokenization scheme that combines vision patches and text tokens. Through a thorough benchmarking of 46 closed-source and open-source LCVLMs, we provide a comprehensive analysis of the current models' vision-language long-context ability. Our results show that: i) performance on a single task is a weak proxy for overall long-context capability; ii) both closed-source and open-source models face challenges in long-context vision-language tasks, indicating substantial room for future improvement; iii) models with stronger reasoning ability tend to exhibit better long-context performance. By offering wide task coverage, various image types, and rigorous length control, MMLongBench provides the missing foundation for diagnosing and advancing the next generation of LCVLMs.

Enhancing Transformers Through Conditioned Embedded Tokens

arXiv:2505.12789v2 Announce Type: replace Abstract: Transformers have transformed modern machine learning, driving breakthroughs in computer vision, natural language processing, and robotics. At the core of their success lies the attention mechanism, which enables the modeling of global dependencies among input tokens. However, we reveal that the attention block in transformers suffers from inherent ill-conditioning, which hampers gradient-based optimization and leads to inefficient training. To address this, we develop a theoretical framework that establishes a direct relationship between the conditioning of the attention block and that of the embedded tokenized data. Building on this insight, we introduce conditioned embedded tokens, a method that systematically modifies the embedded tokens to improve the conditioning of the attention mechanism. Our analysis demonstrates that this approach significantly mitigates ill-conditioning, leading to more stable and efficient training. We validate our methodology across various transformer architectures, achieving consistent improvements in image classification, object detection, instance segmentation, and natural language processing, highlighting its broad applicability and effectiveness.

RGB-to-Polarization Estimation: A New Task and Benchmark Study

arXiv:2505.13050v2 Announce Type: replace Abstract: Polarization images provide rich physical information that is fundamentally absent from standard RGB images, benefiting a wide range of computer vision applications such as reflection separation and material classification. However, the acquisition of polarization images typically requires additional optical components, which increases both the cost and the complexity of the applications. To bridge this gap, we introduce a new task: RGB-to-polarization image estimation, which aims to infer polarization information directly from RGB images. In this work, we establish the first comprehensive benchmark for this task by leveraging existing polarization datasets and evaluating a diverse set of state-of-the-art deep learning models, including both restoration-oriented and generative architectures. Through extensive quantitative and qualitative analysis, our benchmark not only establishes the current performance ceiling of RGB-to-polarization estimation, but also systematically reveals the respective strengths and limitations of different model families -- such as direct reconstruction versus generative synthesis, and task-specific training versus large-scale pre-training. In addition, we provide some potential directions for future research on polarization estimation. This benchmark is intended to serve as a foundational resource to facilitate the design and evaluation of future methods for polarization estimation from standard RGB inputs.

GeoRanker: Distance-Aware Ranking for Worldwide Image Geolocalization

arXiv:2505.13731v2 Announce Type: replace Abstract: Worldwide image geolocalization-the task of predicting GPS coordinates from images taken anywhere on Earth-poses a fundamental challenge due to the vast diversity in visual content across regions. While recent approaches adopt a two-stage pipeline of retrieving candidates and selecting the best match, they typically rely on simplistic similarity heuristics and point-wise supervision, failing to model spatial relationships among candidates. In this paper, we propose GeoRanker, a distance-aware ranking framework that leverages large vision-language models to jointly encode query-candidate interactions and predict geographic proximity. In addition, we introduce a multi-order distance loss that ranks both absolute and relative distances, enabling the model to reason over structured spatial relationships. To support this, we curate GeoRanking, the first dataset explicitly designed for geographic ranking tasks with multimodal candidate information. GeoRanker achieves state-of-the-art results on two well-established benchmarks (IM2GPS3K and YFCC4K), significantly outperforming current best methods.

Constructing a 3D Scene from a Single Image

arXiv:2505.15765v2 Announce Type: replace Abstract: Acquiring detailed 3D scenes typically demands costly equipment, multi-view data, or labor-intensive modeling. Therefore, a lightweight alternative, generating complex 3D scenes from a single top-down image, plays an essential role in real-world applications. While recent 3D generative models have achieved remarkable results at the object level, their extension to full-scene generation often leads to inconsistent geometry, layout hallucinations, and low-quality meshes. In this work, we introduce SceneFuse-3D, a training-free framework designed to synthesize coherent 3D scenes from a single top-down view. Our method is grounded in two principles: region-based generation to improve image-to-3D alignment and resolution, and spatial-aware 3D inpainting to ensure global scene coherence and high-quality geometry generation. Specifically, we decompose the input image into overlapping regions and generate each using a pretrained 3D object generator, followed by a masked rectified flow inpainting process that fills in missing geometry while maintaining structural continuity. This modular design allows us to overcome resolution bottlenecks and preserve spatial structure without requiring 3D supervision or fine-tuning. Extensive experiments across diverse scenes show that SceneFuse-3D outperforms state-of-the-art baselines, including Trellis, Hunyuan3D-2, TripoSG, and LGM, in terms of geometry quality, spatial coherence, and texture fidelity. Our results demonstrate that high-quality coherent 3D scene-level asset generation is achievable from a single top-down image using a principled, training-free pipeline.

Circle-RoPE: Cone-like Decoupled Rotary Positional Embedding for Large Vision-Language Models

arXiv:2505.16416v2 Announce Type: replace Abstract: Rotary Position Embedding (RoPE) is a widely adopted technique for encoding relative positional information in large language models (LLMs). However, when extended to vision-language models (VLMs), RoPE and its variants enforce relative positional dependencies separately within text and image tokens, introducing unintended cross-modal positional biases. For example, image tokens depicting semantically consistent content are assigned distinct positional encodings solely due to spatial location variations. As a result, such tokens exhibit entirely different relative positional relationships with their corresponding text tokens, ultimately leading to misaligned cross-modal representations. To address this, we propose Per-Token Distance, a simple yet effective metric for quantifying the independence of positional encodings across modalities. Informed by this analysis, we introduce Circle-RoPE, a novel encoding scheme designed to eliminate spurious cross-modal biases. Our key idea is to project image token indices onto a \emph{ring} that is orthogonal to the linear axis of text token indices, thereby forming a cone-like structure in the positional encoding space. In this configuration, each text token (point on the linear text axis) becomes the apex of a cone and maintains an equal distance to all image tokens (points on the circular image \emph{ring}), reducing artificial cross-modal biases while preserving intra-image spatial information. To further enhance performance, we propose a staggered strategy that applies different RoPE variants across layers. Extensive experiments demonstrate that our method effectively preserves spatial information from images while reducing relative positional bias, offering a more robust and flexible positional encoding framework for VLMs. The code is available at https://github.com/lose4578/CircleRoPE.

ViP$^2$-CLIP: Visual-Perception Prompting with Unified Alignment for Zero-Shot Anomaly Detection

arXiv:2505.17692v3 Announce Type: replace Abstract: Zero-shot anomaly detection (ZSAD) aims to detect anomalies without any target domain training samples, relying solely on external auxiliary data. Existing CLIP-based methods attempt to activate the model's ZSAD potential via handcrafted or static learnable prompts. The former incur high engineering costs and limited semantic coverage, whereas the latter apply identical descriptions across diverse anomaly types, thus fail to adapt to complex variations. Furthermore, since CLIP is originally pretrained on large-scale classification tasks, its anomaly segmentation quality is highly sensitive to the exact wording of class names, severely constraining prompting strategies that depend on class labels. To address these challenges, we introduce ViP$^{2}$-CLIP. The key insight of ViP$^{2}$-CLIP is a Visual-Perception Prompting (ViP-Prompt) mechanism, which fuses global and multi-scale local visual context to adaptively generate fine-grained textual prompts, eliminating manual templates and class-name priors. This design enables our model to focus on precise abnormal regions, making it particularly valuable when category labels are ambiguous or privacy-constrained. Extensive experiments on 15 industrial and medical benchmarks demonstrate that ViP$^{2}$-CLIP achieves state-of-the-art performance and robust cross-domain generalization.

AutoMiSeg: Automatic Medical Image Segmentation via Test-Time Adaptation of Foundation Models

arXiv:2505.17931v2 Announce Type: replace Abstract: Medical image segmentation is vital for clinical diagnosis, yet current deep learning methods often demand extensive expert effort, i.e., either through annotating large training datasets or providing prompts at inference time for each new case. This paper introduces a zero-shot and automatic segmentation pipeline that combines off-the-shelf vision-language and segmentation foundation models. Given a medical image and a task definition (e.g., "segment the optic disc in an eye fundus image"), our method uses a grounding model to generate an initial bounding box, followed by a visual prompt boosting module that enhance the prompts, which are then processed by a promptable segmentation model to produce the final mask. To address the challenges of domain gap and result verification, we introduce a test-time adaptation framework featuring a set of learnable adaptors that align the medical inputs with foundation model representations. Its hyperparameters are optimized via Bayesian Optimization, guided by a proxy validation model without requiring ground-truth labels. Our pipeline offers an annotation-efficient and scalable solution for zero-shot medical image segmentation across diverse tasks. Our pipeline is evaluated on seven diverse medical imaging datasets and shows promising results. By proper decomposition and test-time adaptation, our fully automatic pipeline not only substantially surpasses the previously best-performing method, yielding a 69\% relative improvement in accuracy (Dice Score from 42.53 to 71.81), but also performs competitively with weakly-prompted interactive foundation models.

Do We Need All the Synthetic Data? Targeted Synthetic Image Augmentation via Diffusion Models

arXiv:2505.21574v2 Announce Type: replace Abstract: Synthetically augmenting training datasets with diffusion models has been an effective strategy for improving generalization of image classifiers. However, existing techniques struggle to ensure the diversity of generation and increase the size of the data by up to 10-30x to improve the in-distribution performance. In this work, we show that synthetically augmenting part of the data that is not learned early in training with faithful images-containing same features but different noise-outperforms augmenting the entire dataset. By analyzing a two-layer CNN, we prove that this strategy improves generalization by promoting homogeneity in feature learning speed without amplifying noise. Our extensive experiments show that by augmenting only 30%-40% of the data, our method boosts generalization by up to 2.8% in a variety of scenarios, including training ResNet, ViT, ConvNeXt, and Swin Transformer on CIFAR-10/100, and TinyImageNet, with various optimizers including SGD and SAM. Notably, our method applied with SGD outperforms the SOTA optimizer, SAM, on CIFAR-100 and TinyImageNet.

CryoCCD: Conditional Cycle-consistent Diffusion with Biophysical Modeling for Cryo-EM Synthesis

arXiv:2505.23444v4 Announce Type: replace Abstract: Single-particle cryo-electron microscopy (cryo-EM) has become a cornerstone of structural biology, enabling near-atomic resolution analysis of macromolecules through advanced computational methods. However, the development of cryo-EM processing tools is constrained by the scarcity of high-quality annotated datasets. Synthetic data generation offers a promising alternative, but existing approaches lack thorough biophysical modeling of heterogeneity and fail to reproduce the complex noise observed in real imaging. To address these limitations, we present CryoCCD, a synthesis framework that unifies versatile biophysical modeling with the first conditional cycle-consistent diffusion model tailored for cryo-EM. The biophysical engine provides multi-functional generation capabilities to capture authentic biological organization, and the diffusion model is enhanced with cycle consistency and mask-guided contrastive learning to ensure realistic noise while preserving structural fidelity. Extensive experiments demonstrate that CryoCCD generates structurally faithful micrographs, enhances particle picking and pose estimation, as well as achieves superior performance over state-of-the-art baselines, while also generalizing effectively to held-out protein families.

DS-VTON: An Enhanced Dual-Scale Coarse-to-Fine Framework for Virtual Try-On

arXiv:2506.00908v2 Announce Type: replace Abstract: Despite recent progress, most existing virtual try-on methods still struggle to simultaneously address two core challenges: accurately aligning the garment image with the target human body, and preserving fine-grained garment textures and patterns. These two requirements map directly onto a coarse-to-fine generation paradigm, where the coarse stage handles structural alignment and the fine stage recovers rich garment details. Motivated by this observation, we propose DS-VTON, an enhanced dual-scale coarse-to-fine framework that tackles the try-on problem more effectively. DS-VTON consists of two stages: the first stage generates a low-resolution try-on result to capture the semantic correspondence between garment and body, where reduced detail facilitates robust structural alignment. In the second stage, a blend-refine diffusion process reconstructs high-resolution outputs by refining the residual between scales through noise-image blending, emphasizing texture fidelity and effectively correcting fine-detail errors from the low-resolution stage. In addition, our method adopts a fully mask-free generation strategy, eliminating reliance on human parsing maps or segmentation masks. Extensive experiments show that DS-VTON not only achieves state-of-the-art performance but consistently and significantly surpasses prior methods in both structural alignment and texture fidelity across multiple standard virtual try-on benchmarks.

Resolving Task Objective Conflicts in Unified Model via Task-Aware Mixture-of-Experts

arXiv:2506.03591v2 Announce Type: replace Abstract: Unified multimodal large language models (MLLMs) based on end-to-end autoregressive (AR) transformers effectively integrate both understanding and generation tasks within a single framework. However, intrinsic Task Objective Conflicts between high-level semantic abstraction in understanding and fine-grained detail preservation in generation pose significant challenges, often leading to suboptimal trade-offs and task interference. Existing solutions, such as decoupling shared visual encoders, fall short of fundamentally resolving these conflicts due to inherent AR architecture. In this paper, we propose a novel approach that decouples internal components of AR to resolve task objective conflicts. Specifically, we design UTAMoE, a Unified Task-Aware Mixture-of-Experts (MoE) framework that decouples internal AR modules via a Task-Aware MoE Layer to create task-specific optimization subpaths. To enhance task differentiation while maintaining overall coordination, we introduce a novel Two-Stage Training Strategy. Extensive experiments on multimodal benchmarks demonstrate that UTAMoE mitigates task objective conflicts, achieving state-of-the-art performance across various tasks. Visualizations and ablation studies further validate the effectiveness of our approach.

A Neurosymbolic Agent System for Compositional Visual Reasoning

arXiv:2506.07778v3 Announce Type: replace Abstract: The advancement in large language models (LLMs) and large vision models has fueled the rapid progress in multi-modal vision-language reasoning capabilities. However, existing vision-language models (VLMs) remain challenged by compositional visual reasoning. This paper presents VLAgent, a neuro-symbolic approach to developing a Vision-Language Agent system for efficient compositional visual reasoning with three novel features. First, VLAgent develops an interpretable visualization-enhanced two-stage neuro-symbolic reasoning system. The first stage is managed by a front-end engine that generates a structured visual reasoning plan (symbolic program script) for each compositional visual reasoning task by utilizing a pre-trained LLM powered with few-shot chain-of-thought in-context learning. The second stage is managed by a high-performance back-end engine. It transforms the planning script into executable code based on visual input (image or video) and the combination of neural models and symbolic functions and then performs a sequence of actions for the compositional visual reason task. Second, to ensure and enhance the quality of mapping the logic plan to a sequence of executable instructions, VLAgent introduces the SS-parser, which examines the syntax and semantic correctness of the planning script, detects and repairs the logic errors found in the LLM-generated logic plan before generating the executable program. Third, VLAgent introduces the execution verifier in critical reasoning steps to validate and refine its compositional reasoning results in a stepwise manner, for example, ensemble methods for critical visual reasoning and caption analysis for low-confidence compositional reasoning. Extensive experiments on six visual benchmarks compared to a dozen SoTA visual reasoning models show that VLAgent outperforms existing representative approaches to compositional visual reasoning.

RoboSwap: A GAN-driven Video Diffusion Framework For Unsupervised Robot Arm Swapping

arXiv:2506.08632v2 Announce Type: replace Abstract: Recent advancements in generative models have revolutionized video synthesis and editing. However, the scarcity of diverse, high-quality datasets continues to hinder video-conditioned robotic learning, limiting cross-platform generalization. In this work, we address the challenge of swapping a robotic arm in one video with another: a key step for crossembodiment learning. Unlike previous methods that depend on paired video demonstrations in the same environmental settings, our proposed framework, RoboSwap, operates on unpaired data from diverse environments, alleviating the data collection needs. RoboSwap introduces a novel video editing pipeline integrating both GANs and diffusion models, combining their isolated advantages. Specifically, we segment robotic arms from their backgrounds and train an unpaired GAN model to translate one robotic arm to another. The translated arm is blended with the original video background and refined with a diffusion model to enhance coherence, motion realism and object interaction. The GAN and diffusion stages are trained independently. Our experiments demonstrate that RoboSwap outperforms state-of-the-art video and image editing models on three benchmarks in terms of both structural coherence and motion consistency, thereby offering a robust solution for generating reliable, cross-embodiment data in robotic learning.

WetCat: Enabling Automated Skill Assessment in Wet-Lab Cataract Surgery Videos

arXiv:2506.08896v4 Announce Type: replace Abstract: To meet the growing demand for systematic surgical training, wet-lab environments have become indispensable platforms for hands-on practice in ophthalmology. Yet, traditional wet-lab training depends heavily on manual performance evaluations, which are labor-intensive, time-consuming, and often subject to variability. Recent advances in computer vision offer promising avenues for automated skill assessment, enhancing both the efficiency and objectivity of surgical education. Despite notable progress in ophthalmic surgical datasets, existing resources predominantly focus on real surgeries or isolated tasks, falling short of supporting comprehensive skill evaluation in controlled wet-lab settings. To address these limitations, we introduce WetCat, the first dataset of wet-lab cataract surgery videos specifically curated for automated skill assessment. WetCat comprises high-resolution recordings of surgeries performed by trainees on artificial eyes, featuring comprehensive phase annotations and semantic segmentations of key anatomical structures. These annotations are meticulously designed to facilitate skill assessment during the critical capsulorhexis and phacoemulsification phases, adhering to standardized surgical skill assessment frameworks. By focusing on these essential phases, WetCat enables the development of interpretable, AI-driven evaluation tools aligned with established clinical metrics. This dataset lays a strong foundation for advancing objective, scalable surgical education and sets a new benchmark for automated workflow analysis and skill assessment in ophthalmology training. The dataset and annotations are publicly available in Synapse.

Attention, Please! Revisiting Attentive Probing Through the Lens of Efficiency

arXiv:2506.10178v2 Announce Type: replace Abstract: As fine-tuning becomes increasingly impractical at scale, probing is emerging as the preferred evaluation protocol. Yet, the standard linear probing fails to adequately reflect the potential of models whose pre-training optimizes representations of patch tokens rather than an explicit global representation. This motivates the need for attentive probing, an alternative that uses attention to selectively aggregate patch-level features. Despite its growing adoption, attentive probing remains under-explored, with existing methods suffering from excessive parameterization and poor computational efficiency. In this work, we revisit attentive probing through the lens of the accuracy vs. parameter efficiency trade-off. We present the first comprehensive study of existing methods, analyzing their design choices and benchmarking their performance. Building on this, we propose efficient probing (EP), a simple yet effective multi-query cross-attention mechanism that eliminates redundant projections and reduces the number of trainable parameters. Despite its simplicity, EP outperforms linear probing and prior attentive probing approaches across seven benchmarks, generalizes well to diverse pre-training paradigms, and delivers strong low-shot and layer-wise gains. Beyond evaluation, our analysis uncovers emerging properties of EP, such as complementary attention maps, which open new directions for leveraging probing beyond protocol design. Code available at https://github.com/billpsomas/efficient-probing.

SIRI-Bench: Challenging VLMs' Spatial Intelligence through Complex Reasoning Tasks

arXiv:2506.14512v2 Announce Type: replace Abstract: Large Language Models (LLMs) have undergone rapid progress, largely attributed to reinforcement learning on complex reasoning tasks. In contrast, while spatial intelligence is fundamental for Vision-Language Models (VLMs) in real-world interaction, the systematic study of their complex spatial reasoning remains underexplored. To bridge this gap, we introduce SIRI-Bench, a benchmark designed to evaluate VLMs' structural spatial intelligence through spatial-grounded reasoning tasks. SIRI-Bench comprises 9,000 video-question-answer triplets, where each problem is embedded in a realistic 3D scene. The benchmark is carefully designed so that solving each problem requires both spatial comprehension and structural reasoning. To facilitate large-scale data synthesis, we develop an Automatic Scene Creation Engine that employs collaborative LLM agents to translate abstract mathematical problems into faithful 3D scenes. Experimental results reveal that state-of-the-art VLMs struggle significantly on SIRI-Bench, underscoring the challenge of structural spatial reasoning. We hope that our study will bring researchers' attention to spatially grounded reasoning and advance VLMs in visual problem-solving.

VisualChef: Generating Visual Aids in Cooking via Mask Inpainting

arXiv:2506.18569v2 Announce Type: replace Abstract: Cooking requires not only following instructions but also understanding, executing, and monitoring each step - a process that can be challenging without visual guidance. Although recipe images and videos offer helpful cues, they often lack consistency in focus, tools, and setup. To better support the cooking process, we introduce VisualChef, a method for generating contextual visual aids tailored to cooking scenarios. Given an initial frame and a specified action, VisualChef generates images depicting both the action's execution and the resulting appearance of the object, while preserving the initial frame's environment. Previous work aims to integrate knowledge extracted from large language models by generating detailed textual descriptions to guide image generation, which requires fine-grained visual-textual alignment and involves additional annotations. In contrast, VisualChef simplifies alignment through mask-based visual grounding. Our key insight is identifying action-relevant objects and classifying them to enable targeted modifications that reflect the intended action and outcome while maintaining a consistent environment. In addition, we propose an automated pipeline to extract high-quality initial, action, and final state frames. We evaluate VisualChef quantitatively and qualitatively on three egocentric video datasets and show its improvements over state-of-the-art methods.

Light of Normals: Unified Feature Representation for Universal Photometric Stereo

arXiv:2506.18882v4 Announce Type: replace Abstract: Universal photometric stereo (PS) is defined by two factors: it must (i) operate under arbitrary, unknown lighting conditions and (ii) avoid reliance on specific illumination models. Despite progress (e.g., SDM UniPS), two challenges remain. First, current encoders cannot guarantee that illumination and normal information are decoupled. To enforce decoupling, we introduce LINO UniPS with two key components: (i) Light Register Tokens with light alignment supervision to aggregate point, direction, and environment lights; (ii) Interleaved Attention Block featuring global cross-image attention that takes all lighting conditions together so the encoder can factor out lighting while retaining normal-related evidence. Second, high-frequency geometric details are easily lost. We address this with (i) a Wavelet-based Dual-branch Architecture and (ii) a Normal-gradient Perception Loss. These techniques yield a unified feature space in which lighting is explicitly represented by register tokens, while normal details are preserved via wavelet branch. We further introduce PS-Verse, a large-scale synthetic dataset graded by geometric complexity and lighting diversity, and adopt curriculum training from simple to complex scenes. Extensive experiments show new state-of-the-art results on public benchmarks (e.g., DiLiGenT, Luces), stronger generalization to real materials, and improved efficiency; ablations confirm that Light Register Tokens + Interleaved Attention Block drive better feature decoupling, while Wavelet-based Dual-branch Architecture + Normal-gradient Perception Loss recover finer details.

ImplicitQA: Going beyond frames towards Implicit Video Reasoning

arXiv:2506.21742v2 Announce Type: replace Abstract: Video Question Answering (VideoQA) has made significant strides by leveraging multimodal learning to align visual and textual modalities. However, current benchmarks overwhelmingly focus on questions answerable through explicit visual content - actions, objects, and events directly observable within individual frames or short clips. In contrast, creative and cinematic videos - such as movies, TV shows, and narrative-driven content - employ storytelling techniques that deliberately omit certain depictions, requiring viewers to infer motives, relationships across discontinuous frames with disjoint visual contexts. Humans naturally excel at such implicit reasoning, seamlessly integrating information across time and context to construct coherent narratives. Yet current benchmarks fail to capture this essential dimension of human-like understanding. To bridge this gap, we present ImplicitQA, a novel benchmark specifically designed to test VideoQA models on human-like implicit reasoning. ImplicitQA comprises 1K meticulously annotated QA pairs drawn from 1K high-quality creative video clips covering 15 genres across 7 decades of content. Questions are systematically categorized into nine key reasoning dimensions: lateral and vertical spatial reasoning, depth and proximity, viewpoint and visibility, motion and trajectory, causal and motivational reasoning, social interactions, physical context, and inferred counting. These annotations are deliberately challenging, crafted by authors, validated through multiple annotators, and benchmarked against human performance to ensure high quality. Our extensive evaluations on 11 leading VideoQA models reveals consistent and significant performance degradation, underscoring their reliance on surface-level visual cues and highlighting the difficulty of implicit reasoning. https://huggingface.co/datasets/ucf-crcv/ImplicitQA.

VSRM: A Robust Mamba-Based Framework for Video Super-Resolution

arXiv:2506.22762v3 Announce Type: replace Abstract: Video super-resolution remains a major challenge in low-level vision tasks. To date, CNN- and Transformer-based methods have delivered impressive results. However, CNNs are limited by local receptive fields, while Transformers struggle with quadratic complexity, posing challenges for processing long sequences in VSR. Recently, Mamba has drawn attention for its long-sequence modeling, linear complexity, and large receptive fields. In this work, we propose VSRM, a novel \textbf{V}ideo \textbf{S}uper-\textbf{R}esolution framework that leverages the power of \textbf{M}amba. VSRM introduces Spatial-to-Temporal Mamba and Temporal-to-Spatial Mamba blocks to extract long-range spatio-temporal features and enhance receptive fields efficiently. To better align adjacent frames, we propose Deformable Cross-Mamba Alignment module. This module utilizes a deformable cross-mamba mechanism to make the compensation stage more dynamic and flexible, preventing feature distortions. Finally, we minimize the frequency domain gaps between reconstructed and ground-truth frames by proposing a simple yet effective Frequency Charbonnier-like loss that better preserves high-frequency content and enhances visual quality. Through extensive experiments, VSRM achieves state-of-the-art results on diverse benchmarks, establishing itself as a solid foundation for future research.

Comprehensive Evaluation of Large Multimodal Models for Nutrition Analysis: A New Benchmark Enriched with Contextual Metadata

arXiv:2507.07048v2 Announce Type: replace Abstract: Large Multimodal Models (LMMs) are increasingly applied to meal images for nutrition analysis. However, existing work primarily evaluates proprietary models, such as GPT-4. This leaves the broad range of LLMs underexplored. Additionally, the influence of integrating contextual metadata and its interaction with various reasoning modifiers remains largely uncharted. This work investigates how interpreting contextual metadata derived from GPS coordinates (converted to location/venue type), timestamps (transformed into meal/day type), and the food items present can enhance LMM performance in estimating key nutritional values. These values include calories, macronutrients (protein, carbohydrates, fat), and portion sizes. We also introduce \textbf{ACETADA}, a new food-image dataset slated for public release. This open dataset provides nutrition information verified by the dietitian and serves as the foundation for our analysis. Our evaluation across eight LMMs (four open-weight and four closed-weight) first establishes the benefit of contextual metadata integration over straightforward prompting with images alone. We then demonstrate how this incorporation of contextual information enhances the efficacy of reasoning modifiers, such as Chain-of-Thought, Multimodal Chain-of-Thought, Scale Hint, Few-Shot, and Expert Persona. Empirical results show that integrating metadata intelligently, when applied through straightforward prompting strategies, can significantly reduce the Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) in predicted nutritional values. This work highlights the potential of context-aware LMMs for improved nutrition analysis.

Divergence Minimization Preference Optimization for Diffusion Model Alignment

arXiv:2507.07510v2 Announce Type: replace Abstract: Diffusion models have achieved remarkable success in generating realistic and versatile images from text prompts. Inspired by the recent advancements of language models, there is an increasing interest in further improving the models by aligning with human preferences. However, we investigate alignment from a divergence minimization perspective and reveal that existing preference optimization methods are typically trapped in suboptimal mean-seeking optimization. In this paper, we introduce Divergence Minimization Preference Optimization (DMPO), a novel and principled method for aligning diffusion models by minimizing reverse KL divergence, which asymptotically enjoys the same optimization direction as original RL. We provide rigorous analysis to justify the effectiveness of DMPO and conduct comprehensive experiments to validate its empirical strength across both human evaluations and automatic metrics. Our extensive results show that diffusion models fine-tuned with DMPO can consistently outperform or match existing techniques, specifically consistently outperforming all baseline models across different base models and test sets, achieving the best PickScore in every case, demonstrating the method's superiority in aligning generative behavior with desired outputs. Overall, DMPO unlocks a robust and elegant pathway for preference alignment, bridging principled theory with practical performance in diffusion models.

SurfDist: Interpretable Three-Dimensional Instance Segmentation Using Curved Surface Patches

arXiv:2507.08223v2 Announce Type: replace Abstract: We present SurfDist, a convolutional neural network architecture for three-dimensional volumetric instance segmentation. SurfDist enables prediction of instances represented as closed surfaces composed of smooth parametric surface patches, specifically bicubic B\'ezier triangles. SurfDist is a modification of the popular model architecture StarDist-3D which breaks StarDist-3D's coupling of instance parameterization dimension and instance voxel resolution, and it produces predictions which may be upsampled to arbitrarily high resolutions without introduction of voxelization artifacts. For datasets with blob-shaped instances, common in biomedical imaging, SurfDist can outperform StarDist-3D with more compact instance parameterizations. We detail SurfDist's technical implementation and show one synthetic and one real-world dataset for which it outperforms StarDist-3D. These results demonstrate that interpretable instance surface models can be learned effectively alongside instance membership.

UGC-VideoCaptioner: An Omni UGC Video Detail Caption Model and New Benchmarks

arXiv:2507.11336v2 Announce Type: replace Abstract: Real-world user-generated videos, especially on platforms like TikTok, often feature rich and intertwined audio visual content. However, existing video captioning benchmarks and models remain predominantly visual centric, overlooking the crucial role of audio in conveying scene dynamics, speaker intent, and narrative context. This lack of omni datasets and lightweight, capable models hampers progress in fine grained, multimodal video understanding. To address these challenges, we introduce UGC-VideoCap, a new benchmark and model framework specifically designed for detailed omnimodal captioning of short form user-generated videos. Unlike prior datasets, UGC-VideoCap emphasizes balanced integration of audio and visual modalities, featuring 1000 TikTok videos annotated through a structured three stage human-in-the-loop pipeline covering audio only, visual only, and joint audio visual semantics. The benchmark also includes 4000 carefully crafted QA pairs probing both unimodal and cross modal understanding. Alongside the dataset, we propose UGC-VideoCaptioner(3B), a 3B parameter captioning model distilled from Gemini 2.5 Flash. Using a novel two-stage training strategy supervised fine tuning followed by Group Relative Policy Optimization (GRPO), our approach enables efficient adaptation from limited data while maintaining competitive performance. Together, our benchmark and model offer a high-quality foundation and a data-efficient solution for advancing omnimodal video captioning in unconstrained real-world UGC settings.

DiViD: Disentangled Video Diffusion for Static-Dynamic Factorization

arXiv:2507.13934v2 Announce Type: replace Abstract: Unsupervised disentanglement of static appearance and dynamic motion in video remains a fundamental challenge, often hindered by information leakage and blurry reconstructions in existing VAE- and GAN-based approaches. We introduce DiViD, the first end-to-end video diffusion framework for explicit static-dynamic factorization. DiViD's sequence encoder extracts a global static token from the first frame and per-frame dynamic tokens, explicitly removing static content from the motion code. Its conditional DDPM decoder incorporates three key inductive biases: a shared-noise schedule for temporal consistency, a time-varying KL-based bottleneck that tightens at early timesteps (compressing static information) and relaxes later (enriching dynamics), and cross-attention that routes the global static token to all frames while keeping dynamic tokens frame-specific. An orthogonality regularizer further prevents residual static-dynamic leakage. We evaluate DiViD on real-world benchmarks using swap-based accuracy and cross-leakage metrics. DiViD outperforms state-of-the-art sequential disentanglement methods: it achieves the highest swap-based joint accuracy, preserves static fidelity while improving dynamic transfer, and reduces average cross-leakage.

SIA: Enhancing Safety via Intent Awareness for Vision-Language Models

arXiv:2507.16856v2 Announce Type: replace Abstract: With the growing deployment of Vision-Language Models (VLMs) in real-world applications, previously overlooked safety risks are becoming increasingly evident. In particular, seemingly innocuous multimodal inputs can combine to reveal harmful intent, leading to unsafe model outputs. While multimodal safety has received increasing attention, existing approaches often fail to address such latent risks, especially when harmfulness arises only from the interaction between modalities. We propose SIA (Safety via Intent Awareness), a training-free, intent-aware safety framework that proactively detects harmful intent in multimodal inputs and uses it to guide the generation of safe responses. SIA follows a three-stage process: (1) visual abstraction via captioning; (2) intent inference through few-shot chain-of-thought (CoT) prompting; and (3) intent-conditioned response generation. By dynamically adapting to the implicit intent inferred from an image-text pair, SIA mitigates harmful outputs without extensive retraining. Extensive experiments on safety benchmarks, including SIUO, MM-SafetyBench, and HoliSafe, show that SIA consistently improves safety and outperforms prior training-free methods.

SAR-TEXT: A Large-Scale SAR Image-Text Dataset Built with SAR-Narrator and A Progressive Learning Strategy for Downstream Tasks

arXiv:2507.18743v3 Announce Type: replace Abstract: Vision Language Models (VLMs) have achieved remarkable breakthroughs in the field of remote sensing in recent years. Synthetic Aperture Radar (SAR) imagery, with its all-weather capability, is essential in remote sensing, yet the lack of large-scale, high-quality SAR image-text datasets hinders its semantic understanding. In this paper, we construct SAR-TEXT, a large-scale and high-quality dataset consisting of over 130,000 SAR image-text pairs. To construct the SAR-TEXT dataset, we design the SAR-Narrator framework, which generates textual descriptions for SAR images through a multi-stage strategy. To verify the effectiveness of the SAR-TEXT dataset, we conduct experiments on three typical vision-language tasks: image-text retrieval, image captioning, and visual question answering (VQA). Specifically, we construct three representative models on SAR-TEXT: SAR-RS-CLIP, SAR-RS-CoCa, and SAR-GPT. SAR-RS-CLIP achieves notable improvements in retrieval performance, boosting average recall by 12.97% and 10.0% on the OSdataset_512 and HRSID test sets, respectively. In the captioning task, SAR-RS-CoCa achieves significant improvements over the original CoCa models in terms of BLEU-4, SPICE, and CIDEr scores. In the VQA task, SAR-GPT outperforms baseline and single-stage models on multiple SAR-VQA datasets, demonstrating stronger semantic understanding and reasoning ability, as further confirmed by qualitative results. It is worth noting that, as a flexible captioning tool, SAR-Narrator can be readily adopted by the community to construct larger-scale SAR image-text datasets. All code, pretrained models, and the SAR-Text dataset are publicly available at: https://github.com/YiguoHe/SAR-TEXT.

RESCUE: Crowd Evacuation Simulation via Controlling SDM-United Characters

arXiv:2507.20117v2 Announce Type: replace Abstract: Crowd evacuation simulation is critical for enhancing public safety, and demanded for realistic virtual environments. Current mainstream evacuation models overlook the complex human behaviors that occur during evacuation, such as pedestrian collisions, interpersonal interactions, and variations in behavior influenced by terrain types or individual body shapes. This results in the failure to accurately simulate the escape of people in the real world. In this paper, aligned with the sensory-decision-motor (SDM) flow of the human brain, we propose a real-time 3D crowd evacuation simulation framework that integrates a 3D-adaptive SFM (Social Force Model) Decision Mechanism and a Personalized Gait Control Motor. This framework allows multiple agents to move in parallel and is suitable for various scenarios, with dynamic crowd awareness. Additionally, we introduce Part-level Force Visualization to assist in evacuation analysis. Experimental results demonstrate that our framework supports dynamic trajectory planning and personalized behavior for each agent throughout the evacuation process, and is compatible with uneven terrain. Visually, our method generates evacuation results that are more realistic and plausible, providing enhanced insights for crowd simulation. The code is available at http://cic.tju.edu.cn/faculty/likun/projects/RESCUE.

ReMoMask: Retrieval-Augmented Masked Motion Generation

arXiv:2508.02605v2 Announce Type: replace Abstract: Text-to-Motion (T2M) generation aims to synthesize realistic and semantically aligned human motion sequences from natural language descriptions. However, current approaches face dual challenges: Generative models (e.g., diffusion models) suffer from limited diversity, error accumulation, and physical implausibility, while Retrieval-Augmented Generation (RAG) methods exhibit diffusion inertia, partial-mode collapse, and asynchronous artifacts. To address these limitations, we propose ReMoMask, a unified framework integrating three key innovations: 1) A Bidirectional Momentum Text-Motion Model decouples negative sample scale from batch size via momentum queues, substantially improving cross-modal retrieval precision; 2) A Semantic Spatio-temporal Attention mechanism enforces biomechanical constraints during part-level fusion to eliminate asynchronous artifacts; 3) RAG-Classier-Free Guidance incorporates minor unconditional generation to enhance generalization. Built upon MoMask's RVQ-VAE, ReMoMask efficiently generates temporally coherent motions in minimal steps. Extensive experiments on standard benchmarks demonstrate the state-of-the-art performance of ReMoMask, achieving a 3.88% and 10.97% improvement in FID scores on HumanML3D and KIT-ML, respectively, compared to the previous SOTA method RAG-T2M. Code: https://github.com/AIGeeksGroup/ReMoMask. Website: https://aigeeksgroup.github.io/ReMoMask.

Follow-Your-Shape: Shape-Aware Image Editing via Trajectory-Guided Region Control

arXiv:2508.08134v3 Announce Type: replace Abstract: While recent flow-based image editing models demonstrate general-purpose capabilities across diverse tasks, they often struggle to specialize in challenging scenarios -- particularly those involving large-scale shape transformations. When performing such structural edits, these methods either fail to achieve the intended shape change or inadvertently alter non-target regions, resulting in degraded background quality. We propose Follow-Your-Shape, a training-free and mask-free framework that supports precise and controllable editing of object shapes while strictly preserving non-target content. Motivated by the divergence between inversion and editing trajectories, we compute a Trajectory Divergence Map (TDM) by comparing token-wise velocity differences between the inversion and denoising paths. The TDM enables precise localization of editable regions and guides a Scheduled KV Injection mechanism that ensures stable and faithful editing. To facilitate a rigorous evaluation, we introduce ReShapeBench, a new benchmark comprising 120 new images and enriched prompt pairs specifically curated for shape-aware editing. Experiments demonstrate that our method achieves superior editability and visual fidelity, particularly in tasks requiring large-scale shape replacement.

Deep Spectral Epipolar Representations for Dense Light Field Reconstruction

arXiv:2508.08900v2 Announce Type: replace Abstract: Accurate and efficient dense depth reconstruction from light field imagery remains a central challenge in computer vision, underpinning applications such as augmented reality, biomedical imaging, and 3D scene reconstruction. Existing deep convolutional approaches, while effective, often incur high computational overhead and are sensitive to noise and disparity inconsistencies in real-world scenarios. This paper introduces a novel Deep Spectral Epipolar Representation (DSER) framework for dense light field reconstruction, which unifies deep spectral feature learning with epipolar-domain regularization. The proposed approach exploits frequency-domain correlations across epipolar plane images to enforce global structural coherence, thereby mitigating artifacts and enhancing depth accuracy. Unlike conventional supervised models, DSER operates efficiently with limited training data while maintaining high reconstruction fidelity. Comprehensive experiments on the 4D Light Field Benchmark and a diverse set of real-world datasets demonstrate that DSER achieves superior performance in terms of precision, structural consistency, and computational efficiency compared to state-of-the-art methods. These results highlight the potential of integrating spectral priors with epipolar geometry for scalable and noise-resilient dense light field depth estimation, establishing DSER as a promising direction for next-generation high-dimensional vision systems.

Event-driven Robust Fitting on Neuromorphic Hardware

arXiv:2508.09466v2 Announce Type: replace Abstract: Robust fitting of geometric models is a fundamental task in many computer vision pipelines. Numerous innovations have been produced on the topic, from improving the efficiency and accuracy of random sampling heuristics to generating novel theoretical insights that underpin new approaches with mathematical guarantees. However, one aspect of robust fitting that has received little attention is energy efficiency. This performance metric has become critical as high energy consumption is a growing concern for AI adoption. In this paper, we explore energy-efficient robust fitting via the neuromorphic computing paradigm. Specifically, we designed a novel spiking neural network for robust fitting on real neuromorphic hardware, the Intel Loihi 2. Enabling this are novel event-driven formulations of model estimation that allow robust fitting to be implemented in the unique architecture of Loihi 2, and algorithmic strategies to alleviate the current limited precision and instruction set of the hardware. Results show that our neuromorphic robust fitting consumes only a fraction (15%) of the energy required to run the established robust fitting algorithm on a standard CPU to equivalent accuracy.

TSLA: A Task-Specific Learning Adaptation for Semantic Segmentation on Autonomous Vehicles Platform

arXiv:2508.12279v2 Announce Type: replace Abstract: Autonomous driving platforms encounter diverse driving scenarios, each with varying hardware resources and precision requirements. Given the computational limitations of embedded devices, it is crucial to consider computing costs when deploying on target platforms like the NVIDIA\textsuperscript{\textregistered} DRIVE PX 2. Our objective is to customize the semantic segmentation network according to the computing power and specific scenarios of autonomous driving hardware. We implement dynamic adaptability through a three-tier control mechanism -- width multiplier, classifier depth, and classifier kernel -- allowing fine-grained control over model components based on hardware constraints and task requirements. This adaptability facilitates broad model scaling, targeted refinement of the final layers, and scenario-specific optimization of kernel sizes, leading to improved resource allocation and performance. Additionally, we leverage Bayesian Optimization with surrogate modeling to efficiently explore hyperparameter spaces under tight computational budgets. Our approach addresses scenario-specific and task-specific requirements through automatic parameter search, accommodating the unique computational complexity and accuracy needs of autonomous driving. It scales its Multiply-Accumulate Operations (MACs) for Task-Specific Learning Adaptation (TSLA), resulting in alternative configurations tailored to diverse self-driving tasks. These TSLA customizations maximize computational capacity and model accuracy, optimizing hardware utilization.

The Telephone Game: Evaluating Semantic Drift in Unified Models

arXiv:2509.04438v2 Announce Type: replace Abstract: Employing a single, unified model (UM) for both visual understanding (image-to-text: I2T) and visual generation (text-to-image: T2I) has opened a new direction in Visual Language Model (VLM) research. While UMs can also support broader unimodal tasks (e.g., text-to-text, image-to-image), we focus on the core cross-modal pair T2I and I2T. Existing evaluation benchmarks consider these capabilities in isolation: FID and GenEval for T2I, and benchmarks such as MME, MMBench for I2T. These isolated single-pass metrics do not reveal cross-consistency: whether a model that "understands" a concept can also "render" it, nor whether semantic meaning is preserved when cycling between image and text modalities. To address this, we introduce the Semantic Drift Protocol (SDP) for Unified Models, a cyclic evaluation protocol that alternates I2T and T2I over multiple generations to quantify semantic drift. We propose two metrics: (i) Mean Cumulative Drift (MCD), an embedding-based measure of overall semantic drift; and (ii) Multi-Generation GenEval (MGG), an object-level compliance score extending GenEval. To assess generalization beyond COCO dataset, which is widely used in training; we create a new benchmark Nocaps+Docci400, sampled from NoCaps and DOCCI and evaluated on seven recent models. SDP reveals substantial variation in cross-modal stability: some models like BAGEL maintain semantic meaning over many alternations, whereas others like VILA-U drift quickly despite strong single-pass scores. Our results highlight SDP as a necessary complement to standard I2T and T2I evaluations. Code is available at https://github.com/mollahsabbir/Semantic-Drift-in-Unified-Models

Bias in Gender Bias Benchmarks: How Spurious Features Distort Evaluation

arXiv:2509.07596v2 Announce Type: replace Abstract: Gender bias in vision-language foundation models (VLMs) raises concerns about their safe deployment and is typically evaluated using benchmarks with gender annotations on real-world images. However, as these benchmarks often contain spurious correlations between gender and non-gender features, such as objects and backgrounds, we identify a critical oversight in gender bias evaluation: Do spurious features distort gender bias evaluation? To address this question, we systematically perturb non-gender features across four widely used benchmarks (COCO-gender, FACET, MIAP, and PHASE) and various VLMs to quantify their impact on bias evaluation. Our findings reveal that even minimal perturbations, such as masking just 10% of objects or weakly blurring backgrounds, can dramatically alter bias scores, shifting metrics by up to 175% in generative VLMs and 43% in CLIP variants. This suggests that current bias evaluations often reflect model responses to spurious features rather than gender bias, undermining their reliability. Since creating spurious feature-free benchmarks is fundamentally challenging, we recommend reporting bias metrics alongside feature-sensitivity measurements to enable a more reliable bias assessment.

Graph Algorithm Unrolling with Douglas-Rachford Iterations for Image Interpolation with Guaranteed Initialization

arXiv:2509.11926v2 Announce Type: replace Abstract: Conventional deep neural nets (DNNs) initialize network parameters at random and then optimize each one via stochastic gradient descent (SGD), resulting in substantial risk of poor-performing local minima.Focusing on the image interpolation problem and leveraging a recent theorem that maps a (pseudo-)linear interpolator {\Theta} to a directed graph filter that is a solution to a MAP problem regularized with a graph shift variation (GSV) prior, we first initialize a directed graph adjacency matrix A based on a known interpolator {\Theta}, establishing a baseline performance.Then, towards further gain, we learn perturbation matrices P and P(2) from data to augment A, whose restoration effects are implemented via Douglas-Rachford (DR) iterations, which we unroll into a lightweight interpretable neural net.Experimental results demonstrate state-of-the-art image interpolation results, while drastically reducing network parameters.

Do Vision-Language Models See Urban Scenes as People Do? An Urban Perception Benchmark

arXiv:2509.14574v2 Announce Type: replace Abstract: Understanding how people read city scenes can inform design and planning. We introduce a small benchmark for testing vision-language models (VLMs) on urban perception using 100 Montreal street images, evenly split between photographs and photorealistic synthetic scenes. Twelve participants from seven community groups supplied 230 annotation forms across 30 dimensions mixing physical attributes and subjective impressions. French responses were normalized to English. We evaluated seven VLMs in a zero-shot setup with a structured prompt and deterministic parser. We use accuracy for single-choice items and Jaccard overlap for multi-label items; human agreement uses Krippendorff's alpha and pairwise Jaccard. Results suggest stronger model alignment on visible, objective properties than subjective appraisals. The top system (claude-sonnet) reaches macro 0.31 and mean Jaccard 0.48 on multi-label items. Higher human agreement coincides with better model scores. Synthetic images slightly lower scores. We release the benchmark, prompts, and harness for reproducible, uncertainty-aware evaluation in participatory urban analysis.

Smaller is Better: Enhancing Transparency in Vehicle AI Systems via Pruning

arXiv:2509.20148v2 Announce Type: replace Abstract: Connected and autonomous vehicles continue to heavily rely on AI systems, where transparency and security are critical for trust and operational safety. Post-hoc explanations provide transparency to these black-box like AI models but the quality and reliability of these explanations is often questioned due to inconsistencies and lack of faithfulness in representing model decisions. This paper systematically examines the impact of three widely used training approaches, namely natural training, adversarial training, and pruning, affect the quality of post-hoc explanations for traffic sign classifiers. Through extensive empirical evaluation, we demonstrate that pruning significantly enhances the comprehensibility and faithfulness of explanations (using saliency maps). Our findings reveal that pruning not only improves model efficiency but also enforces sparsity in learned representation, leading to more interpretable and reliable decisions. Additionally, these insights suggest that pruning is a promising strategy for developing transparent deep learning models, especially in resource-constrained vehicular AI systems.

A Comparative Benchmark of Real-time Detectors for Blueberry Detection towards Precision Orchard Management

arXiv:2509.20580v2 Announce Type: replace Abstract: Blueberry detection in natural environments remains challenging due to variable lighting, occlusions, and motion blur due to environmental factors and imaging devices. Deep learning-based object detectors promise to address these challenges, but they demand a large-scale, diverse dataset that captures the real-world complexities. Moreover, deploying these models in practical scenarios often requires the right accuracy/speed/memory trade-off in model selection. This study presents a novel comparative benchmark analysis of advanced real-time object detectors, including YOLO (You Only Look Once) (v8-v12) and RT-DETR (Real-Time Detection Transformers) (v1-v2) families, consisting of 36 model variants, evaluated on a newly curated dataset for blueberry detection. This dataset comprises 661 canopy images collected with smartphones during the 2022-2023 seasons, consisting of 85,879 labelled instances (including 36,256 ripe and 49,623 unripe blueberries) across a wide range of lighting conditions, occlusions, and fruit maturity stages. Among the YOLO models, YOLOv12m achieved the best accuracy with a mAP@50 of 93.3%, while RT-DETRv2-X obtained a mAP@50 of 93.6%, the highest among all the RT-DETR variants. The inference time varied with the model scale and complexity, and the mid-sized models appeared to offer a good accuracy-speed balance. To further enhance detection performance, all the models were fine-tuned using Unbiased Mean Teacher-based semi-supervised learning (SSL) on a separate set of 1,035 unlabeled images acquired by a ground-based machine vision platform in 2024. This resulted in accuracy gains ranging from -1.4% to 2.9%, with RT-DETR-v2-X achieving the best mAP@50 of 94.8%. More in-depth research into SSL is needed to better leverage cross-domain unlabeled data. Both the dataset and software programs of this study are made publicly available to support further research.

Region-of-Interest Augmentation for Mammography Classification under Patient-Level Cross-Validation

arXiv:2509.20585v2 Announce Type: replace Abstract: Breast cancer screening with mammography remains central to early detection and mortality reduction. Deep learning has shown strong potential for automating mammogram interpretation, yet limited-resolution datasets and small sample sizes continue to restrict performance. We revisit the Mini-DDSM dataset (9,684 images; 2,414 patients) and introduce a lightweight region-of-interest (ROI) augmentation strategy. During training, full images are probabilistically replaced with random ROI crops sampled from a precomputed, label-free bounding-box bank, with optional jitter to increase variability. We evaluate under strict patient-level cross-validation and report ROC-AUC, PR-AUC, and training-time efficiency metrics (throughput and GPU memory). Because ROI augmentation is training-only, inference-time cost remains unchanged. On Mini-DDSM, ROI augmentation (best: p_roi = 0.10, alpha = 0.10) yields modest average ROC-AUC gains, with performance varying across folds; PR-AUC is flat to slightly lower. These results demonstrate that simple, data-centric ROI strategies can enhance mammography classification in constrained settings without requiring additional labels or architectural modifications.

Do Sparse Subnetworks Exhibit Cognitively Aligned Attention? Effects of Pruning on Saliency Map Fidelity, Sparsity, and Concept Coherence

arXiv:2509.21387v3 Announce Type: replace Abstract: Prior works have shown that neural networks can be heavily pruned while preserving performance, but the impact of pruning on model interpretability remains unclear. In this work, we investigate how magnitude-based pruning followed by fine-tuning affects both low-level saliency maps and high-level concept representations. Using a ResNet-18 trained on ImageNette, we compare post-hoc explanations from Vanilla Gradients (VG) and Integrated Gradients (IG) across pruning levels, evaluating sparsity and faithfulness. We further apply CRAFT-based concept extraction to track changes in semantic coherence of learned concepts. Our results show that light-to-moderate pruning improves saliency-map focus and faithfulness while retaining distinct, semantically meaningful concepts. In contrast, aggressive pruning merges heterogeneous features, reducing saliency map sparsity and concept coherence despite maintaining accuracy. These findings suggest that while pruning can shape internal representations toward more human-aligned attention patterns, excessive pruning undermines interpretability.

Reasoning-Enhanced Domain-Adaptive Pretraining of Multimodal Large Language Models for Short Video Content Governance

arXiv:2509.21486v2 Announce Type: replace Abstract: Short video platforms are evolving rapidly, making the identification of inappropriate content increasingly critical. Existing approaches typically train separate and small classification models for each type of issue, which requires extensive human-labeled data and lacks cross-issue generalization. We propose a reasoning-enhanced multimodal large language model (MLLM) pretraining paradigm for unified inappropriate content detection. To address the distribution gap between short video content and the original pretraining data of MLLMs, as well as the complex issue definitions, we introduce three targeted pretraining tasks: (1) \textit{Caption}, to enhance the MLLM's perception of video details; (2) \textit{Visual Question Answering (VQA)}, to deepen the MLLM's understanding of issue definitions and annotation guidelines; (3) \textit{Chain-of-Thought (CoT)}, to enhance the MLLM's reasoning capability. Experimental results show that our pretraining approach significantly improves the MLLM's performance in both zero-shot and supervised fine-tuning (SFT) settings. In addition, our pretrained model demonstrates strong generalization capabilities to emergent, previously unseen issues.

A Tale of Two Experts: Cooperative Learning for Source-Free Unsupervised Domain Adaptation

arXiv:2509.22229v2 Announce Type: replace Abstract: Source-Free Unsupervised Domain Adaptation (SFUDA) addresses the realistic challenge of adapting a source-trained model to a target domain without access to the source data, driven by concerns over privacy and cost. Existing SFUDA methods either exploit only the source model's predictions or fine-tune large multimodal models, yet both neglect complementary insights and the latent structure of target data. In this paper, we propose the Experts Cooperative Learning (EXCL). EXCL contains the Dual Experts framework and Retrieval-Augmentation-Interaction optimization pipeline. The Dual Experts framework places a frozen source-domain model (augmented with Conv-Adapter) and a pretrained vision-language model (with a trainable text prompt) on equal footing to mine consensus knowledge from unlabeled target samples. To effectively train these plug-in modules under purely unsupervised conditions, we introduce Retrieval-Augmented-Interaction(RAIN), a three-stage pipeline that (1) collaboratively retrieves pseudo-source and complex target samples, (2) separately fine-tunes each expert on its respective sample set, and (3) enforces learning object consistency via a shared learning result. Extensive experiments on four benchmark datasets demonstrate that our approach matches state-of-the-art performance.

Copyright Infringement Detection in Text-to-Image Diffusion Models via Differential Privacy

arXiv:2509.23022v2 Announce Type: replace Abstract: The widespread deployment of large vision models such as Stable Diffusion raises significant legal and ethical concerns, as these models can memorize and reproduce copyrighted content without authorization. Existing detection approaches often lack robustness and fail to provide rigorous theoretical underpinnings. To address these gaps, we formalize the concept of copyright infringement and its detection from the perspective of Differential Privacy (DP), and introduce the conditional sensitivity metric, a concept analogous to sensitivity in DP, that quantifies the deviation in a diffusion model's output caused by the inclusion or exclusion of a specific training data point. To operationalize this metric, we propose D-Plus-Minus (DPM), a novel post-hoc detection framework that identifies copyright infringement in text-to-image diffusion models. Specifically, DPM simulates inclusion and exclusion processes by fine-tuning models in two opposing directions: learning or unlearning. Besides, to disentangle concept-specific influence from the global parameter shifts induced by fine-tuning, DPM computes confidence scores over orthogonal prompt distributions using statistical metrics. Moreover, to facilitate standardized benchmarking, we also construct the Copyright Infringement Detection Dataset (CIDD), a comprehensive resource for evaluating detection across diverse categories. Our results demonstrate that DPM reliably detects infringement content without requiring access to the original training dataset or text prompts, offering an interpretable and practical solution for safeguarding intellectual property in the era of generative AI.

Mask What Matters: Controllable Text-Guided Masking for Self-Supervised Medical Image Analysis

arXiv:2509.23054v2 Announce Type: replace Abstract: The scarcity of annotated data in specialized domains such as medical imaging presents significant challenges to training robust vision models. While self-supervised masked image modeling (MIM) offers a promising solution, existing approaches largely rely on random high-ratio masking, leading to inefficiency and poor semantic alignment. Moreover, region-aware variants typically depend on reconstruction heuristics or supervised signals, limiting their adaptability across tasks and modalities. We propose Mask What Matters, a controllable text-guided masking framework for self-supervised medical image analysis. By leveraging vision-language models for prompt-based region localization, our method flexibly applies differentiated masking to emphasize diagnostically relevant regions while reducing redundancy in background areas. This controllable design enables better semantic alignment, improved representation learning, and stronger cross-task generalizability. Comprehensive evaluation across multiple medical imaging modalities, including brain MRI, chest CT, and lung X-ray, shows that Mask What Matters consistently outperforms existing MIM methods (e.g., SparK), achieving gains of up to +3.1 percentage points in classification accuracy, +1.3 in box average precision (BoxAP), and +1.1 in mask average precision (MaskAP) for detection. Notably, it achieves these improvements with substantially lower overall masking ratios (e.g., 40\% vs. 70\%). This work demonstrates that controllable, text-driven masking can enable semantically aligned self-supervised learning, advancing the development of robust vision models for medical image analysis.

OracleGS: Grounding Generative Priors for Sparse-View Gaussian Splatting

arXiv:2509.23258v2 Announce Type: replace Abstract: Sparse-view novel view synthesis is fundamentally ill-posed due to severe geometric ambiguity. Current methods are caught in a trade-off: regressive models are geometrically faithful but incomplete, whereas generative models can complete scenes but often introduce structural inconsistencies. We propose OracleGS, a novel framework that reconciles generative completeness with regressive fidelity for sparse view Gaussian Splatting. Instead of using generative models to patch incomplete reconstructions, our "propose-and-validate" framework first leverages a pre-trained 3D-aware diffusion model to synthesize novel views to propose a complete scene. We then repurpose a multi-view stereo (MVS) model as a 3D-aware oracle to validate the 3D uncertainties of generated views, using its attention maps to reveal regions where the generated views are well-supported by multi-view evidence versus where they fall into regions of high uncertainty due to occlusion, lack of texture, or direct inconsistency. This uncertainty signal directly guides the optimization of a 3D Gaussian Splatting model via an uncertainty-weighted loss. Our approach conditions the powerful generative prior on multi-view geometric evidence, filtering hallucinatory artifacts while preserving plausible completions in under-constrained regions, outperforming state-of-the-art methods on datasets including Mip-NeRF 360 and NeRF Synthetic.

EfficientMIL: Efficient Linear-Complexity MIL Method for WSI Classification

arXiv:2509.23640v2 Announce Type: replace Abstract: Whole slide images (WSIs) classification represents a fundamental challenge in computational pathology, where multiple instance learning (MIL) has emerged as the dominant paradigm. Current state-of-the-art (SOTA) MIL methods rely on attention mechanisms, achieving good performance but requiring substantial computational resources due to quadratic complexity when processing hundreds of thousands of patches. To address this computational bottleneck, we introduce EfficientMIL, a novel linear-complexity MIL approach for WSIs classification with the patches selection module Adaptive Patch Selector (APS) that we designed, replacing the quadratic-complexity self-attention mechanisms in Transformer-based MIL methods with efficient sequence models including RNN-based GRU, LSTM, and State Space Model (SSM) Mamba. EfficientMIL achieves significant computational efficiency improvements while outperforming other MIL methods across multiple histopathology datasets. On TCGA-Lung dataset, EfficientMIL-Mamba achieved AUC of 0.976 and accuracy of 0.933, while on CAMELYON16 dataset, EfficientMIL-GRU achieved AUC of 0.990 and accuracy of 0.975, surpassing previous state-of-the-art methods. Extensive experiments demonstrate that APS is also more effective for patches selection than conventional selection strategies.

AutoPrune: Each Complexity Deserves a Pruning Policy

arXiv:2509.23931v2 Announce Type: replace Abstract: The established redundancy in visual tokens within large vision-language models allows pruning to effectively reduce their substantial computational demands. Previous methods typically employ heuristic layer-specific pruning strategies where, although the number of tokens removed may differ across decoder layers, the overall pruning schedule is fixed and applied uniformly to all input samples and tasks, failing to align token elimination with the model's holistic reasoning trajectory. Cognitive science indicates that human visual processing often begins with broad exploration to accumulate evidence before narrowing focus as the target becomes distinct. Our experiments reveal an analogous pattern in these models. This observation suggests that neither a fixed pruning schedule nor a heuristic layer-wise strategy can optimally accommodate the diverse complexities inherent in different inputs. To overcome this limitation, we introduce Complexity-Adaptive Pruning (AutoPrune), a training-free, plug-and-play framework that tailors pruning policies to varying sample and task complexities. Specifically, AutoPrune quantifies the mutual information between visual and textual tokens, then projects this signal to a budget-constrained logistic retention curve. Each such logistic curve, defined by its unique shape, corresponds to the specific complexity of different tasks and can guarantee adherence to predefined computational constraints. We evaluate AutoPrune on standard vision-language tasks and on Vision-Language-Action models for autonomous driving. Notably, when applied to LLaVA-1.5-7B, our method prunes 89% of visual tokens and reduces inference FLOPs by 76.8% while retaining 96.7% of the original accuracy averaged over all tasks. This corresponds to a 9.1% improvement over the recent work PDrop, demonstrating the effectiveness. Code is available at https://github.com/AutoLab-SAI-SJTU/AutoPrune.

FrameMind: Frame-Interleaved Video Reasoning via Reinforcement Learning

arXiv:2509.24008v3 Announce Type: replace Abstract: Current video understanding models rely on fixed frame sampling strategies, processing predetermined visual inputs regardless of the specific reasoning requirements of each question. This static approach limits their ability to adaptively gather visual evidence, leading to suboptimal performance on tasks that require either broad temporal coverage or fine-grained spatial detail. In this paper, we introduce FrameMind, an end-to-end framework trained with reinforcement learning that enables models to dynamically request visual information during reasoning through Frame-Interleaved Chain-of-Thought (FiCOT). Unlike traditional approaches, FrameMind operates in multiple turns where the model alternates between textual reasoning and active visual perception, using tools to extract targeted frames or video clips based on identified knowledge gaps. To train effective dynamic sampling policies, we propose Dynamic Resolution Frame Sampling (DRFS), which exposes models to diverse temporal-spatial trade-offs during learning, and DRFS-GRPO, a group-relative policy optimization algorithm that learns from outcome-based rewards without requiring frame-level annotations. Extensive experiments on challenging benchmarks like MLVU and VideoMME demonstrate that our method significantly outperforms existing models, advancing the state of the art in flexible and efficient video understanding.

Uncovering Grounding IDs: How External Cues Shape Multi-Modal Binding

arXiv:2509.24072v2 Announce Type: replace Abstract: Large vision-language models (LVLMs) show strong performance across multimodal benchmarks but remain limited in structured reasoning and precise grounding. Recent work has demonstrated that adding simple visual structures, such as partitions and annotations, improves accuracy, yet the internal mechanisms underlying these gains remain unclear. We investigate this phenomenon and propose the concept of Grounding IDs, latent identifiers induced by external cues that bind objects to their designated partitions across modalities. Through representation analysis, we find that these identifiers emerge as robust within-partition alignment in embedding space and reduce the modality gap between image and text. Causal interventions further confirm that these identifiers mediate binding between objects and symbolic cues. We show that Grounding IDs strengthen attention between related components, which in turn improves cross-modal grounding and reduces hallucinations. Taken together, our results identify Grounding IDs as a key symbolic mechanism explaining how external cues enhance multimodal binding, offering both interpretability and practical improvements in robustness.

Latent Visual Reasoning

arXiv:2509.24251v2 Announce Type: replace Abstract: Multimodal Large Language Models (MLLMs) have achieved notable gains in various tasks by incorporating Chain-of-Thought (CoT) reasoning in language spaces. Recent work extends this direction by leveraging external tools for visual editing, thereby enhancing the visual signal along the reasoning trajectories. Nevertheless, these approaches remain fundamentally constrained: reasoning is still confined to the language space, with visual information treated as static preconditions. We introduce Latent Visual Reasoning (LVR), a new paradigm that enables autoregressive reasoning directly in the visual embedding space. A visual encoder first projects images into visual tokens within a joint semantic space shared with the language model. The language model is then trained to generate latent states that reconstruct key visual tokens critical for answering the query, constituting the process of latent visual reasoning. By interleaving LVR with standard text generation, our model achieves substantial gains on perception-intensive visual question answering tasks. In addition, we adapt the GRPO algorithm to conduct reinforcement learning on latent reasoning, further balancing LVR and textual generation. We show that LVR substantially improves fine-grained visual understanding and perception, achieving 71.67% on MMVP compared to 66.67% with Qwen2.5-VL. Code base and model weights will be released later.

Towards Foundation Models for Cryo-ET Subtomogram Analysis

arXiv:2509.24311v2 Announce Type: replace Abstract: Cryo-electron tomography (cryo-ET) enables in situ visualization of macromolecular structures, where subtomogram analysis tasks such as classification, alignment, and averaging are critical for structural determination. However, effective analysis is hindered by scarce annotations, severe noise, and poor generalization. To address these challenges, we take the first step towards foundation models for cryo-ET subtomograms. First, we introduce CryoEngine, a large-scale synthetic data generator that produces over 904k subtomograms from 452 particle classes for pretraining. Second, we design an Adaptive Phase Tokenization-enhanced Vision Transformer (APT-ViT), which incorporates adaptive phase tokenization as an equivariance-enhancing module that improves robustness to both geometric and semantic variations. Third, we introduce a Noise-Resilient Contrastive Learning (NRCL) strategy to stabilize representation learning under severe noise conditions. Evaluations across 24 synthetic and real datasets demonstrate state-of-the-art (SOTA) performance on all three major subtomogram tasks and strong generalization to unseen datasets, advancing scalable and robust subtomogram analysis in cryo-ET.

ExGS: Extreme 3D Gaussian Compression with Diffusion Priors

arXiv:2509.24758v3 Announce Type: replace Abstract: Neural scene representations, such as 3D Gaussian Splatting (3DGS), have enabled high-quality neural rendering; however, their large storage and transmission costs hinder deployment in resource-constrained environments. Existing compression methods either rely on costly optimization, which is slow and scene-specific, or adopt training-free pruning and quantization, which degrade rendering quality under high compression ratios. In contrast, recent data-driven approaches provide a promising direction to overcome this trade-off, enabling efficient compression while preserving high rendering quality.We introduce ExGS, a novel feed-forward framework that unifies Universal Gaussian Compression (UGC) with GaussPainter for Extreme 3DGS compression. UGC performs re-optimization-free pruning to aggressively reduce Gaussian primitives while retaining only essential information, whereas GaussPainter leverages powerful diffusion priors with mask-guided refinement to restore high-quality renderings from heavily pruned Gaussian scenes. Unlike conventional inpainting, GaussPainter not only fills in missing regions but also enhances visible pixels, yielding substantial improvements in degraded renderings.To ensure practicality, it adopts a lightweight VAE and a one-step diffusion design, enabling real-time restoration. Our framework can even achieve over 100X compression (reducing a typical 354.77 MB model to about 3.31 MB) while preserving fidelity and significantly improving image quality under challenging conditions. These results highlight the central role of diffusion priors in bridging the gap between extreme compression and high-quality neural rendering.Our code repository will be released at: https://github.com/chenttt2001/ExGS

Self-Evolving Vision-Language Models for Image Quality Assessment via Voting and Ranking

arXiv:2509.25787v2 Announce Type: replace Abstract: Improving vision-language models (VLMs) in the post-training stage typically relies on supervised fine-tuning or reinforcement learning, methods that necessitate costly, human-annotated data. While self-supervised techniques such as self-consistency have proven effective for enhancing reasoning capabilities, their application to perceptual domains such as image quality assessment (IQA) remains largely unexplored. In this work, we introduce EvoQuality, a novel framework that enables a VLM to autonomously refine its quality perception capabilities without any ground-truth labels. EvoQuality adapts the principle of self-consistency to the ranking-based nature of IQA. It generates pseudo-labels by performing pairwise majority voting on the VLM's own outputs to establish a consensus on relative quality. These pseudo-rankings are then formulated into a fidelity reward that guides the model's iterative evolution through group relative policy optimization (GRPO). By iteratively leveraging its own predictions, EvoQuality progressively refines the VLM's perceptual capability. Extensive experiments show that EvoQuality boosts the base VLM's zero-shot performance by 31.8\% on PLCC across diverse IQA benchmarks. Remarkably, despite being entirely self-supervised, EvoQuality achieves performance that is competitive with, or even surpasses, state-of-the-art supervised VLM-based IQA models, outperforming these models on 5 out of 7 IQA benchmarks.

PAL-UI: Planning with Active Look-back for Vision-Based GUI Agents

arXiv:2510.00413v2 Announce Type: replace Abstract: Graphical User Interface (GUI) agents powered by Multimodal Large Language Models (MLLMs) promise human-like interaction with software applications, yet long-horizon tasks remain challenging due to memory limitations. Existing approaches either truncate history or rely on simple textual summaries, which risk losing critical information when past visual details become necessary for future decisions. In this paper, we propose \textbf{PAL-UI} (\textbf{P}lanning with \textbf{A}ctive \textbf{L}ook-back), a novel framework that enables GUI agents to adaptively retrieve past observations when required. PAL-UI combines a dual-level summarization agent, capturing both observation-level cues and action-level outcomes, with a dedicated retrieval tool that allows the agent to recall specific historical screenshots during planning. We curate a step-level instruction dataset of 8.6K samples from mobile GUI navigation trajectories and train \textbf{PAL-UI-3B} and \textbf{PAL-UI-7B} models based on Qwen2.5-VL. Extensive experiments demonstrate that PAL-UI significantly outperforms baseline models and prior methods in mobile GUI navigation tasks, even under data-efficient settings. Moreover, PAL-UI exhibits strong cross-domain generalization, achieving notable improvements in web navigation without additional training. Our work highlights the potential of active memory retrieval for long-horizon planning capabilities of vision-based GUI agents.

Erased, But Not Forgotten: Erased Rectified Flow Transformers Still Remain Unsafe Under Concept Attack

arXiv:2510.00635v2 Announce Type: replace Abstract: Recent advances in text-to-image (T2I) diffusion models have enabled impressive generative capabilities, but they also raise significant safety concerns due to the potential to produce harmful or undesirable content. While concept erasure has been explored as a mitigation strategy, most existing approaches and corresponding attack evaluations are tailored to Stable Diffusion (SD) and exhibit limited effectiveness when transferred to next-generation rectified flow transformers such as Flux. In this work, we present ReFlux, the first concept attack method specifically designed to assess the robustness of concept erasure in the latest rectified flow-based T2I framework. Our approach is motivated by the observation that existing concept erasure techniques, when applied to Flux, fundamentally rely on a phenomenon known as attention localization. Building on this insight, we propose a simple yet effective attack strategy that specifically targets this property. At its core, a reverse-attention optimization strategy is introduced to effectively reactivate suppressed signals while stabilizing attention. This is further reinforced by a velocity-guided dynamic that enhances the robustness of concept reactivation by steering the flow matching process, and a consistency-preserving objective that maintains the global layout and preserves unrelated content. Extensive experiments consistently demonstrate the effectiveness and efficiency of the proposed attack method, establishing a reliable benchmark for evaluating the robustness of concept erasure strategies in rectified flow transformers.

VirDA: Reusing Backbone for Unsupervised Domain Adaptation with Visual Reprogramming

arXiv:2510.01660v3 Announce Type: replace Abstract: Existing UDA pipelines fine-tune already well-trained backbone parameters for every new source-and-target pair, resulting in the number of training parameters and storage memory growing linearly with each new pair, and also preventing the reuse of these well-trained backbone parameters. Inspired by recent implications that existing backbones have textural biases, we propose making use of domain-specific textural bias for domain adaptation via visual reprogramming, namely VirDA. Instead of fine-tuning the full backbone, VirDA prepends a domain-specific visual reprogramming layer to the backbone. This layer produces visual prompts that act as an added textural bias to the input image, adapting its "style" to a target domain. To optimize these visual reprogramming layers, we use multiple objective functions that optimize the intra- and inter-domain distribution differences when domain-adapting visual prompts are applied. This process does not require modifying the backbone parameters, allowing the same backbone to be reused across different domains. We evaluate VirDA on Office-31 and obtain 92.8% mean accuracy with only 1.5M trainable parameters. VirDA surpasses PDA, the state-of-the-art parameter-efficient UDA baseline, by +1.6% accuracy while using just 46% of its parameters. Compared with full-backbone fine-tuning, VirDA outperforms CDTrans and FixBi by +0.2% and +1.4%, respectively, while requiring only 1.7% and 2.8% of their trainable parameters. Relative to the strongest current methods (PMTrans and TVT), VirDA uses ~1.7% of their parameters and trades off only 2.2% and 1.1% accuracy, respectively.

Automated Defect Detection for Mass-Produced Electronic Components Based on YOLO Object Detection Models

arXiv:2510.01914v2 Announce Type: replace Abstract: Since the defect detection of conventional industry components is time-consuming and labor-intensive, it leads to a significant burden on quality inspection personnel and makes it difficult to manage product quality. In this paper, we propose an automated defect detection system for the dual in-line package (DIP) that is widely used in industry, using digital camera optics and a deep learning (DL)-based model. The two most common defect categories of DIP are examined: (1) surface defects, and (2) pin-leg defects. However, the lack of defective component images leads to a challenge for detection tasks. To solve this problem, the ConSinGAN is used to generate a suitable-sized dataset for training and testing. Four varieties of the YOLO model are investigated (v3, v4, v7, and v9), both in isolation and with the ConSinGAN augmentation. The proposed YOLOv7 with ConSinGAN is superior to the other YOLO versions in accuracy of 95.50\%, detection time of 285 ms, and is far superior to threshold-based approaches. In addition, the supervisory control and data acquisition (SCADA) system is developed, and the associated sensor architecture is described. The proposed automated defect detection can be easily established with numerous types of defects or insufficient defect data.

Generating Findings for Jaw Cysts in Dental Panoramic Radiographs Using GPT-4o: Building a Two-Stage Self-Correction Loop with Structured Output (SLSO) Framework

arXiv:2510.02001v2 Announce Type: replace Abstract: In this study, we utilized the multimodal capabilities of OpenAI GPT-4o to automatically generate jaw cyst findings on dental panoramic radiographs. To improve accuracy, we constructed a Self-correction Loop with Structured Output (SLSO) framework and verified its effectiveness. A 10-step process was implemented for 22 cases of jaw cysts, including image input and analysis, structured data generation, tooth number extraction and consistency checking, iterative regeneration when inconsistencies were detected, and finding generation with subsequent restructuring and consistency verification. A comparative experiment was conducted using the conventional Chain-of-Thought (CoT) method across seven evaluation items: transparency, internal structure, borders, root resorption, tooth movement, relationships with other structures, and tooth number. The results showed that the proposed SLSO framework improved output accuracy for many items, with 66.9%, 33.3%, and 28.6% improvement rates for tooth number, tooth movement, and root resorption, respectively. In the successful cases, a consistently structured output was achieved after up to five regenerations. Although statistical significance was not reached because of the small size of the dataset, the overall SLSO framework enforced negative finding descriptions, suppressed hallucinations, and improved tooth number identification accuracy. However, the accurate identification of extensive lesions spanning multiple teeth is limited. Nevertheless, further refinement is required to enhance overall performance and move toward a practical finding generation system.

VidGuard-R1: AI-Generated Video Detection and Explanation via Reasoning MLLMs and RL

arXiv:2510.02282v2 Announce Type: replace Abstract: With the rapid advancement of AI-generated videos, there is an urgent need for effective detection tools to mitigate societal risks such as misinformation and reputational harm. In addition to accurate classification, it is essential that detection models provide interpretable explanations to ensure transparency for regulators and end users. To address these challenges, we introduce VidGuard-R1, the first video authenticity detector that fine-tunes a multi-modal large language model (MLLM) using group relative policy optimization (GRPO). Our model delivers both highly accurate judgments and insightful reasoning. We curate a challenging dataset of 140k real and AI-generated videos produced by state-of-the-art generation models, carefully designing the generation process to maximize discrimination difficulty. We then fine-tune Qwen-VL using GRPO with two specialized reward models that target temporal artifacts and generation complexity. Extensive experiments demonstrate that VidGuard-R1 achieves state-of-the-art zero-shot performance on existing benchmarks, with additional training pushing accuracy above 95%. Case studies further show that VidGuard-R1 produces precise and interpretable rationales behind its predictions. The code is publicly available at https://VidGuard-R1.github.io.

One Patch to Caption Them All: A Unified Zero-Shot Captioning Framework

arXiv:2510.02898v2 Announce Type: replace Abstract: Zero-shot captioners are recently proposed models that utilize common-space vision-language representations to caption images without relying on paired image-text data. To caption an image, they proceed by textually decoding a text-aligned image feature, but they limit their scope to global representations and whole-image captions. We present Patch-ioner, a unified framework for zero-shot captioning that shifts from an image-centric to a patch-centric paradigm, enabling the captioning of arbitrary regions without the need of region-level supervision. Instead of relying on global image representations, we treat individual patches as atomic captioning units and aggregate them to describe arbitrary regions, from single patches to non-contiguous areas and entire images. We analyze the key ingredients that enable current latent captioners to work in our novel proposed framework. Experiments demonstrate that backbones producing meaningful, dense visual features, such as DINO, are key to achieving state-of-the-art performance in multiple region-based captioning tasks. Compared to other baselines and state-of-the-art competitors, our models achieve better performance on zero-shot dense, region-set, and a newly introduced trace captioning task, highlighting the effectiveness of patch-wise semantic representations for scalable caption generation. Project page at https://paciosoft.com/Patch-ioner/ .

What Drives Compositional Generalization in Visual Generative Models?

arXiv:2510.03075v2 Announce Type: replace Abstract: Compositional generalization, the ability to generate novel combinations of known concepts, is a key ingredient for visual generative models. Yet, not all mechanisms that enable or inhibit it are fully understood. In this work, we conduct a systematic study of how various design choices influence compositional generalization in image and video generation in a positive or negative way. Through controlled experiments, we identify two key factors: (i) whether the training objective operates on a discrete or continuous distribution, and (ii) to what extent conditioning provides information about the constituent concepts during training. Building on these insights, we show that relaxing the MaskGIT discrete loss with an auxiliary continuous JEPA-based objective can improve compositional performance in discrete models like MaskGIT.

Robust MRI Reconstruction by Smoothed Unrolling (SMUG)

arXiv:2312.07784v3 Announce Type: replace-cross Abstract: As the popularity of deep learning (DL) in the field of magnetic resonance imaging (MRI) continues to rise, recent research has indicated that DL-based MRI reconstruction models might be excessively sensitive to minor input disturbances, including worst-case additive perturbations. This sensitivity often leads to unstable, aliased images. This raises the question of how to devise DL techniques for MRI reconstruction that can be robust to train-test variations. To address this problem, we propose a novel image reconstruction framework, termed Smoothed Unrolling (SMUG), which advances a deep unrolling-based MRI reconstruction model using a randomized smoothing (RS)-based robust learning approach. RS, which improves the tolerance of a model against input noises, has been widely used in the design of adversarial defense approaches for image classification tasks. Yet, we find that the conventional design that applies RS to the entire DL-based MRI model is ineffective. In this paper, we show that SMUG and its variants address the above issue by customizing the RS process based on the unrolling architecture of a DL-based MRI reconstruction model. Compared to the vanilla RS approach, we show that SMUG improves the robustness of MRI reconstruction with respect to a diverse set of instability sources, including worst-case and random noise perturbations to input measurements, varying measurement sampling rates, and different numbers of unrolling steps. Furthermore, we theoretically analyze the robustness of our method in the presence of perturbations.

RealKIE: Five Novel Datasets for Enterprise Key Information Extraction

arXiv:2403.20101v2 Announce Type: replace-cross Abstract: We introduce RealKIE, a benchmark of five challenging datasets aimed at advancing key information extraction methods, with an emphasis on enterprise applications. The datasets include a diverse range of documents including SEC S1 Filings, US Non-disclosure Agreements, UK Charity Reports, FCC Invoices, and Resource Contracts. Each presents unique challenges: poor text serialization, sparse annotations in long documents, and complex tabular layouts. These datasets provide a realistic testing ground for key information extraction tasks like investment analysis and contract analysis. In addition to presenting these datasets, we offer an in-depth description of the annotation process, document processing techniques, and baseline modeling approaches. This contribution facilitates the development of NLP models capable of handling practical challenges and supports further research into information extraction technologies applicable to industry-specific problems. The annotated data, OCR outputs, and code to reproduce baselines are available to download at https://indicodatasolutions.github.io/RealKIE/.

Temporal Source Recovery for Time-Series Source-Free Unsupervised Domain Adaptation

arXiv:2409.19635v2 Announce Type: replace-cross Abstract: Time-Series (TS) data has grown in importance with the rise of Internet of Things devices like sensors, but its labeling remains costly and complex. While Unsupervised Domain Adaptation (UDAs) offers an effective solution, growing data privacy concerns have led to the development of Source-Free UDA (SFUDAs), enabling model adaptation to target domains without accessing source data. Despite their potential, applying existing SFUDAs to TS data is challenging due to the difficulty of transferring temporal dependencies, an essential characteristic of TS data, particularly in the absence of source samples. Although prior works attempt to address this by specific source pretraining designs, such requirements are often impractical, as source data owners cannot be expected to adhere to particular pretraining schemes. To address this, we propose Temporal Source Recovery (TemSR), a framework that leverages the intrinsic properties of TS data to generate a source-like domain and recover source temporal dependencies. With this domain, TemSR enables dependency transfer to the target domain without accessing source data or relying on source-specific designs, thereby facilitating effective and practical TS-SFUDA. TemSR features a masking recovery optimization process to generate a source-like distribution with restored temporal dependencies. This distribution is further refined through local context-aware regularization to preserve local dependencies, and anchor-based recovery diversity maximization to promote distributional diversity. Together, these components enable effective temporal dependency recovery and facilitate transfer across domains using standard UDA techniques. Extensive experiments across multiple TS tasks demonstrate the effectiveness of TemSR, which even surpasses existing TS-SFUDA methods that require source-specific designs.

Segmenting Bi-Atrial Structures Using ResNext Based Framework

arXiv:2503.02892v3 Announce Type: replace-cross Abstract: Atrial Fibrillation (AF), the most common sustained cardiac arrhythmia worldwide, increasingly requires accurate bi-atrial structural assessment to guide ablation strategies, particularly in persistent AF. Late gadolinium-enhanced magnetic resonance imaging (LGE-MRI) enables visualisation of atrial fibrosis, but precise manual segmentation remains time-consuming, operator-dependent, and prone to variability. We propose TASSNet, a novel two-stage deep learning framework for fully automated segmentation of both left atrium (LA) and right atrium (RA), including atrial walls and cavities, from 3D LGE-MRI. TASSNet introduces two main innovations: (i) a ResNeXt-based encoder to enhance feature extraction from limited medical datasets, and (ii) a cyclical learning rate schedule to address convergence instability in highly imbalanced, small-batch 3D segmentation tasks. We evaluated our method on two datasets, one of which was completely out-of-distribution, without any additional training. In both cases, TASSNet successfully segmented atrial structures with high accuracy. These results highlight TASSNet's potential for robust and reproducible bi-atrial segmentation, enabling advanced fibrosis quantification and personalised ablation planning in clinical AF management.

Humanoid Policy ~ Human Policy

arXiv:2503.13441v3 Announce Type: replace-cross Abstract: Training manipulation policies for humanoid robots with diverse data enhances their robustness and generalization across tasks and platforms. However, learning solely from robot demonstrations is labor-intensive, requiring expensive tele-operated data collection which is difficult to scale. This paper investigates a more scalable data source, egocentric human demonstrations, to serve as cross-embodiment training data for robot learning. We mitigate the embodiment gap between humanoids and humans from both the data and modeling perspectives. We collect an egocentric task-oriented dataset (PH2D) that is directly aligned with humanoid manipulation demonstrations. We then train a human-humanoid behavior policy, which we term Human Action Transformer (HAT). The state-action space of HAT is unified for both humans and humanoid robots and can be differentiably retargeted to robot actions. Co-trained with smaller-scale robot data, HAT directly models humanoid robots and humans as different embodiments without additional supervision. We show that human data improves both generalization and robustness of HAT with significantly better data collection efficiency. Code and data: https://human-as-robot.github.io/

Periodontal Bone Loss Analysis via Keypoint Detection With Heuristic Post-Processing

arXiv:2503.13477v2 Announce Type: replace-cross Abstract: This study proposes a deep learning framework and annotation methodology for the automatic detection of periodontal bone loss landmarks, associated conditions, and staging. 192 periapical radiographs were collected and annotated with a stage agnostic methodology, labelling clinically relevant landmarks regardless of disease presence or extent. We propose a heuristic post-processing module that aligns predicted keypoints to tooth boundaries using an auxiliary instance segmentation model. An evaluation metric, Percentage of Relative Correct Keypoints (PRCK), is proposed to capture keypoint performance in dental imaging domains. Four donor pose estimation models were adapted with fine-tuning for our keypoint problem. Post-processing improved fine-grained localisation, raising average PRCK^{0.05} by +0.028, but reduced coarse performance for PRCK^{0.25} by -0.0523 and PRCK^{0.5} by -0.0345. Orientation estimation shows excellent performance for auxiliary segmentation when filtered with either stage 1 object detection model. Periodontal staging was detected sufficiently, with the best mesial and distal Dice scores of 0.508 and 0.489, while furcation involvement and widened periodontal ligament space tasks remained challenging due to scarce positive samples. Scalability is implied with similar validation and external set performance. The annotation methodology enables stage agnostic training with balanced representation across disease severities for some detection tasks. The PRCK metric provides a domain-specific alternative to generic pose metrics, while the heuristic post-processing module consistently corrected implausible predictions with occasional catastrophic failures. The proposed framework demonstrates the feasibility of clinically interpretable periodontal bone loss assessment, with potential to reduce diagnostic variability and clinician workload.

Praxis-VLM: Vision-Grounded Decision Making via Text-Driven Reinforcement Learning

arXiv:2503.16965v3 Announce Type: replace-cross Abstract: Vision Language Models exhibit impressive performance for various tasks, yet they often lack the sophisticated situational reasoning required for complex decision-making. This paper shows that VLMs can achieve surprisingly strong decision-making performance when visual scenes are replaced by textual descriptions, suggesting foundational reasoning can be effectively learned from language. Motivated by this insight, we propose Praxis-VLM, a reasoning VLM for vision-grounded decision-making. Praxis-VLM employs the GRPO algorithm on textual scenarios to instill robust reasoning capabilities, where models learn to evaluate actions and their consequences. These reasoning skills, acquired purely from text, successfully transfer to multimodal inference with visual inputs, significantly reducing reliance on scarce paired image-text training data. Experiments across diverse decision-making benchmarks demonstrate that Praxis-VLM substantially outperforms standard supervised fine-tuning, exhibiting superior performance and generalizability. Further analysis confirms that our models engage in explicit and effective reasoning, underpinning their enhanced performance and adaptability.

Poisson multi-Bernoulli mixture filter for trajectory measurements

arXiv:2504.08421v2 Announce Type: replace-cross Abstract: This paper presents a Poisson multi-Bernoulli mixture (PMBM) filter for multi-target filtering based on sensor measurements that are sets of trajectories in the last two-time step window. The proposed filter, the trajectory measurement PMBM (TM-PMBM) filter, propagates a PMBM density on the set of target states. In prediction, the filter obtains the PMBM density on the set of trajectories over the last two time steps. This density is then updated with the set of trajectory measurements. After the update step, the PMBM posterior on the set of two-step trajectories is marginalised to obtain a PMBM density on the set of target states. The filter provides a closed-form solution for multi-target filtering based on sets of trajectory measurements, estimating the set of target states at the end of each time window. Additionally, the paper proposes computationally lighter alternatives to the TM-PMBM filter by deriving a Poisson multi-Bernoulli (PMB) density through Kullback-Leibler divergence minimisation in an augmented space with auxiliary variables. The performance of the proposed filters are evaluated in a simulation study.

What Lurks Within? Concept Auditing for Shared Diffusion Models at Scale

arXiv:2504.14815v2 Announce Type: replace-cross Abstract: Diffusion models (DMs) have revolutionized text-to-image generation, enabling the creation of highly realistic and customized images from text prompts. With the rise of parameter-efficient fine-tuning (PEFT) techniques, users can now customize powerful pre-trained models using minimal computational resources. However, the widespread sharing of fine-tuned DMs on open platforms raises growing ethical and legal concerns, as these models may inadvertently or deliberately generate sensitive or unauthorized content. Despite increasing regulatory attention on generative AI, there are currently no practical tools for systematically auditing these models before deployment. In this paper, we address the problem of concept auditing: determining whether a fine-tuned DM has learned to generate a specific target concept. Existing approaches typically rely on prompt-based input crafting and output-based image classification but they suffer from critical limitations, including prompt uncertainty, concept drift, and poor scalability. To overcome these challenges, we introduce Prompt-Agnostic Image-Free Auditing (PAIA), a novel, model-centric concept auditing framework. By treating the DM as the object of inspection, PAIA enables direct analysis of internal model behavior, bypassing the need for optimized prompts or generated images. We evaluate PAIA on 320 controlled models trained with curated concept datasets and 771 real-world community models sourced from a public DM sharing platform. Evaluation results show that PAIA achieves over 90% detection accuracy while reducing auditing time by 18 - 40X compared to existing baselines. To our knowledge, PAIA is the first scalable and practical solution for pre-deployment concept auditing of diffusion models, providing a practical foundation for safer and more transparent diffusion model sharing.

Physics-Based Motion Imitation with Adversarial Differential Discriminators

arXiv:2505.04961v2 Announce Type: replace-cross Abstract: Multi-objective optimization problems, which require the simultaneous optimization of multiple objectives, are prevalent across numerous applications. Existing multi-objective optimization methods often rely on manually-tuned aggregation functions to formulate a joint optimization objective. The performance of such hand-tuned methods is heavily dependent on careful weight selection, a time-consuming and laborious process. These limitations also arise in the setting of reinforcement-learning-based motion tracking methods for physically simulated characters, where intricately crafted reward functions are typically used to achieve high-fidelity results. Such solutions not only require domain expertise and significant manual tuning, but also limit the applicability of the resulting reward function across diverse skills. To bridge this gap, we present a novel adversarial multi-objective optimization technique that is broadly applicable to a range of multi-objective reinforcement-learning tasks, including motion tracking. Our proposed Adversarial Differential Discriminator (ADD) receives a single positive sample, yet is still effective at guiding the optimization process. We demonstrate that our technique can enable characters to closely replicate a variety of acrobatic and agile behaviors, achieving comparable quality to state-of-the-art motion-tracking methods, without relying on manually-designed reward functions. Code and results are available at https://add-moo.github.io/.

Neural Brain: A Neuroscience-inspired Framework for Embodied Agents

arXiv:2505.07634v3 Announce Type: replace-cross Abstract: The rapid evolution of artificial intelligence (AI) has shifted from static, data-driven models to dynamic systems capable of perceiving and interacting with real-world environments. Despite advancements in pattern recognition and symbolic reasoning, current AI systems, such as large language models, remain disembodied, unable to physically engage with the world. This limitation has driven the rise of embodied AI, where autonomous agents, such as humanoid robots, must navigate and manipulate unstructured environments with human-like adaptability. At the core of this challenge lies the concept of Neural Brain, a central intelligence system designed to drive embodied agents with human-like adaptability. A Neural Brain must seamlessly integrate multimodal sensing and perception with cognitive capabilities. Achieving this also requires an adaptive memory system and energy-efficient hardware-software co-design, enabling real-time action in dynamic environments. This paper introduces a unified framework for the Neural Brain of embodied agents, addressing two fundamental challenges: (1) defining the core components of Neural Brain and (2) bridging the gap between static AI models and the dynamic adaptability required for real-world deployment. To this end, we propose a biologically inspired architecture that integrates multimodal active sensing, perception-cognition-action function, neuroplasticity-based memory storage and updating, and neuromorphic hardware/software optimization. Furthermore, we also review the latest research on embodied agents across these four aspects and analyze the gap between current AI systems and human intelligence. By synthesizing insights from neuroscience, we outline a roadmap towards the development of generalizable, autonomous agents capable of human-level intelligence in real-world scenarios.

TACO: Enhancing Multimodal In-context Learning via Task Mapping-Guided Sequence Configuration

arXiv:2505.17098v3 Announce Type: replace-cross Abstract: Multimodal in-context learning (ICL) has emerged as a key mechanism for harnessing the capabilities of large vision-language models (LVLMs). However, its effectiveness remains highly sensitive to the quality of input ICL sequences, particularly for tasks involving complex reasoning or open-ended generation. A major limitation is our limited understanding of how LVLMs actually exploit these sequences during inference. To bridge this gap, we systematically interpret multimodal ICL through the lens of task mapping, which reveals how local and global relationships within and among demonstrations guide model reasoning. Building on this insight, we present TACO, a lightweight transformer-based model equipped with task-aware attention that dynamically configures ICL sequences. By injecting task-mapping signals into the autoregressive decoding process, TACO creates a bidirectional synergy between sequence construction and task reasoning. Experiments on five LVLMs and nine datasets demonstrate that TACO consistently surpasses baselines across diverse ICL tasks. These results position task mapping as a novel and valuable perspective for interpreting and improving multimodal ICL.

How to build a consistency model: Learning flow maps via self-distillation

arXiv:2505.18825v2 Announce Type: replace-cross Abstract: Flow-based generative models achieve state-of-the-art sample quality, but require the expensive solution of a differential equation at inference time. Flow map models, commonly known as consistency models, encompass many recent efforts to improve inference-time efficiency by learning the solution operator of this differential equation. Yet despite their promise, these models lack a unified description that clearly explains how to learn them efficiently in practice. Here, building on the methodology proposed in Boffi et. al. (2024), we present a systematic algorithmic framework for directly learning the flow map associated with a flow or diffusion model. By exploiting a relationship between the velocity field underlying a continuous-time flow and the instantaneous rate of change of the flow map, we show how to convert any distillation scheme into a direct training algorithm via self-distillation, eliminating the need for pre-trained teachers. We introduce three algorithmic families based on different mathematical characterizations of the flow map: Eulerian, Lagrangian, and Progressive methods, which we show encompass and extend all known distillation and direct training schemes for consistency models. We find that the novel class of Lagrangian methods, which avoid both spatial derivatives and bootstrapping from small steps by design, achieve significantly more stable training and higher performance than more standard Eulerian and Progressive schemes. Our methodology unifies existing training schemes under a single common framework and reveals new design principles for accelerated generative modeling. Associated code is available at https://github.com/nmboffi/flow-maps.

FreqPolicy: Frequency Autoregressive Visuomotor Policy with Continuous Tokens

arXiv:2506.01583v2 Announce Type: replace-cross Abstract: Learning effective visuomotor policies for robotic manipulation is challenging, as it requires generating precise actions while maintaining computational efficiency. Existing methods remain unsatisfactory due to inherent limitations in the essential action representation and the basic network architectures. We observe that representing actions in the frequency domain captures the structured nature of motion more effectively: low-frequency components reflect global movement patterns, while high-frequency components encode fine local details. Additionally, robotic manipulation tasks of varying complexity demand different levels of modeling precision across these frequency bands. Motivated by this, we propose a novel paradigm for visuomotor policy learning that progressively models hierarchical frequency components. To further enhance precision, we introduce continuous latent representations that maintain smoothness and continuity in the action space. Extensive experiments across diverse 2D and 3D robotic manipulation benchmarks demonstrate that our approach outperforms existing methods in both accuracy and efficiency, showcasing the potential of a frequency-domain autoregressive framework with continuous tokens for generalized robotic manipulation.Code is available at https://github.com/4DVLab/Freqpolicy

TCDiff++: An End-to-end Trajectory-Controllable Diffusion Model for Harmonious Music-Driven Group Choreography

arXiv:2506.18671v4 Announce Type: replace-cross Abstract: Music-driven dance generation has garnered significant attention due to its wide range of industrial applications, particularly in the creation of group choreography. During the group dance generation process, however, most existing methods still face three primary issues: multi-dancer collisions, single-dancer foot sliding and abrupt swapping in the generation of long group dance. In this paper, we propose TCDiff++, a music-driven end-to-end framework designed to generate harmonious group dance. Specifically, to mitigate multi-dancer collisions, we utilize a dancer positioning embedding to encode temporal and identity information. Additionally, we incorporate a distance-consistency loss to ensure that inter-dancer distances remain within plausible ranges. To address the issue of single-dancer foot sliding, we introduce a swap mode embedding to indicate dancer swapping patterns and design a Footwork Adaptor to refine raw motion, thereby minimizing foot sliding. For long group dance generation, we present a long group diffusion sampling strategy that reduces abrupt position shifts by injecting positional information into the noisy input. Furthermore, we integrate a Sequence Decoder layer to enhance the model's ability to selectively process long sequences. Extensive experiments demonstrate that our TCDiff++ achieves state-of-the-art performance, particularly in long-duration scenarios, ensuring high-quality and coherent group dance generation.

Filling of incomplete sinograms from sparse PET detector configurations using a residual U-Net

arXiv:2506.19600v2 Announce Type: replace-cross Abstract: Long axial field-of-view PET scanners offer increased field-of-view and sensitivity compared to traditional PET scanners. However, a significant cost is associated with the densely packed photodetectors required for the extended-coverage systems, limiting clinical utilisation. To mitigate the cost limitations, alternative sparse system configurations have been proposed, allowing an extended field-of-view PET design with detector costs similar to a standard PET system, albeit at the expense of image quality. In this work, we propose a deep sinogram restoration network to fill in the missing sinogram data. Our method utilises a modified Residual U-Net, trained on clinical PET scans from a GE Signa PET/MR, simulating the removal of 50% of the detectors in a chessboard pattern (retaining only 25% of all lines of response). The model successfully recovers missing counts, with a mean absolute error below two events per pixel, outperforming 2D interpolation in both sinogram and reconstructed image domain. Notably, the predicted sinograms exhibit a smoothing effect, leading to reconstructed images lacking sharpness in finer details. Despite these limitations, the model demonstrates a substantial capacity for compensating for the undersampling caused by the sparse detector configuration. This proof-of-concept study suggests that sparse detector configurations, combined with deep learning techniques, offer a viable alternative to conventional PET scanner designs. This approach supports the development of cost-effective, total body PET scanners, allowing a significant step forward in medical imaging technology.

RAM-W600: A Multi-Task Wrist Dataset and Benchmark for Rheumatoid Arthritis

arXiv:2507.05193v3 Announce Type: replace-cross Abstract: Rheumatoid arthritis (RA) is a common autoimmune disease that has been the focus of research in computer-aided diagnosis (CAD) and disease monitoring. In clinical settings, conventional radiography (CR) is widely used for the screening and evaluation of RA due to its low cost and accessibility. The wrist is a critical region for the diagnosis of RA. However, CAD research in this area remains limited, primarily due to the challenges in acquiring high-quality instance-level annotations. (i) The wrist comprises numerous small bones with narrow joint spaces, complex structures, and frequent overlaps, requiring detailed anatomical knowledge for accurate annotation. (ii) Disease progression in RA often leads to osteophyte, bone erosion (BE), and even bony ankylosis, which alter bone morphology and increase annotation difficulty, necessitating expertise in rheumatology. This work presents a multi-task dataset for wrist bone in CR, including two tasks: (i) wrist bone instance segmentation and (ii) Sharp/van der Heijde (SvdH) BE scoring, which is the first public resource for wrist bone instance segmentation. This dataset comprises 1048 wrist conventional radiographs of 388 patients from six medical centers, with pixel-level instance segmentation annotations for 618 images and SvdH BE scores for 800 images. This dataset can potentially support a wide range of research tasks related to RA, including joint space narrowing (JSN) progression quantification, BE detection, bone deformity evaluation, and osteophyte detection. It may also be applied to other wrist-related tasks, such as carpal bone fracture localization. We hope this dataset will significantly lower the barrier to research on wrist RA and accelerate progress in CAD research within the RA-related domain.

Conformal Prediction for Long-Tailed Classification

arXiv:2507.06867v2 Announce Type: replace-cross Abstract: Many real-world classification problems, such as plant identification, have extremely long-tailed class distributions. In order for prediction sets to be useful in such settings, they should (i) provide good class-conditional coverage, ensuring that rare classes are not systematically omitted from the prediction sets, and (ii) be a reasonable size, allowing users to easily verify candidate labels. Unfortunately, existing conformal prediction methods, when applied to the long-tailed setting, force practitioners to make a binary choice between small sets with poor class-conditional coverage or sets with very good class-conditional coverage but that are extremely large. We propose methods with guaranteed marginal coverage that smoothly trade off between set size and class-conditional coverage. First, we introduce a new conformal score function called prevalence-adjusted softmax that targets macro-coverage, a relaxed notion of class-conditional coverage. Second, we propose a new procedure that interpolates between marginal and class-conditional conformal prediction by linearly interpolating their conformal score thresholds. We demonstrate our methods on Pl@ntNet-300K and iNaturalist-2018, two long-tailed image datasets with 1,081 and 8,142 classes, respectively.

Depth-Sequence Transformer (DST) for Segment-Specific ICA Calcification Mapping on Non-Contrast CT

arXiv:2507.08214v3 Announce Type: replace-cross Abstract: While total intracranial carotid artery calcification (ICAC) volume is an established stroke biomarker, growing evidence shows this aggregate metric ignores the critical influence of plaque location, since calcification in different segments carries distinct prognostic and procedural risks. However, a finer-grained, segment-specific quantification has remained technically infeasible. Conventional 3D models are forced to process downsampled volumes or isolated patches, sacrificing the global context required to resolve anatomical ambiguity and render reliable landmark localization. To overcome this, we reformulate the 3D challenge as a \textbf{Parallel Probabilistic Landmark Localization} task along the 1D axial dimension. We propose the \textbf{Depth-Sequence Transformer (DST)}, a framework that processes full-resolution CT volumes as sequences of 2D slices, learning to predict $N=6$ independent probability distributions that pinpoint key anatomical landmarks. Our DST framework demonstrates exceptional accuracy and robustness. Evaluated on a 100-patient clinical cohort with rigorous 5-fold cross-validation, it achieves a Mean Absolute Error (MAE) of \textbf{0.1 slices}, with \textbf{96\%} of predictions falling within a $\pm1$ slice tolerance. Furthermore, to validate its architectural power, the DST backbone establishes the best result on the public Clean-CC-CCII classification benchmark under an end-to-end evaluation protocol. Our work delivers the first practical tool for automated segment-specific ICAC analysis. The proposed framework provides a foundation for further studies on the role of location-specific biomarkers in diagnosis, prognosis, and procedural planning.

MapIQ: Evaluating Multimodal Large Language Models for Map Question Answering

arXiv:2507.11625v2 Announce Type: replace-cross Abstract: Recent advancements in multimodal large language models (MLLMs) have driven researchers to explore how well these models read data visualizations, e.g., bar charts, scatter plots. More recently, attention has shifted to visual question answering with maps (Map-VQA). However, Map-VQA research has primarily focused on choropleth maps, which cover only a limited range of thematic categories and visual analytical tasks. To address these gaps, we introduce MapIQ, a benchmark dataset comprising 14,706 question-answer pairs across three map types: choropleth maps, cartograms, and proportional symbol maps spanning topics from six distinct themes (e.g., housing, crime). We evaluate multiple MLLMs using six visual analytical tasks, comparing their performance against one another and a human baseline. An additional experiment examining the impact of map design changes (e.g., altered color schemes, modified legend designs, and removal of map elements) provides insights into the robustness and sensitivity of MLLMs, their reliance on internal geographic knowledge, and potential avenues for improving Map-VQA performance.

EMedNeXt: An Enhanced Brain Tumor Segmentation Framework for Sub-Saharan Africa using MedNeXt V2 with Deep Supervision

arXiv:2507.23256v2 Announce Type: replace-cross Abstract: Brain cancer affects millions worldwide, and in nearly every clinical setting, doctors rely on magnetic resonance imaging (MRI) to diagnose and monitor gliomas. However, the current standard for tumor quantification through manual segmentation of multi-parametric MRI is time-consuming, requires expert radiologists, and is often infeasible in under-resourced healthcare systems. This problem is especially pronounced in low-income regions, where MRI scanners are of lower quality and radiology expertise is scarce, leading to incorrect segmentation and quantification. In addition, the number of acquired MRI scans in Africa is typically small. To address these challenges, the BraTS-Lighthouse 2025 Challenge focuses on robust tumor segmentation in sub-Saharan Africa (SSA), where resource constraints and image quality degradation introduce significant shifts. In this study, we present EMedNeXt -- an enhanced brain tumor segmentation framework based on MedNeXt V2 with deep supervision and optimized post-processing pipelines tailored for SSA. EMedNeXt introduces three key contributions: a larger region of interest, an improved nnU-Net v2-based architectural skeleton, and a robust model ensembling system. Evaluated on the hidden validation set, our solution achieved an average LesionWise DSC of 0.897 with an average LesionWise NSD of 0.541 and 0.84 at a tolerance of 0.5 mm and 1.0 mm, respectively.

OpenCUA: Open Foundations for Computer-Use Agents

arXiv:2508.09123v3 Announce Type: replace-cross Abstract: Vision-language models have demonstrated impressive capabilities as computer-use agents (CUAs) capable of automating diverse computer tasks. As their commercial potential grows, critical details of the most capable CUA systems remain closed. As these agents will increasingly mediate digital interactions and execute consequential decisions on our behalf, the research community needs access to open CUA frameworks to study their capabilities, limitations, and risks. To bridge this gap, we propose OpenCUA, a comprehensive open-source framework for scaling CUA data and foundation models. Our framework consists of: (1) an annotation infrastructure that seamlessly captures human computer-use demonstrations; (2) AgentNet, the first large-scale computer-use task dataset spanning 3 operating systems and 200+ applications and websites; (3) a scalable pipeline that transforms demonstrations into state-action pairs with reflective long Chain-of-Thought reasoning that sustain robust performance gains as data scales. Our end-to-end agent models demonstrate strong performance across CUA benchmarks. In particular, OpenCUA-72B achieves an average success rate of 45.0% on OSWorld-Verified, establishing a new state-of-the-art (SOTA) among open-source models. Further analysis confirms that our approach generalizes well across domains and benefits significantly from increased test-time computation. We release our annotation tool, datasets, code, and models to build open foundations for further CUA research.

Negative Shanshui: Real-time Interactive Ink Painting Synthesis

arXiv:2508.16612v2 Announce Type: replace-cross Abstract: This paper presents Negative Shanshui, a real-time interactive AI synthesis approach that reinterprets classical Chinese landscape ink painting, i.e., shanshui, to engage with ecological crises in the Anthropocene. Negative Shanshui optimizes a fine-tuned Stable Diffusion model for real-time inferences and integrates it with gaze-driven inpainting, frame interpolation; it enables dynamic morphing animations in response to the viewer's gaze and presents as an interactive virtual reality (VR) experience. The paper describes the complete technical pipeline, covering the system framework, optimization strategies, gaze-based interaction, and multimodal deployment in an art festival. Further analysis of audience feedback collected during its public exhibition highlights how participants variously engaged with the work through empathy, ambivalence, and critical reflection.

Mixture of Contexts for Long Video Generation

arXiv:2508.21058v2 Announce Type: replace-cross Abstract: Long video generation is fundamentally a long context memory problem: models must retain and retrieve salient events across a long range without collapsing or drifting. However, scaling diffusion transformers to generate long-context videos is fundamentally limited by the quadratic cost of self-attention, which makes memory and computation intractable and difficult to optimize for long sequences. We recast long-context video generation as an internal information retrieval task and propose a simple, learnable sparse attention routing module, Mixture of Contexts (MoC), as an effective long-term memory retrieval engine. In MoC, each query dynamically selects a few informative chunks plus mandatory anchors (caption, local windows) to attend to, with causal routing that prevents loop closures. As we scale the data and gradually sparsify the routing, the model allocates compute to salient history, preserving identities, actions, and scenes over minutes of content. Efficiency follows as a byproduct of retrieval (near-linear scaling), which enables practical training and synthesis, and the emergence of memory and consistency at the scale of minutes.

QDFlow: A Python package for physics simulations of quantum dot devices

arXiv:2509.13298v2 Announce Type: replace-cross Abstract: Recent advances in machine learning (ML) have accelerated progress in calibrating and operating quantum dot (QD) devices. However, most ML approaches rely on access to large, representative datasets designed to capture the full spectrum of data quality encountered in practice, with both high- and low-quality data for training, benchmarking, and validation, with labels capturing key features of the device state. Collating such datasets experimentally is challenging due to limited data availability, slow measurement bandwidths, and the labor-intensive nature of labeling. QDFlow is an open-source physics simulator for multi-QD arrays that generates realistic synthetic data with ground-truth labels. QDFlow combines a self-consistent Thomas-Fermi solver, a dynamic capacitance model, and flexible noise modules to simulate charge stability diagrams and ray-based data closely resembling experiments. With an extensive set of parameters that can be varied and customizable noise models, QDFlow supports the creation of large, diverse datasets for ML development, benchmarking, and quantum device research.

Interpreting deep learning-based stellar mass estimation via causal analysis and mutual information decomposition

arXiv:2509.23901v2 Announce Type: replace-cross Abstract: End-to-end deep learning models fed with multi-band galaxy images are powerful data-driven tools used to estimate galaxy physical properties in the absence of spectroscopy. However, due to a lack of interpretability and the associational nature of such models, it is difficult to understand how the information that is included in addition to integrated photometry (e.g., morphology) contributes to the estimation task. Improving our understanding in this field would enable further advances into unraveling the physical connections among galaxy properties and optimizing data exploitation. Therefore, our work is aimed at interpreting the deep learning-based estimation of stellar mass via two interpretability techniques: causal analysis and mutual information decomposition. The former reveals the causal paths between multiple variables beyond nondirectional statistical associations, while the latter quantifies the multicomponent contributions (i.e., redundant, unique, and synergistic) of different input data to the stellar mass estimation. Using data from the Sloan Digital Sky Survey (SDSS) and the Wide-field Infrared Survey Explorer (WISE), we obtained meaningful results that provide physical interpretations for image-based models. Our work demonstrates the gains from combining deep learning with interpretability techniques, and holds promise in promoting more data-driven astrophysical research (e.g., astrophysical parameter estimations and investigations on complex multivariate physical processes).

InfMasking: Unleashing Synergistic Information by Contrastive Multimodal Interactions

arXiv:2509.25270v2 Announce Type: replace-cross Abstract: In multimodal representation learning, synergistic interactions between modalities not only provide complementary information but also create unique outcomes through specific interaction patterns that no single modality could achieve alone. Existing methods may struggle to effectively capture the full spectrum of synergistic information, leading to suboptimal performance in tasks where such interactions are critical. This is particularly problematic because synergistic information constitutes the fundamental value proposition of multimodal representation. To address this challenge, we introduce InfMasking, a contrastive synergistic information extraction method designed to enhance synergistic information through an Infinite Masking strategy. InfMasking stochastically occludes most features from each modality during fusion, preserving only partial information to create representations with varied synergistic patterns. Unmasked fused representations are then aligned with masked ones through mutual information maximization to encode comprehensive synergistic information. This infinite masking strategy enables capturing richer interactions by exposing the model to diverse partial modality combinations during training. As computing mutual information estimates with infinite masking is computationally prohibitive, we derive an InfMasking loss to approximate this calculation. Through controlled experiments, we demonstrate that InfMasking effectively enhances synergistic information between modalities. In evaluations on large-scale real-world datasets, InfMasking achieves state-of-the-art performance across seven benchmarks. Code is released at https://github.com/brightest66/InfMasking.

PARS: Low-Latency LLM Serving via Pairwise Learning-to-Rank

arXiv:2510.03243v1 Announce Type: new Abstract: Efficient scheduling of LLM inference tasks is essential for achieving low latency and high throughput, particularly with the growing use of reasoning-capable LLMs. Traditional strategies like First-Come-First-Serve (FCFS) often suffer from Head-of-Line (HOL) blocking, where long-running tasks delay shorter ones queued behind them. In this paper, we introduce PARS, a prompt-aware LLM task scheduler that improves serving efficiency by approximating shortest-job-first (SJF) scheduling through pairwise ranking with margin ranking loss. PARS focuses on impactful scheduling decisions and is seamlessly integrated into the state-of-the-art LLM serving system vLLM. It effectively predicts response-length-based task ordering, reducing latency with minimal overhead. Extensive experiments across multiple LLMs and real-world inference datasets show that PARS significantly improves performance, including for reasoning workloads. Furthermore, our cross-model evaluations demonstrate that the design generalizes well, enabling effective scheduling even when predictors are trained on different LLMs.

VIFO: Visual Feature Empowered Multivariate Time Series Forecasting with Cross-Modal Fusion

arXiv:2510.03244v1 Announce Type: new Abstract: Large time series foundation models often adopt channel-independent architectures to handle varying data dimensions, but this design ignores crucial cross-channel dependencies. Concurrently, existing multimodal approaches have not fully exploited the power of large vision models (LVMs) to interpret spatiotemporal data. Additionally, there remains significant unexplored potential in leveraging the advantages of information extraction from different modalities to enhance time series forecasting performance. To address these gaps, we propose the VIFO, a cross-modal forecasting model. VIFO uniquely renders multivariate time series into image, enabling pre-trained LVM to extract complex cross-channel patterns that are invisible to channel-independent models. These visual features are then aligned and fused with representations from the time series modality. By freezing the LVM and training only 7.45% of its parameters, VIFO achieves competitive performance on multiple benchmarks, offering an efficient and effective solution for capturing cross-variable relationships in

Frequency-Aware Model Parameter Explorer: A new attribution method for improving explainability

arXiv:2510.03245v1 Announce Type: new Abstract: Ensuring the reliability of deep neural networks (DNNs) in the presence of real world noise and intentional perturbations remains a significant challenge. To address this, attribution methods have been proposed, though their efficacy remains suboptimal and necessitates further refinement. In this paper, we propose a novel category of transferable adversarial attacks, called transferable frequency-aware attacks, enabling frequency-aware exploration via both high-and low-frequency components. Based on this type of attacks, we also propose a novel attribution method, named Frequency-Aware Model Parameter Explorer (FAMPE), which improves the explainability for DNNs. Relative to the current state-of-the-art method AttEXplore, our FAMPE attains an average gain of 13.02% in Insertion Score, thereby outperforming existing approaches. Through detailed ablation studies, we also investigate the role of both high- and low-frequency components in explainability.

StructPrune: Structured Global Pruning asymptotics with $\mathcal{O}(\sqrt{N})$ GPU Memory

arXiv:2510.03246v1 Announce Type: new Abstract: Pruning is critical for scaling large language models (LLMs). Global pruning achieves strong performance but requires $\mathcal{O}(N)$ memory, which is infeasible for billion-parameter models. Local pruning reduces GPU memory usage to that of a single layer by pruning layers independently, but it neglects inter-layer dependencies and often leads to suboptimal performance in high-sparsity regimes. Unlike unstructured pruning, structured pruning produces regular sparsity patterns that align well with GPU kernels and library optimizations, making it more hardware-efficient. However, structured pruning typically relies on global pruning, since structured patterns are more prone to severe performance degradation under local optimization. To jointly achieve structured pruning and the memory efficiency of local pruning, we propose a divide-and-conquer strategy that decomposes the global pruning problem into coordinated subproblems across different modules, each of which fits within limited GPU memory. Building on this idea, we design \textbf{STRUPRUNE}, an ADMM-based framework that integrates structured sparsity into the pruning process, combining the memory efficiency of local pruning with the hardware compatibility of structured methods. We derive a closed-form analytical solution for structured pruning masks that provides an explicit rule for layer-wise sparsity allocation, and further develop an energy-based asymptotic framework yielding a softmax-form allocation scheme that simplifies optimization while adapting to heterogeneous layer importance. Experiments demonstrate that STRUPRUNE matches the perplexity of global structured pruning while reducing memory cost from $\mathcal{O}(N)$ to $\mathcal{O}(\sqrt{N})$, enabling practical deployment at the billion-parameter scale.

Towards Multimodal Active Learning: Efficient Learning with Limited Paired Data

arXiv:2510.03247v1 Announce Type: new Abstract: Active learning (AL) is a principled strategy to reduce annotation cost in data-hungry deep learning. However, existing AL algorithms focus almost exclusively on unimodal data, overlooking the substantial annotation burden in multimodal learning. We introduce the first framework for multimodal active learning with unaligned data, where the learner must actively acquire cross-modal alignments rather than labels on pre-aligned pairs. This setting captures the practical bottleneck in modern multimodal pipelines such as CLIP and SigLIP, where unimodal features are easy to obtain but high-quality alignment is costly. We develop a new algorithm that combines uncertainty and diversity principles in a modality-aware design, achieves linear-time acquisition, and applies seamlessly to both pool-based and streaming-based settings. Extensive experiments on benchmark datasets demonstrate that our approach consistently reduces multimodal annotation cost while preserving performance; for instance, on the ColorSwap dataset it cuts annotation requirements by up to $40\%$ without loss in accuracy.

Real-Time Brain Biomechanics Prediction with Neural Operators: Toward Clinically Deployable Traumatic Brain Injury Models

arXiv:2510.03248v1 Announce Type: new Abstract: Traumatic brain injury (TBI) remains a major public health concern, with over 69 million cases annually worldwide. Finite element (FE) models offer high-fidelity predictions of brain deformation but are computationally expensive, requiring hours per simulation and limiting their clinical utility for rapid decision-making. This study benchmarks state-of-the-art neural operator (NO) architectures for rapid, patient-specific prediction of brain displacement fields, aiming to enable real-time TBI modeling in clinical and translational settings. We formulated TBI modeling as an operator learning problem, mapping subject-specific anatomical MRI, magnetic resonance elastography (MRE) stiffness maps, and demographic features to full-field 3D brain displacement predictions. Four architectures - Fourier Neural Operator (FNO), Factorized FNO (F-FNO), Multi-Grid FNO (MG-FNO), and Deep Operator Network (DeepONet) were trained and evaluated on 249 MRE datasets across physiologically relevant frequencies (20 - 90 Hz). MG-FNO achieved the highest accuracy (MSE = 0.0023, 94.3\% spatial fidelity) and preserved fine-scale features, while F-FNO converged 2$\times$ faster than standard FNO. DeepONet offered the fastest inference (14.5 iterations/s) with a 7$\times$ computational speed-up over MG-FNO, suggesting utility for embedded or edge computing applications. All NOs reduced computation time from hours to milliseconds without sacrificing anatomical realism. NOs provide an efficient, resolution-invariant approach for predicting brain deformation, opening the door to real-time, patient-specific TBI risk assessment, clinical triage support, and optimization of protective equipment. These results highlight the potential for NO-based digital twins of the human brain, enabling scalable, on-demand biomechanical modeling in both clinical and population health contexts.

Light Differentiable Logic Gate Networks

arXiv:2510.03250v1 Announce Type: new Abstract: Differentiable logic gate networks (DLGNs) exhibit extraordinary efficiency at inference while sustaining competitive accuracy. But vanishing gradients, discretization errors, and high training cost impede scaling these networks. Even with dedicated parameter initialization schemes from subsequent works, increasing depth still harms accuracy. We show that the root cause of these issues lies in the underlying parametrization of logic gate neurons themselves. To overcome this issue, we propose a reparametrization that also shrinks the parameter size logarithmically in the number of inputs per gate. For binary inputs, this already reduces the model size by 4x, speeds up the backward pass by up to 1.86x, and converges in 8.5x fewer training steps. On top of that, we show that the accuracy on CIFAR-100 remains stable and sometimes superior to the original parametrization.

Numerion: A Multi-Hypercomplex Model for Time Series Forecasting

arXiv:2510.03251v1 Announce Type: new Abstract: Many methods aim to enhance time series forecasting by decomposing the series through intricate model structures and prior knowledge, yet they are inevitably limited by computational complexity and the robustness of the assumptions. Our research uncovers that in the complex domain and higher-order hypercomplex spaces, the characteristic frequencies of time series naturally decrease. Leveraging this insight, we propose Numerion, a time series forecasting model based on multiple hypercomplex spaces. Specifically, grounded in theoretical support, we generalize linear layers and activation functions to hypercomplex spaces of arbitrary power-of-two dimensions and introduce a novel Real-Hypercomplex-Real Domain Multi-Layer Perceptron (RHR-MLP) architecture. Numerion utilizes multiple RHR-MLPs to map time series into hypercomplex spaces of varying dimensions, naturally decomposing and independently modeling the series, and adaptively fuses the latent patterns exhibited in different spaces through a dynamic fusion mechanism. Experiments validate the model`s performance, achieving state-of-the-art results on multiple public datasets. Visualizations and quantitative analyses comprehensively demonstrate the ability of multi-dimensional RHR-MLPs to naturally decompose time series and reveal the tendency of higher dimensional hypercomplex spaces to capture lower frequency features.

Universal Multi-Domain Translation via Diffusion Routers

arXiv:2510.03252v1 Announce Type: new Abstract: Multi-domain translation (MDT) aims to learn translations between multiple domains, yet existing approaches either require fully aligned tuples or can only handle domain pairs seen in training, limiting their practicality and excluding many cross-domain mappings. We introduce universal MDT (UMDT), a generalization of MDT that seeks to translate between any pair of $K$ domains using only $K-1$ paired datasets with a central domain. To tackle this problem, we propose Diffusion Router (DR), a unified diffusion-based framework that models all central$\leftrightarrow$non-central translations with a single noise predictor conditioned on the source and target domain labels. DR enables indirect non-central translations by routing through the central domain. We further introduce a novel scalable learning strategy with a variational-bound objective and an efficient Tweedie refinement procedure to support direct non-central mappings. Through evaluation on three large-scale UMDT benchmarks, DR achieves state-of-the-art results for both indirect and direct translations, while lowering sampling cost and unlocking novel tasks such as sketch$\leftrightarrow$segmentation. These results establish DR as a scalable and versatile framework for universal translation across multiple domains.

Solving the Granularity Mismatch: Hierarchical Preference Learning for Long-Horizon LLM Agents

arXiv:2510.03253v1 Announce Type: new Abstract: Large Language Models (LLMs) as autonomous agents are increasingly tasked with solving complex, long-horizon problems. Aligning these agents via preference-based offline methods like Direct Preference Optimization (DPO) is a promising direction, yet it faces a critical granularity mismatch. Trajectory-level DPO provides a signal that is too coarse for precise credit assignment, while step-level DPO is often too myopic to capture the value of multi-step behaviors. To resolve this challenge, we introduce Hierarchical Preference Learning (HPL), a hierarchical framework that optimizes LLM agents by leveraging preference signals at multiple, synergistic granularities. While HPL incorporates trajectory- and step-level DPO for global and local policy stability, its core innovation lies in group-level preference optimization guided by a dual-layer curriculum. Our approach first decomposes expert trajectories into semantically coherent action groups and then generates contrasting suboptimal groups to enable preference learning at a fine-grained, sub-task level. Then, instead of treating all preference pairs equally, HPL introduces a curriculum scheduler that organizes the learning process from simple to complex. This curriculum is structured along two axes: the group length, representing sub-task complexity, and the sample difficulty, defined by the reward gap between preferred and dispreferred action groups. Experiments on three challenging agent benchmarks show that HPL outperforms existing state-of-the-art methods. Our analyses demonstrate that the hierarchical DPO loss effectively integrates preference signals across multiple granularities, while the dual-layer curriculum is crucial for enabling the agent to solve a wide range of tasks, from simple behaviors to complex multi-step sequences.

Adversarial training with restricted data manipulation

arXiv:2510.03254v1 Announce Type: new Abstract: Adversarial machine learning concerns situations in which learners face attacks from active adversaries. Such scenarios arise in applications such as spam email filtering, malware detection and fake image generation, where security methods must be actively updated to keep up with the everimproving generation of malicious data. Pessimistic Bilevel optimisation has been shown to be an effective method of training resilient classifiers against such adversaries. By modelling these scenarios as a game between the learner and the adversary, we anticipate how the adversary will modify their data and then train a resilient classifier accordingly. However, since existing pessimistic bilevel approaches feature an unrestricted adversary, the model is vulnerable to becoming overly pessimistic and unrealistic. When finding the optimal solution that defeats the classifier, it is possible that the adversary's data becomes nonsensical and loses its intended nature. Such an adversary will not properly reflect reality, and consequently, will lead to poor classifier performance when implemented on real-world data. By constructing a constrained pessimistic bilevel optimisation model, we restrict the adversary's movements and identify a solution that better reflects reality. We demonstrate through experiments that this model performs, on average, better than the existing approach.

SciTS: Scientific Time Series Understanding and Generation with LLMs

arXiv:2510.03255v1 Announce Type: new Abstract: The scientific reasoning ability of large language models (LLMs) has recently attracted significant attention. Time series, as a fundamental modality in scientific data, presents unique challenges that are often overlooked in current multimodal LLMs, which either encode numerical sequences as text or convert them into images. Such approaches may be insufficient for comprehensive scientific time series understanding and generation. Existing unified time series models typically specialise in either forecasting or analysis, and their effectiveness on non-periodic, heterogeneous scientific signals remains unclear. To address these gaps, we introduce SciTS, a benchmark spanning 12 scientific domains and 43 tasks, with over 50k+ instances, both univariate and multivariate signals ranging from $10^0$ to $10^7$ in length and up to 10~MHz in frequency. We benchmark 17 models, including text-only LLMs, multimodal LLMs, and unified time series models, and find that general-purpose LLMs exhibit stronger generalisability than specialised time series models, while representing time series as text or images limits their performance due to excessively long sequences and loss of numerical precision, respectively. We then introduce TimeOmni, a framework that equips LLMs with the ability to understand and generate time series while remaining compatible with general-purpose LLM training. This work fills a gap in both dedicated benchmarks and modelling frameworks for scientific time series, paving the way for LLMs to understand and generate complex temporal scientific data.

Triple-BERT: Do We Really Need MARL for Order Dispatch on Ride-Sharing Platforms?

arXiv:2510.03257v1 Announce Type: new Abstract: On-demand ride-sharing platforms, such as Uber and Lyft, face the intricate real-time challenge of bundling and matching passengers-each with distinct origins and destinations-to available vehicles, all while navigating significant system uncertainties. Due to the extensive observation space arising from the large number of drivers and orders, order dispatching, though fundamentally a centralized task, is often addressed using Multi-Agent Reinforcement Learning (MARL). However, independent MARL methods fail to capture global information and exhibit poor cooperation among workers, while Centralized Training Decentralized Execution (CTDE) MARL methods suffer from the curse of dimensionality. To overcome these challenges, we propose Triple-BERT, a centralized Single Agent Reinforcement Learning (MARL) method designed specifically for large-scale order dispatching on ride-sharing platforms. Built on a variant TD3, our approach addresses the vast action space through an action decomposition strategy that breaks down the joint action probability into individual driver action probabilities. To handle the extensive observation space, we introduce a novel BERT-based network, where parameter reuse mitigates parameter growth as the number of drivers and orders increases, and the attention mechanism effectively captures the complex relationships among the large pool of driver and orders. We validate our method using a real-world ride-hailing dataset from Manhattan. Triple-BERT achieves approximately an 11.95% improvement over current state-of-the-art methods, with a 4.26% increase in served orders and a 22.25% reduction in pickup times. Our code, trained model parameters, and processed data are publicly available at the repository https://github.com/RS2002/Triple-BERT .

POEM: Explore Unexplored Reliable Samples to Enhance Test-Time Adaptation

arXiv:2510.03258v1 Announce Type: new Abstract: Test-time adaptation (TTA) aims to transfer knowledge from a source model to unknown test data with potential distribution shifts in an online manner. Many existing TTA methods rely on entropy as a confidence metric to optimize the model. However, these approaches are sensitive to the predefined entropy threshold, influencing which samples are chosen for model adaptation. Consequently, potentially reliable target samples are often overlooked and underutilized. For instance, a sample's entropy might slightly exceed the threshold initially, but fall below it after the model is updated. Such samples can provide stable supervised information and offer a normal range of gradients to guide model adaptation. In this paper, we propose a general approach, \underline{POEM}, to promote TTA via ex\underline{\textbf{p}}loring the previously unexpl\underline{\textbf{o}}red reliabl\underline{\textbf{e}} sa\underline{\textbf{m}}ples. Additionally, we introduce an extra Adapt Branch network to strike a balance between extracting domain-agnostic representations and achieving high performance on target data. Comprehensive experiments across multiple architectures demonstrate that POEM consistently outperforms existing TTA methods in both challenging scenarios and real-world domain shifts, while remaining computationally efficient. The effectiveness of POEM is evaluated through extensive analyses and thorough ablation studies. Moreover, the core idea behind POEM can be employed as an augmentation strategy to boost the performance of existing TTA approaches. The source code is publicly available at \emph{https://github.com/ycarobot/POEM}

Meta-Awareness Enhances Reasoning Models: Self-Alignment Reinforcement Learning

arXiv:2510.03259v1 Announce Type: new Abstract: Recent studies on reasoning models explore the meta-awareness of language models, the ability to know how to think by itself. We argue that large reasoning models lack this meta-awareness property by proving severe misalignment between true rollouts and predicted meta information. We posit that aligning meta-prediction with true rollouts will lead to significant performance gains. To verify this hypothesis, we design a training pipeline that boosts Meta-Awareness via Self-Alignment (MASA), and prove that enhanced meta-awareness directly translates to improved accuracy. Unlike existing meta-cognitive reasoning models, our method does not require external training sources but leverages self-generated signals to train meta-awareness. Moreover, our method enables efficient training by i) filtering out zero-variance prompts that are either trivial or unsolvable and ii) cutting off lengthy rollouts when they are unlikely to lead to correct answers. The results are inspiring: our strategy yields significant improvements in both accuracy and training efficiency on in-domain tasks and shows strong generalization to out-of-domain benchmarks. More specifically, our method can speed up GRPO training by over 1.28x to reach the same performance, and achieve a 19.3% gain in accuracy on AIME25, and a 6.2 % average gain over six mathematics benchmarks. Training with meta-cognitive guidance enhances out-of-domain generalization, giving a 3.87 % boost on GPQA-Diamond and a 2.08 % overall accuracy gain across 13 benchmarks spanning logical, scientific, and coding domains.

Semantic-Inductive Attribute Selection for Zero-Shot Learning

arXiv:2510.03260v1 Announce Type: new Abstract: Zero-Shot Learning is an important paradigm within General-Purpose Artificial Intelligence Systems, particularly in those that operate in open-world scenarios where systems must adapt to new tasks dynamically. Semantic spaces play a pivotal role as they bridge seen and unseen classes, but whether human-annotated or generated by a machine learning model, they often contain noisy, redundant, or irrelevant attributes that hinder performance. To address this, we introduce a partitioning scheme that simulates unseen conditions in an inductive setting (which is the most challenging), allowing attribute relevance to be assessed without access to semantic information from unseen classes. Within this framework, we study two complementary feature-selection strategies and assess their generalisation. The first adapts embedded feature selection to the particular demands of ZSL, turning model-driven rankings into meaningful semantic pruning; the second leverages evolutionary computation to directly explore the space of attribute subsets more broadly. Experiments on five benchmark datasets (AWA2, CUB, SUN, aPY, FLO) show that both methods consistently improve accuracy on unseen classes by reducing redundancy, but in complementary ways: RFS is efficient and competitive though dependent on critical hyperparameters, whereas GA is more costly yet explores the search space more broadly and avoids such dependence. These results confirm that semantic spaces are inherently redundant and highlight the proposed partitioning scheme as an effective tool to refine them under inductive conditions.

Data-Driven Temperature Modelling of Machine Tools by Neural Networks: A Benchmark

arXiv:2510.03261v1 Announce Type: new Abstract: Thermal errors in machine tools significantly impact machining precision and productivity. Traditional thermal error correction/compensation methods rely on measured temperature-deformation fields or on transfer functions. Most existing data-driven compensation strategies employ neural networks (NNs) to directly predict thermal errors or specific compensation values. While effective, these approaches are tightly bound to particular error types, spatial locations, or machine configurations, limiting their generality and adaptability. In this work, we introduce a novel paradigm in which NNs are trained to predict high-fidelity temperature and heat flux fields within the machine tool. The proposed framework enables subsequent computation and correction of a wide range of error types using modular, swappable downstream components. The NN is trained using data obtained with the finite element method under varying initial conditions and incorporates a correlation-based selection strategy that identifies the most informative measurement points, minimising hardware requirements during inference. We further benchmark state-of-the-art time-series NN architectures, namely Recurrent NN, Gated Recurrent Unit, Long-Short Term Memory (LSTM), Bidirectional LSTM, Transformer, and Temporal Convolutional Network, by training both specialised models, tailored for specific initial conditions, and general models, capable of extrapolating to unseen scenarios. The results show accurate and low-cost prediction of temperature and heat flux fields, laying the basis for enabling flexible and generalisable thermal error correction in machine tool environments.

Rethinking Inter-LoRA Orthogonality in Adapter Merging: Insights from Orthogonal Monte Carlo Dropout

arXiv:2510.03262v1 Announce Type: new Abstract: We propose Orthogonal Monte Carlo Dropout, a mechanism that enforces strict orthogonality when combining sparse semantic vectors without extra time complexity. LoRA, a popular fine-tuning method for large models, typically trains a module to represent a specific concept such as an object or a style. When multiple LoRAs are merged, for example to generate an object in a particular style, their semantic vectors may interfere with each other. Our method guarantees, at the theoretical and runtime levels, that merged LoRAs remain orthogonal and thus free from direct interference. However, empirical analysis reveals that such orthogonality does not lead to the semantic disentanglement or compositionality highlighted in prior work on compositional adaptation. This finding suggests that inter-LoRA orthogonality alone may be insufficient for achieving true semantic compositionality, prompting a re-examination of its role in adapter merging.

Memory Self-Regeneration: Uncovering Hidden Knowledge in Unlearned Models

arXiv:2510.03263v1 Announce Type: new Abstract: The impressive capability of modern text-to-image models to generate realistic visuals has come with a serious drawback: they can be misused to create harmful, deceptive or unlawful content. This has accelerated the push for machine unlearning. This new field seeks to selectively remove specific knowledge from a model's training data without causing a drop in its overall performance. However, it turns out that actually forgetting a given concept is an extremely difficult task. Models exposed to attacks using adversarial prompts show the ability to generate so-called unlearned concepts, which can be not only harmful but also illegal. In this paper, we present considerations regarding the ability of models to forget and recall knowledge, introducing the Memory Self-Regeneration task. Furthermore, we present MemoRa strategy, which we consider to be a regenerative approach supporting the effective recovery of previously lost knowledge. Moreover, we propose that robustness in knowledge retrieval is a crucial yet underexplored evaluation measure for developing more robust and effective unlearning techniques. Finally, we demonstrate that forgetting occurs in two distinct ways: short-term, where concepts can be quickly recalled, and long-term, where recovery is more challenging.

Front-Loading Reasoning: The Synergy between Pretraining and Post-Training Data

arXiv:2510.03264v1 Announce Type: new Abstract: The prevailing paradigm for enhancing the reasoning abilities of LLMs revolves around post-training on high-quality, reasoning-intensive data. While emerging literature suggests that reasoning data is increasingly incorporated also during the mid-training stage-a practice that is relatively more proprietary and less openly characterized-the role of such data in pretraining remains unclear. In particular, due to the opaqueness of pretraining corpora in most frontier models, the effect of reasoning data introduced at different phases of pre- and/or post-training is relatively less reported in the scientific literature. This raises several important questions: Is adding reasoning data earlier during pretraining any better than introducing it during post-training? Could earlier inclusion risk overfitting and harm generalization, or instead establish durable foundations that later fine-tuning cannot recover? We conduct the first systematic study of how reasoning data-varying in scale, diversity, and quality-affects LLM performance when introduced at different stages of training. We find that front-loading reasoning data into pretraining is critical (19% avg gain), establishing foundational capabilities that cannot be fully replicated by later-stage SFT, even with more data. We uncover an asymmetric principle for optimal data allocation: pretraining benefits most from broad diversity in reasoning patterns (11% avg gain), while SFT is more sensitive to data quality (15% avg gain). We show that high-quality pretraining data has latent effects, activated only after SFT, and that naively scaling SFT data can be detrimental, washing away the benefits of early reasoning injection. Our results challenge the conventional separation of language modeling and reasoning, providing a principled guide for strategically allocating data across the entire training pipeline to build more capable models.

MindCraft: How Concept Trees Take Shape In Deep Models

arXiv:2510.03265v1 Announce Type: new Abstract: Large-scale foundation models demonstrate strong performance across language, vision, and reasoning tasks. However, how they internally structure and stabilize concepts remains elusive. Inspired by causal inference, we introduce the MindCraft framework built upon Concept Trees. By applying spectral decomposition at each layer and linking principal directions into branching Concept Paths, Concept Trees reconstruct the hierarchical emergence of concepts, revealing exactly when they diverge from shared representations into linearly separable subspaces. Empirical evaluations across diverse scenarios across disciplines, including medical diagnosis, physics reasoning, and political decision-making, show that Concept Trees recover semantic hierarchies, disentangle latent concepts, and can be widely applied across multiple domains. The Concept Tree establishes a widely applicable and powerful framework that enables in-depth analysis of conceptual representations in deep models, marking a significant step forward in the foundation of interpretable AI.

Variational Autoencoders-based Detection of Extremes in Plant Productivity in an Earth System Model

arXiv:2510.03266v1 Announce Type: new Abstract: Climate anomalies significantly impact terrestrial carbon cycle dynamics, necessitating robust methods for detecting and analyzing anomalous behavior in plant productivity. This study presents a novel application of variational autoencoders (VAE) for identifying extreme events in gross primary productivity (GPP) from Community Earth System Model version 2 simulations across four AR6 regions in the Continental United States. We compare VAE-based anomaly detection with traditional singular spectral analysis (SSA) methods across three time periods: 1850-80, 1950-80, and 2050-80 under the SSP585 scenario. The VAE architecture employs three dense layers and a latent space with an input sequence length of 12 months, trained on a normalized GPP time series to reconstruct the GPP and identifying anomalies based on reconstruction errors. Extreme events are defined using 5th percentile thresholds applied to both VAE and SSA anomalies. Results demonstrate strong regional agreement between VAE and SSA methods in spatial patterns of extreme event frequencies, despite VAE producing higher threshold values (179-756 GgC for VAE vs. 100-784 GgC for SSA across regions and periods). Both methods reveal increasing magnitudes and frequencies of negative carbon cycle extremes toward 2050-80, particularly in Western and Central North America. The VAE approach shows comparable performance to established SSA techniques, while offering computational advantages and enhanced capability for capturing non-linear temporal dependencies in carbon cycle variability. Unlike SSA, the VAE method does not require one to define the periodicity of the signals in the data; it discovers them from the data.

PT$^2$-LLM: Post-Training Ternarization for Large Language Models

arXiv:2510.03267v1 Announce Type: new Abstract: Large Language Models (LLMs) have shown impressive capabilities across diverse tasks, but their large memory and compute demands hinder deployment. Ternarization has gained attention as a promising compression technique, delivering substantial size reduction and high computational efficiency. However, its potential in the post-training quantization (PTQ) setting remains underexplored, due to the challenge of training-free parameter optimization and the quantization difficulty posed by outliers and dispersed weights. To address these issues, we propose PT$^2$-LLM, a post-training ternarization framework tailored for LLMs. At its core is an Asymmetric Ternary Quantizer equipped with a two-stage refinement pipeline: (1) Iterative Ternary Fitting (ITF), which alternates between optimal ternary grid construction and flexible rounding to minimize quantization error, and (2) Activation-aware Grid Alignment (AGA), which further refines the ternary grid to better match full-precision outputs. In addition, we propose a plug-and-play Structural Similarity-based Reordering (SSR) strategy that leverages inter-column structural similarity to ease quantization and mitigate outlier effects, further enhancing overall performance. Extensive experiments demonstrate that PT$^2$-LLM delivers competitive performance against state-of-the-art (SOTA) 2-bit PTQ methods with lower memory cost, while also accelerating both prefill and decoding to achieve end-to-end speedup. The code and models will be available at https://github.com/XIANGLONGYAN/PT2-LLM.

Decrypt Modality Gap in Multimodal Contrastive Learning: From Convergent Representation to Pair Alignment

arXiv:2510.03268v1 Announce Type: new Abstract: Multimodal contrastive learning (MCL) aims to embed data from different modalities in a shared embedding space. However, empirical evidence shows that representations from different modalities occupy completely separate regions of embedding space, a phenomenon referred to as the modality gap. Moreover, experimental findings on how the size of the modality gap influences downstream performance are inconsistent. These observations raise two key questions: (1) What causes the modality gap? (2) How does it affect downstream tasks? To address these questions, this paper introduces the first theoretical framework for analyzing the convergent optimal representations of MCL and the modality alignment when training is optimized. Specifically, we prove that without any constraint or under the cone constraint, the modality gap converges to zero. Under the subspace constraint (i.e., representations of two modalities fall into two distinct hyperplanes due to dimension collapse), the modality gap converges to the smallest angle between the two hyperplanes. This result identifies \emph{dimension collapse} as the fundamental origin of the modality gap. Furthermore, our theorems demonstrate that paired samples cannot be perfectly aligned under the subspace constraint. The modality gap influences downstream performance by affecting the alignment between sample pairs. We prove that, in this case, perfect alignment between two modalities can still be achieved via two ways: hyperplane rotation and shared space projection.

General Exploratory Bonus for Optimistic Exploration in RLHF

arXiv:2510.03269v1 Announce Type: new Abstract: Optimistic exploration is central to improving sample efficiency in reinforcement learning with human feedback, yet existing exploratory bonus methods to incentivize exploration often fail to realize optimism. We provide a theoretical analysis showing that current formulations, under KL or $\alpha$-divergence regularization, unintentionally bias exploration toward high-probability regions of the reference model, thereby reinforcing conservative behavior instead of promoting discovery of uncertain regions. To address this pitfall, we introduce the General Exploratory Bonus (GEB), a novel theoretical framework that provably satisfies the optimism principle. GEB counteracts divergence-induced bias via reference-dependent reward regulation and unifies prior heuristic bonuses as special cases, while extending naturally across the full $\alpha$-divergence family. Empirically, GEB consistently outperforms baselines on alignment tasks across multiple divergence settings and large language model backbones. These results demonstrate that GEB offers both a principled and practical solution for optimistic exploration in RLHF.

CoDA: Coding LM via Diffusion Adaptation

arXiv:2510.03270v1 Announce Type: new Abstract: Diffusion language models promise bidirectional context and infilling capabilities that autoregressive coders lack, yet practical systems remain heavyweight. We introduce CoDA, a 1.7B-parameter diffusion coder trained on TPU with a fully open-source training pipeline. CoDA pairs large-scale diffusion pre-training with code-centric mid-training and instruction tuning, enabling confidence-guided sampling that keeps inference latency competitive. On Humaneval, MBPP, and EvalPlus, CoDA-1.7B-Instruct matches or surpasses diffusion models up to 7B parameters. Our release includes model checkpoints, evaluation harnesses, and TPU training pipelines to accelerate research on lightweight diffusion-based coding assistants.

Decision Potential Surface: A Theoretical and Practical Approximation of LLM's Decision Boundary

arXiv:2510.03271v1 Announce Type: new Abstract: Decision boundary, the subspace of inputs where a machine learning model assigns equal classification probabilities to two classes, is pivotal in revealing core model properties and interpreting behaviors. While analyzing the decision boundary of large language models (LLMs) has raised increasing attention recently, constructing it for mainstream LLMs remains computationally infeasible due to the enormous vocabulary-sequence sizes and the auto-regressive nature of LLMs. To address this issue, in this paper we propose Decision Potential Surface (DPS), a new notion for analyzing LLM decision boundary. DPS is defined on the confidences in distinguishing different sampling sequences for each input, which naturally captures the potential of decision boundary. We prove that the zero-height isohypse in DPS is equivalent to the decision boundary of an LLM, with enclosed regions representing decision regions. By leveraging DPS, for the first time in the literature, we propose an approximate decision boundary construction algorithm, namely $K$-DPS, which only requires K-finite times of sequence sampling to approximate an LLM's decision boundary with negligible error. We theoretically derive the upper bounds for the absolute error, expected error, and the error concentration between K-DPS and the ideal DPS, demonstrating that such errors can be trade-off with sampling times. Our results are empirically validated by extensive experiments across various LLMs and corpora.

PDE-Transformer: A Continuous Dynamical Systems Approach to Sequence Modeling

arXiv:2510.03272v1 Announce Type: new Abstract: The Transformer architecture has revolutionized artificial intelligence, yet a principled theoretical understanding of its internal mechanisms remains elusive. This paper introduces a novel analytical framework that reconceptualizes the Transformer's discrete, layered structure as a continuous spatiotemporal dynamical system governed by a master Partial Differential Equation (PDE). Within this paradigm, we map core architectural components to distinct mathematical operators: self-attention as a non-local interaction, the feed-forward network as a local reaction, and, critically, residual connections and layer normalization as indispensable stabilization mechanisms. We do not propose a new model, but rather employ the PDE system as a theoretical probe to analyze the mathematical necessity of these components. By comparing a standard Transformer with a PDE simulator that lacks explicit stabilizers, our experiments provide compelling empirical evidence for our central thesis. We demonstrate that without residual connections, the system suffers from catastrophic representational drift, while the absence of layer normalization leads to unstable, explosive training dynamics. Our findings reveal that these seemingly heuristic "tricks" are, in fact, fundamental mathematical stabilizers required to tame an otherwise powerful but inherently unstable continuous system. This work offers a first-principles explanation for the Transformer's design and establishes a new paradigm for analyzing deep neural networks through the lens of continuous dynamics.

Learning without Global Backpropagation via Synergistic Information Distillation

arXiv:2510.03273v1 Announce Type: new Abstract: Backpropagation (BP), while foundational to deep learning, imposes two critical scalability bottlenecks: update locking, where network modules remain idle until the entire backward pass completes, and high memory consumption due to storing activations for gradient computation. To address these limitations, we introduce Synergistic Information Distillation (SID), a novel training framework that reframes deep learning as a cascade of local cooperative refinement problems. In SID, a deep network is structured as a pipeline of modules, each imposed with a local objective to refine a probabilistic belief about the ground-truth target. This objective balances fidelity to the target with consistency to the belief from its preceding module. By decoupling the backward dependencies between modules, SID enables parallel training and hence eliminates update locking and drastically reduces memory requirements. Meanwhile, this design preserves the standard feed-forward inference pass, making SID a versatile drop-in replacement for BP. We provide a theoretical foundation, proving that SID guarantees monotonic performance improvement with network depth. Empirically, SID consistently matches or surpasses the classification accuracy of BP, exhibiting superior scalability and pronounced robustness to label noise.Code is available at: https://github.com/ychAlbert/sid-bp

Quant-dLLM: Post-Training Extreme Low-Bit Quantization for Diffusion Large Language Models

arXiv:2510.03274v1 Announce Type: new Abstract: Diffusion large language models (dLLMs), which offer bidirectional context and flexible masked-denoising generation, are emerging as a compelling alternative to autoregressive (AR) LLMs. However, like AR LLMs, their model sizes continue to grow, motivating weight compression for deployment. Although post-training quantization (PTQ) is effective for AR LLMs, directly transferring it to dLLMs at 2-bit leads to unsatisfactory performance. To tackle these challenges, we propose Quant-dLLM, an ultra-low-bit PTQ framework tailored to dLLMs. Since masked-denoising activations in dLLMs differ from the fully visible signals assumed by standard PTQ methods, we introduce Masked Calibration Simulation (MCS) to align calibration with the timestep-dependent masking, which yields more reliable calibrations. Moreover, we propose a Data-aware Any-order Quantizer (DAQ) that learns ultra-low-bit weight representations via an optimization algorithm. It performs iterative approximation guided by our simulated calibration data. In addition, under a strict 2-bit budget, we introduce Adaptive Blockwise Mixed Precision (ABMP), a sensitivity-based precision allocation scheme that adaptively assigns bit width across channel groups. When restricted to 2-bit precision, Quant-dLLM consistently achieves higher accuracy than state-of-the-art (SOTA) AR-transfer PTQ methods on dLLMs. The code and models will be available at: https://github.com/ZTA2785/Quant-dLLM.

SDQ-LLM: Sigma-Delta Quantization for 1-bit LLMs of any size

arXiv:2510.03275v1 Announce Type: new Abstract: Large language models (LLMs) face significant computational and memory challenges, making extremely low-bit quantization crucial for their efficient deployment. In this work, we introduce SDQ-LLM: Sigma-Delta Quantization for 1-bit LLMs of any size, a novel framework that enables extremely low-bit quantization of LLMs while preserving their linguistic reasoning capabilities. A distinctive feature of SDQ-LLM is the continuous adjustability of the Over-Sampling Ratio (OSR), enabling dynamic adaptation to memory or VRAM constraints by selecting fractional OSR (e.g. 2.5 times) for an optimal trade-off between model size and accuracy. SDQ-LLM uses upsampling combined with Sigma-Delta Quantizer to binarize or ternarize LLMs weights, encoding high-precision parameters into 1-bit or 1.58-bit representations, replacing the multiplication operations within linear layers with addition. This approach significantly enhances inference efficiency under extremely low-bit quantization. To further reduce the loss of quantization precision, we incorporate Hadamard-based weight smoothing prior to quantization, improving the stability and robustness of the weight representations. Furthermore, to fully leverage the continuity of the OSR and reduce precision loss, recognizing the correlation between quantization sensitivity and weight variance, we propose a fine-grained, layer- and linear-wise OSR allocation strategy, MultiOSR. This strategy distributes OSR both across layers and within each layer, based on weight variance and parameter scale. Finally, extensive experiments on OPT and LLaMA model families demonstrate that SDQ-LLM achieves a more efficient and high-precision performance even under highly aggressive low-OSR settings. Our code is available at https://github.com/Dreamlittlecat/LLM-Quant-Factory.

QuadEnhancer: Leveraging Quadratic Transformations to Enhance Deep Neural Networks

arXiv:2510.03276v1 Announce Type: new Abstract: The combination of linear transformations and non-linear activation functions forms the foundation of most modern deep neural networks, enabling them to approximate highly complex functions. This paper explores the introduction of quadratic transformations to further increase nonlinearity in neural networks, with the aim of enhancing the performance of existing architectures. To reduce parameter complexity and computational complexity, we propose a lightweight quadratic enhancer that uses low-rankness, weight sharing, and sparsification techniques. For a fixed architecture, the proposed approach introduces quadratic interactions between features at every layer, while only adding negligible amounts of additional model parameters and forward computations. We conduct a set of proof-of-concept experiments for the proposed method across three tasks: image classification, text classification, and fine-tuning large-language models. In all tasks, the proposed approach demonstrates clear and substantial performance gains.

Quantifying constraint hierarchies in Bayesian PINNs via per-constraint Hessian decomposition

arXiv:2510.03278v1 Announce Type: new Abstract: Bayesian physics-informed neural networks (B-PINNs) merge data with governing equations to solve differential equations under uncertainty. However, interpreting uncertainty and overconfidence in B-PINNs requires care due to the poorly understood effects the physical constraints have on the network; overconfidence could reflect warranted precision, enforced by the constraints, rather than miscalibration. Motivated by the need to further clarify how individual physical constraints shape these networks, we introduce a scalable, matrix-free Laplace framework that decomposes the posterior Hessian into contributions from each constraint and provides metrics to quantify their relative influence on the loss landscape. Applied to the Van der Pol equation, our method tracks how constraints sculpt the network's geometry and shows, directly through the Hessian, how changing a single loss weight non-trivially redistributes curvature and effective dominance across the others.

MemMamba: Rethinking Memory Patterns in State Space Model

arXiv:2510.03279v1 Announce Type: new Abstract: With the explosive growth of data, long-sequence modeling has become increasingly important in tasks such as natural language processing and bioinformatics. However, existing methods face inherent trade-offs between efficiency and memory. Recurrent neural networks suffer from gradient vanishing and explosion, making them hard to scale. Transformers can model global dependencies but are constrained by quadratic complexity. Recently, selective state-space models such as Mamba have demonstrated high efficiency with O(n) time and O(1) recurrent inference, yet their long-range memory decays exponentially. In this work, we conduct mathematical derivations and information-theoretic analysis to systematically uncover the memory decay mechanism of Mamba, answering a fundamental question: what is the nature of Mamba's long-range memory and how does it retain information? To quantify key information loss, we further introduce horizontal-vertical memory fidelity metrics that capture degradation both within and across layers. Inspired by how humans distill and retain salient information when reading long documents, we propose MemMamba, a novel architectural framework that integrates state summarization mechanism together with cross-layer and cross-token attention, which alleviates long-range forgetting while preserving linear complexity. MemMamba achieves significant improvements over existing Mamba variants and Transformers on long-sequence benchmarks such as PG19 and Passkey Retrieval, while delivering a 48% speedup in inference efficiency. Both theoretical analysis and empirical results demonstrate that MemMamba achieves a breakthrough in the complexity-memory trade-off, offering a new paradigm for ultra-long sequence modeling.

Training Optimal Large Diffusion Language Models

arXiv:2510.03280v1 Announce Type: new Abstract: We introduce Quokka, the first systematic scaling law for diffusion language models (DLMs), encompassing both compute-constrained and data-constrained regimes, and studying the key modeling and optimization designs. Quokka is a good friend of Chinchilla and provides wider scopes. We hope the results would bring short-term practical guidance in DLMs training and long-term inspirations for the whole AI community.

Discovering Transformer Circuits via a Hybrid Attribution and Pruning Framework

arXiv:2510.03282v1 Announce Type: new Abstract: Interpreting language models often involves circuit analysis, which aims to identify sparse subnetworks, or circuits, that accomplish specific tasks. Existing circuit discovery algorithms face a fundamental trade-off: attribution patching is fast but unfaithful to the full model, while edge pruning is faithful but computationally expensive. This research proposes a hybrid attribution and pruning (HAP) framework that uses attribution patching to identify a high-potential subgraph, then applies edge pruning to extract a faithful circuit from it. We show that HAP is 46\% faster than baseline algorithms without sacrificing circuit faithfulness. Furthermore, we present a case study on the Indirect Object Identification task, showing that our method preserves cooperative circuit components (e.g. S-inhibition heads) that attribution patching methods prune at high sparsity. Our results show that HAP could be an effective approach for improving the scalability of mechanistic interpretability research to larger models. Our code is available at https://anonymous.4open.science/r/HAP-circuit-discovery.

MACE: A Hybrid LLM Serving System with Colocated SLO-aware Continuous Retraining Alignment

arXiv:2510.03283v1 Announce Type: new Abstract: Large language models (LLMs) deployed on edge servers are increasingly used in latency-sensitive applications such as personalized assistants, recommendation, and content moderation. However, the non-stationary nature of user data necessitates frequent retraining, which introduces a fundamental tension between inference latency and model accuracy under constrained GPU resources. Existing retraining strategies either delay model updates, over-commit resources to retraining, or overlook iteration-level retraining granularity. In this paper, we identify that iteration-level scheduling is crucial for adapting retraining frequency to model drift without violating service-level objectives (SLOs). We propose MACE, a hybrid LLM system that colocates concurrent inference (prefill, decode) and fine-tuning, with intelligent memory management to maximize task performance while promising inference throughput. MACE leverages the insight that not all model updates equally affect output alignment and allocates GPU cycles accordingly to balance throughput, latency, and update freshness. Our trace-driven evaluation shows that MACE matches or exceeds continuous retraining while reducing inference latency by up to 63% and maintaining throughput under resource constraints. Compared to periodic retraining, MACE improves latency breakdown across prefill, decode, and finetune stages, and sustains GPU utilization above 85% in NVIDIA AGX Orin. These results demonstrate that iteration-level hybrid scheduling is a promising direction for deploying LLMs with continual learning capabilities on edge platforms.

Edge-FIT: Federated Instruction Tuning of Quantized LLMs for Privacy-Preserving Smart Home Environments

arXiv:2510.03284v1 Announce Type: new Abstract: This paper proposes Edge-FIT (Federated Instruction Tuning on the Edge), a scalable framework for Federated Instruction Tuning (FIT) of Large Language Models (LLMs). Traditional Federated Learning (TFL) methods, like FedAvg, fail when confronted with the massive parameter size of LLMs [3], [6]. Our Edge-FIT framework combines federated learning with 4-bit Quantized Low-Rank Adaptation (QLORA), mitigating the core issues of communication and computational overhead. We demonstrate this by filtering the general-purpose Databricks Dolly 15k dataset for the IoT domain. Experimental results show the Edge-FIT tuned Llama 2(7B) achieves an F1-Score of 0.89. We also demonstrate a viable trade-off using the 3.8B Phi-3-mini model, validating Edge-FIT as a scalable framework for decentralized LLM deployment on home compute gateways.

LogAction: Consistent Cross-system Anomaly Detection through Logs via Active Domain

arXiv:2510.03288v1 Announce Type: new Abstract: Log-based anomaly detection is a essential task for ensuring the reliability and performance of software systems. However, the performance of existing anomaly detection methods heavily relies on labeling, while labeling a large volume of logs is highly challenging. To address this issue, many approaches based on transfer learning and active learning have been proposed. Nevertheless, their effectiveness is hindered by issues such as the gap between source and target system data distributions and cold-start problems. In this paper, we propose LogAction, a novel log-based anomaly detection model based on active domain adaptation. LogAction integrates transfer learning and active learning techniques. On one hand, it uses labeled data from a mature system to train a base model, mitigating the cold-start issue in active learning. On the other hand, LogAction utilize free energy-based sampling and uncertainty-based sampling to select logs located at the distribution boundaries for manual labeling, thus addresses the data distribution gap in transfer learning with minimal human labeling efforts. Experimental results on six different combinations of datasets demonstrate that LogAction achieves an average 93.01% F1 score with only 2% of manual labels, outperforming some state-of-the-art methods by 26.28%. Website: https://logaction.github.io

Why mask diffusion does not work

arXiv:2510.03289v1 Announce Type: new Abstract: The main advantages of diffusion language models over autoregressive (AR) models lie in their ability to support parallel generation and bidirectional attention, enabling a more controllable generation process. In recent years, open-source mask diffusion language models have emerged, most of which are based on a variant known as absorbing diffusion. However, this paper demonstrates why mask diffusion faces inherent difficulties in achieving parallel generation and bidirectional attention. We also propose the most effective training and inference strategies for mask diffusion.

Single-Core Superscalar Optimization of Clifford Neural Layers

arXiv:2510.03290v1 Announce Type: new Abstract: Within the growing interest in the physical sciences in developing networks with equivariance properties, Clifford neural layers shine as one approach that delivers $E(n)$ and $O(n)$ equivariances given specific group actions. In this paper, we analyze the inner structure of the computation within Clifford convolutional layers and propose and implement several optimizations to speed up the inference process while maintaining correctness. In particular, we begin by analyzing the theoretical foundations of Clifford algebras to eliminate redundant matrix allocations and computations, then systematically apply established optimization techniques to enhance performance further. We report a final average speedup of 21.35x over the baseline implementation of eleven functions and runtimes comparable to and faster than the original PyTorch implementation in six cases. In the remaining cases, we achieve performance in the same order of magnitude as the original library.

UniPruning: Unifying Local Metric and Global Feedback for Scalable Sparse LLMs

arXiv:2510.03291v1 Announce Type: new Abstract: Large Language Models (LLMs) achieve strong performance across diverse tasks but face prohibitive computational and memory costs. Pruning offers a promising path by inducing sparsity while preserving architectural flexibility. However, existing methods struggle to balance efficiency and robustness: local metric approaches prune layer by layer but often collapse under high sparsity, whereas global feedback methods enforce consistency at the cost of expensive weight updates or restrictive semi-structured formats. We present UniPruning, a unified post-training pruning framework that combines the speed of local saliency metrics with the stability of global coordination, enabled by a mirror descent based optimization, all without updating model weights. UniPruning leverages fast layer-wise scoring and a lightweight global controller to allocate a single sparsity budget, supporting both unstructured and semi-structured N :M pruning within one framework. After a brief calibration, it can generate pruning masks for arbitrary sparsity levels in one shot, and adapts seamlessly to hardware-aware constraints. Extensive experiments on multiple pretrained LLM families and standard benchmarks show that UniPruning consistently delivers competitive or superior perplexity and zero-shot accuracy. Ablation studies further highlight the importance of mirror descent and local saliency anchoring. Overall, UniPruning provides an efficient, principled, and scalable solution for sparsifying large-scale LLMs. Our code is available at: https://github.com/RainbowQTT/UniPruning.

From Score Distributions to Balance: Plug-and-Play Mixture-of-Experts Routing

arXiv:2510.03293v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) models can scale parameter capacity by routing each token to a subset of experts through a learned gate function. While conditional routing reduces training costs, it shifts the burden on inference memory: expert parameters and activations consume memory, limiting the number of experts per device. As tokens are routed, some experts become overloaded while others are underutilized. Because experts are mapped to GPUs, this imbalance translates directly into degraded system performance in terms of latency, throughput, and cost. We present LASER, a plug-and-play, inference-time routing algorithm that balances load while preserving accuracy. LASER adapts to the shape of the gate's score distribution. When scores provide a clear preference, it routes to the strongest experts; when scores are more uniform, it broadens the set of viable experts and routes to the least-loaded among them. Because LASER relies only on gate scores from a trained model, it integrates directly into existing MoE inference pipelines without retraining or finetuning. We evaluate LASER on Mixtral-8x7B and DeepSeek-MoE-16b-chat across four datasets (ARC-Easy, ARC-Challenge, MMLU, and GSM8K). LASER improves load balancing, translating into lower latency and higher throughput, while keeping the accuracy changes negligible.

CAFL-L: Constraint-Aware Federated Learning with Lagrangian Dual Optimization for On-Device Language Models

arXiv:2510.03298v1 Announce Type: new Abstract: We introduce Constraint-Aware Federated Learning with Lagrangian Dual Optimization (CAFL-L), a principled extension of FedAvg that explicitly incorporates device-level resource constraints including energy, communication, memory, and thermal budgets. CAFL-L employs Lagrangian dual optimization to dynamically adapt training hyperparameters -- freezing depth, local steps, batch size, and communication compression -- while preserving training stability through token-budget preservation via gradient accumulation. Experiments on a character-level language model demonstrate that CAFL-L achieves superior constraint satisfaction compared to standard FedAvg (reducing memory usage by 20% and communication by 95%) while maintaining competitive validation performance, making it practical for deployment on resource-constrained edge devices.

Dynamic Meta-Learning for Adaptive XGBoost-Neural Ensembles

arXiv:2510.03301v1 Announce Type: new Abstract: This paper introduces a novel adaptive ensemble framework that synergistically combines XGBoost and neural networks through sophisticated meta-learning. The proposed method leverages advanced uncertainty quantification techniques and feature importance integration to dynamically orchestrate model selection and combination. Experimental results demonstrate superior predictive performance and enhanced interpretability across diverse datasets, contributing to the development of more intelligent and flexible machine learning systems.

Revoking Amnesia: RL-based Trajectory Optimization to Resurrect Erased Concepts in Diffusion Models

arXiv:2510.03302v1 Announce Type: new Abstract: Concept erasure techniques have been widely deployed in T2I diffusion models to prevent inappropriate content generation for safety and copyright considerations. However, as models evolve to next-generation architectures like Flux, established erasure methods (\textit{e.g.}, ESD, UCE, AC) exhibit degraded effectiveness, raising questions about their true mechanisms. Through systematic analysis, we reveal that concept erasure creates only an illusion of ``amnesia": rather than genuine forgetting, these methods bias sampling trajectories away from target concepts, making the erasure fundamentally reversible. This insight motivates the need to distinguish superficial safety from genuine concept removal. In this work, we propose \textbf{RevAm} (\underline{Rev}oking \underline{Am}nesia), an RL-based trajectory optimization framework that resurrects erased concepts by dynamically steering the denoising process without modifying model weights. By adapting Group Relative Policy Optimization (GRPO) to diffusion models, RevAm explores diverse recovery trajectories through trajectory-level rewards, overcoming local optima that limit existing methods. Extensive experiments demonstrate that RevAm achieves superior concept resurrection fidelity while reducing computational time by 10$\times$, exposing critical vulnerabilities in current safety mechanisms and underscoring the need for more robust erasure techniques beyond trajectory manipulation.

Machine Learning Workflows in Climate Modeling: Design Patterns and Insights from Case Studies

arXiv:2510.03305v1 Announce Type: new Abstract: Machine learning has been increasingly applied in climate modeling on system emulation acceleration, data-driven parameter inference, forecasting, and knowledge discovery, addressing challenges such as physical consistency, multi-scale coupling, data sparsity, robust generalization, and integration with scientific workflows. This paper analyzes a series of case studies from applied machine learning research in climate modeling, with a focus on design choices and workflow structure. Rather than reviewing technical details, we aim to synthesize workflow design patterns across diverse projects in ML-enabled climate modeling: from surrogate modeling, ML parameterization, probabilistic programming, to simulation-based inference, and physics-informed transfer learning. We unpack how these workflows are grounded in physical knowledge, informed by simulation data, and designed to integrate observations. We aim to offer a framework for ensuring rigor in scientific machine learning through more transparent model development, critical evaluation, informed adaptation, and reproducibility, and to contribute to lowering the barrier for interdisciplinary collaboration at the interface of data science and climate modeling.

Thin Bridges for Drug Text Alignment: Lightweight Contrastive Learning for Target Specific Drug Retrieval

arXiv:2510.03309v1 Announce Type: new Abstract: Multimodal foundation models hold promise for drug discovery and biomedical applications, but most existing approaches rely on heavy pretraining or large scale multimodal corpora. We investigate whether thin contrastive bridges, lightweight projection heads over frozen unimodal encoders can align chemical and textual representations without training a full multimodal model. Using paired mechanisms from ChEMBL, we align ECFP4 molecular fingerprints with biomedical sentence embeddings through dual linear projections trained with a contrastive objective. To better handle drugs sharing the same therapeutic target, we incorporate hard negative weighting and a margin loss. Evaluation under scaffold based splits, which require generalization across disjoint chemical cores, demonstrates that our approach achieves non-trivial cross modal alignment and substantially improves within target discrimination compared to frozen baselines. These results suggest that thin bridges offer a compute efficient alternative to large scale multimodal pretraining, enabling scaffold aware drug text alignment and target specific retrieval in precision medicine.

Predicting Effects, Missing Distributions: Evaluating LLMs as Human Behavior Simulators in Operations Management

arXiv:2510.03310v1 Announce Type: new Abstract: LLMs are emerging tools for simulating human behavior in business, economics, and social science, offering a lower-cost complement to laboratory experiments, field studies, and surveys. This paper evaluates how well LLMs replicate human behavior in operations management. Using nine published experiments in behavioral operations, we assess two criteria: replication of hypothesis-test outcomes and distributional alignment via Wasserstein distance. LLMs reproduce most hypothesis-level effects, capturing key decision biases, but their response distributions diverge from human data, including for strong commercial models. We also test two lightweight interventions -- chain-of-thought prompting and hyperparameter tuning -- which reduce misalignment and can sometimes let smaller or open-source models match or surpass larger systems.

Scaling Laws Revisited: Modeling the Role of Data Quality in Language Model Pretraining

arXiv:2510.03313v1 Announce Type: new Abstract: Scaling laws for language model training traditionally characterize how performance scales with model size and dataset volume. Prior work has explored architecture variants and data treatments such as dataset filtering and noise injection in language model pretraining; however, these studies have not formalized data quality within a principled scaling law. We introduce a dimensionless data-quality parameter Q, and propose a quality-aware scaling law extending the Chinchilla framework to predict loss as a joint function of model size, data volume, and data quality. The law is motivated by an effective-sample-size and information-theoretic view of noisy or redundant corpora, and it admits two practical estimators for Q: (i) a corruption rate proxy and (ii) a deficiency measure. Through synthetic experiments in neural machine translation and autoregressive modeling -- where we systematically control data quality via multiple levels of noise injection and coverage variation -- we show that loss scales predictably with data quality and that higher-quality data can substantially reduce model size and hence compute requirements. Our results demonstrate a sublinear decay of effective data with quality and robustness to moderate data corruption; out-of-sample evaluations further validate the predictive form of the law. Unlike prior empirical analyses, our work establishes an explicit, generalizable law for data quality, offering concrete guidance for balancing data curation effort and model scale in large-scale pretraining.

Fast frequency reconstruction using Deep Learning for event recognition in ring laser data

arXiv:2510.03325v1 Announce Type: new Abstract: The reconstruction of a frequency with minimal delay from a sinusoidal signal is a common task in several fields; for example Ring Laser Gyroscopes, since their output signal is a beat frequency. While conventional methods require several seconds of data, we present a neural network approach capable of reconstructing frequencies of several hundred Hertz within approximately 10 milliseconds. This enables rapid trigger generation. The method outperforms standard Fourier-based techniques, improving frequency estimation precision by a factor of 2 in the operational range of GINGERINO, our Ring Laser Gyroscope.\ In addition to fast frequency estimation, we introduce an automated classification framework to identify physical disturbances in the signal, such as laser instabilities and seismic events, achieving accuracy rates between 99\% and 100\% on independent test datasets for the seismic class. These results mark a step forward in integrating artificial intelligence into signal analysis for geophysical applications.

Constant in an Ever-Changing World

arXiv:2510.03330v1 Announce Type: new Abstract: The training process of reinforcement learning often suffers from severe oscillations, leading to instability and degraded performance. In this paper, we propose a Constant in an Ever-Changing World (CIC) framework that enhances algorithmic stability to improve performance. CIC maintains both a representative policy and a current policy. Instead of updating the representative policy blindly, CIC selectively updates it only when the current policy demonstrates superiority. Furthermore, CIC employs an adaptive adjustment mechanism, enabling the representative and current policies to jointly facilitate critic training. We evaluate CIC on five MuJoCo environments, and the results show that CIC improves the performance of conventional algorithms without incurring additional computational cost.

Semantic-Aware Scheduling for GPU Clusters with Large Language Models

arXiv:2510.03334v1 Announce Type: new Abstract: Deep learning (DL) schedulers are pivotal in optimizing resource allocation in GPU clusters, but operate with a critical limitation: they are largely blind to the semantic context of the jobs they manage. This forces them to rely on limited metadata, leading to high profiling overhead, unreliable duration estimation, inadequate failure handling, and poor observability. To this end, we propose SchedMate, a framework that bridges this semantic gap by systematically extracting deep insights from overlooked, unstructured data sources: source code, runtime logs, and historical jobs. SchedMate enhances existing schedulers non-intrusively through three LLM-based components. Our implementation integrates seamlessly with existing deep learning schedulers. Evaluations on a 128-GPU physical cluster and extensive simulations on production traces show SchedMate reduces average job completion times by up to 1.91x, substantially enhancing the scheduling performance, demonstrating the critical role of semantic-awareness in modern DL scheduling.

Matching the Optimal Denoiser in Point Cloud Diffusion with (Improved) Rotational Alignment

arXiv:2510.03335v1 Announce Type: new Abstract: Diffusion models are a popular class of generative models trained to reverse a noising process starting from a target data distribution. Training a diffusion model consists of learning how to denoise noisy samples at different noise levels. When training diffusion models for point clouds such as molecules and proteins, there is often no canonical orientation that can be assigned. To capture this symmetry, the true data samples are often augmented by transforming them with random rotations sampled uniformly over $SO(3)$. Then, the denoised predictions are often rotationally aligned via the Kabsch-Umeyama algorithm to the ground truth samples before computing the loss. However, the effect of this alignment step has not been well studied. Here, we show that the optimal denoiser can be expressed in terms of a matrix Fisher distribution over $SO(3)$. Alignment corresponds to sampling the mode of this distribution, and turns out to be the zeroth order approximation for small noise levels, explaining its effectiveness. We build on this perspective to derive better approximators to the optimal denoiser in the limit of small noise. Our experiments highlight that alignment is often a `good enough' approximation for the noise levels that matter most for training diffusion models.

Pool Me Wisely: On the Effect of Pooling in Transformer-Based Models

arXiv:2510.03339v1 Announce Type: new Abstract: Transformer models have become the dominant backbone for sequence modeling, leveraging self-attention to produce contextualized token representations. These are typically aggregated into fixed-size vectors via pooling operations for downstream tasks. While much of the literature has focused on attention mechanisms, the role of pooling remains underexplored despite its critical impact on model behavior. In this paper, we introduce a theoretical framework that rigorously characterizes the expressivity of Transformer-based models equipped with widely used pooling methods by deriving closed-form bounds on their representational capacity and the ability to distinguish similar inputs. Our analysis extends to different variations of attention formulations, demonstrating that these bounds hold across diverse architectural variants. We empirically evaluate pooling strategies across tasks requiring both global and local contextual understanding, spanning three major modalities: computer vision, natural language processing, and time-series analysis. Results reveal consistent trends in how pooling choices affect accuracy, sensitivity, and optimization behavior. Our findings unify theoretical and empirical perspectives, providing practical guidance for selecting or designing pooling mechanisms suited to specific tasks. This work positions pooling as a key architectural component in Transformer models and lays the foundation for more principled model design beyond attention alone.

Learning Pareto-Optimal Pandemic Intervention Policies with MORL

arXiv:2510.03340v1 Announce Type: new Abstract: The COVID-19 pandemic underscored a critical need for intervention strategies that balance disease containment with socioeconomic stability. We approach this challenge by designing a framework for modeling and evaluating disease-spread prevention strategies. Our framework leverages multi-objective reinforcement learning (MORL) - a formulation necessitated by competing objectives - combined with a new stochastic differential equation (SDE) pandemic simulator, calibrated and validated against global COVID-19 data. Our simulator reproduces national-scale pandemic dynamics with orders of magnitude higher fidelity than other models commonly used in reinforcement learning (RL) approaches to pandemic intervention. Training a Pareto-Conditioned Network (PCN) agent on this simulator, we illustrate the direct policy trade-offs between epidemiological control and economic stability for COVID-19. Furthermore, we demonstrate the framework's generality by extending it to pathogens with different epidemiological profiles, such as polio and influenza, and show how these profiles lead the agent to discover fundamentally different intervention policies. To ground our work in contemporary policymaking challenges, we apply the model to measles outbreaks, quantifying how a modest 5% drop in vaccination coverage necessitates significantly more stringent and costly interventions to curb disease spread. This work provides a robust and adaptable framework to support transparent, evidence-based policymaking for mitigating public health crises.

Pilot selection in the era of Virtual reality: algorithms for accurate and interpretable machine learning models

arXiv:2510.03345v1 Announce Type: new Abstract: With the rapid growth of the aviation industry, there is a need for a large number of flight crew. How to select the right pilots in a cost-efficient manner has become an important research question. In the current study, twenty-three pilots were recruited from China Eastern Airlines, and 23 novices were from the community of Tsinghua University. A novel approach incorporating machine learning and virtual reality technology was applied to distinguish features between these participants with different flight skills. Results indicate that SVM with the MIC feature selection method consistently achieved the highest prediction performance on all metrics with an Accuracy of 0.93, an AUC of 0.96, and an F1 of 0.93, which outperforms four other classifier algorithms and two other feature selection methods. From the perspective of feature selection methods, the MIC method can select features with a nonlinear relationship to sampling labels, instead of a simple filter-out. Our new implementation of the SVM + MIC algorithm outperforms all existing pilot selection algorithms and perhaps provides the first implementation based on eye tracking and flight dynamics data. This study's VR simulation platforms and algorithms can be used for pilot selection and training.

KVComm: Enabling Efficient LLM Communication through Selective KV Sharing

arXiv:2510.03346v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in multi-agent systems, where effective inter-model communication is crucial. Existing communication protocols either rely on natural language, incurring high inference costs and information loss, or on hidden states, which suffer from information concentration bias and inefficiency. To address these limitations, we propose KVComm, a novel communication framework that enables efficient communication between LLMs through selective sharing of KV pairs. KVComm leverages the rich information encoded in the KV pairs while avoiding the pitfalls of hidden states. We introduce a KV layer-wise selection strategy based on attention importance scores with a Gaussian prior to identify the most informative KV pairs for communication. Extensive experiments across diverse tasks and model pairs demonstrate that KVComm achieves comparable performance to the upper-bound method, which directly merges inputs to one model without any communication, while transmitting as few as 30\% of layers' KV pairs. Our study highlights the potential of KV pairs as an effective medium for inter-LLM communication, paving the way for scalable and efficient multi-agent systems.

AgentCaster: Reasoning-Guided Tornado Forecasting

arXiv:2510.03349v1 Announce Type: new Abstract: There is a growing need to evaluate Large Language Models (LLMs) on complex, high-impact, real-world tasks to assess their true readiness as reasoning agents. To address this gap, we introduce AgentCaster, a contamination-free framework employing multimodal LLMs end-to-end for the challenging, long-horizon task of tornado forecasting. Within AgentCaster, models interpret heterogeneous spatiotemporal data from a high-resolution convection-allowing forecast archive. We assess model performance over a 40-day period featuring diverse historical data, spanning several major tornado outbreaks and including over 500 tornado reports. Each day, models query interactively from a pool of 3,625 forecast maps and 40,125 forecast soundings for a forecast horizon of 12-36 hours. Probabilistic tornado-risk polygon predictions are verified against ground truths derived from geometric comparisons across disjoint risk bands in projected coordinate space. To quantify accuracy, we propose domain-specific TornadoBench and TornadoHallucination metrics, with TornadoBench highly challenging for both LLMs and domain expert human forecasters. Notably, human experts significantly outperform state-of-the-art models, which demonstrate a strong tendency to hallucinate and overpredict risk intensity, struggle with precise geographic placement, and exhibit poor spatiotemporal reasoning in complex, dynamically evolving systems. AgentCaster aims to advance research on improving LLM agents for challenging reasoning tasks in critical domains.

Interpretable Neuropsychiatric Diagnosis via Concept-Guided Graph Neural Networks

arXiv:2510.03351v1 Announce Type: new Abstract: Nearly one in five adolescents currently live with a diagnosed mental or behavioral health condition, such as anxiety, depression, or conduct disorder, underscoring the urgency of developing accurate and interpretable diagnostic tools. Resting-state functional magnetic resonance imaging (rs-fMRI) provides a powerful lens into large-scale functional connectivity, where brain regions are modeled as nodes and inter-regional synchrony as edges, offering clinically relevant biomarkers for psychiatric disorders. While prior works use graph neural network (GNN) approaches for disorder prediction, they remain complex black-boxes, limiting their reliability and clinical translation. In this work, we propose CONCEPTNEURO, a concept-based diagnosis framework that leverages large language models (LLMs) and neurobiological domain knowledge to automatically generate, filter, and encode interpretable functional connectivity concepts. Each concept is represented as a structured subgraph linking specific brain regions, which are then passed through a concept classifier. Our design ensures predictions through clinically meaningful connectivity patterns, enabling both interpretability and strong predictive performance. Extensive experiments across multiple psychiatric disorder datasets demonstrate that CONCEPTNEURO-augmented GNNs consistently outperform their vanilla counterparts, improving accuracy while providing transparent, clinically aligned explanations. Furthermore, concept analyses highlight disorder-specific connectivity patterns that align with expert knowledge and suggest new hypotheses for future investigation, establishing CONCEPTNEURO as an interpretable, domain-informed framework for psychiatric disorder diagnosis.

High Cycle S-N curve prediction for Al 7075-T6 alloy using Recurrent Neural Networks (RNNs)

arXiv:2510.03355v1 Announce Type: new Abstract: Aluminum is a widely used alloy, which is susceptible to fatigue failure. Characterizing fatigue performance for materials is extremely time and cost demanding, especially for high cycle data. To help mitigate this, a transfer learning based framework has been developed using Long short-term memory networks (LSTMs) in which a source LSTM model is trained based on pure axial fatigue data for Aluminum 7075-T6 alloy which is then transferred to predict high cycle torsional S-N curves. The framework was able to accurately predict Al torsional S-N curves for a much higher cycle range. It is the belief that this framework will help to drastically mitigate the cost of gathering fatigue characteristics for different materials and help prioritize tests with better cost and time constraints.

Understanding Transformers for Time Series: Rank Structure, Flow-of-ranks, and Compressibility

arXiv:2510.03358v1 Announce Type: new Abstract: Transformers are widely used across data modalities, and yet the principles distilled from text models often transfer imperfectly to models trained to other modalities. In this paper, we analyze Transformers through the lens of rank structure. Our focus is on the time series setting, where the structural properties of the data differ remarkably from those of text or vision. We show that time-series embeddings, unlike text or vision, exhibit sharply decaying singular value spectra: small patch sizes and smooth continuous mappings concentrate the data into low-rank subspaces. From this, we prove that the associated $Q/K/V$ projections admit accurate low-rank approximations, and that attention layers become compressible in proportion to the decay of the embedding spectrum. We introduce the concept of flow-of-ranks, a phenomenon by which nonlinear mixing across depth inflates the rank, explaining why early layers are most amenable to compression and why ranks grow with depth. Guided by these theoretical and empirical results, we use these insights to compress Chronos, a large time series foundation model, achieving a reduction of $65\%$ in inference time and $81\%$ in memory, without loss of accuracy. Our findings provide principled guidance for allocating width, depth, and heads in time series foundation models, and for exploiting their inherent compressibility.

Physics-informed Neural-operator Predictive Control for Drag Reduction in Turbulent Flows

arXiv:2510.03360v1 Announce Type: new Abstract: Assessing turbulence control effects for wall friction numerically is a significant challenge since it requires expensive simulations of turbulent fluid dynamics. We instead propose an efficient deep reinforcement learning (RL) framework for modeling and control of turbulent flows. It is model-based RL for predictive control (PC), where both the policy and the observer models for turbulence control are learned jointly using Physics Informed Neural Operators (PINO), which are discretization invariant and can capture fine scales in turbulent flows accurately. Our PINO-PC outperforms prior model-free reinforcement learning methods in various challenging scenarios where the flows are of high Reynolds numbers and unseen, i.e., not provided during model training. We find that PINO-PC achieves a drag reduction of 39.0\% under a bulk-velocity Reynolds number of 15,000, outperforming previous fluid control methods by more than 32\%.

Estimating link level traffic emissions: enhancing MOVES with open-source data

arXiv:2510.03362v1 Announce Type: new Abstract: Open-source data offers a scalable and transparent foundation for estimating vehicle activity and emissions in urban regions. In this study, we propose a data-driven framework that integrates MOVES and open-source GPS trajectory data, OpenStreetMap (OSM) road networks, regional traffic datasets and satellite imagery-derived feature vectors to estimate the link level operating mode distribution and traffic emissions. A neural network model is trained to predict the distribution of MOVES-defined operating modes using only features derived from readily available data. The proposed methodology was applied using open-source data related to 45 municipalities in the Boston Metropolitan area. The "ground truth" operating mode distribution was established using OSM open-source GPS trajectories. Compared to the MOVES baseline, the proposed model reduces RMSE by over 50% for regional scale traffic emissions of key pollutants including CO, NOx, CO2, and PM2.5. This study demonstrates the feasibility of low-cost, replicable, and data-driven emissions estimation using fully open data sources.

Diffusion-Based, Data-Assimilation-Enabled Super-Resolution of Hub-height Winds

arXiv:2510.03364v1 Announce Type: new Abstract: High-quality observations of hub-height winds are valuable but sparse in space and time. Simulations are widely available on regular grids but are generally biased and too coarse to inform wind-farm siting or to assess extreme-weather-related risks (e.g., gusts) at infrastructure scales. To fully utilize both data types for generating high-quality, high-resolution hub-height wind speeds (tens to ~100m above ground), this study introduces WindSR, a diffusion model with data assimilation for super-resolution downscaling of hub-height winds. WindSR integrates sparse observational data with simulation fields during downscaling using state-of-the-art diffusion models. A dynamic-radius blending method is introduced to merge observations with simulations, providing conditioning for the diffusion process. Terrain information is incorporated during both training and inference to account for its role as a key driver of winds. Evaluated against convolutional-neural-network and generative-adversarial-network baselines, WindSR outperforms them in both downscaling efficiency and accuracy. Our data assimilation reduces WindSR's model bias by approximately 20% relative to independent observations.

Disentangling Recall and Reasoning in Transformer Models through Layer-wise Attention and Activation Analysis

arXiv:2510.03366v1 Announce Type: new Abstract: Transformer-based language models excel at both recall (retrieving memorized facts) and reasoning (performing multi-step inference), but whether these abilities rely on distinct internal mechanisms remains unclear. Distinguishing recall from reasoning is crucial for predicting model generalization, designing targeted evaluations, and building safer interventions that affect one ability without disrupting the other.We approach this question through mechanistic interpretability, using controlled datasets of synthetic linguistic puzzles to probe transformer models at the layer, head, and neuron level. Our pipeline combines activation patching and structured ablations to causally measure component contributions to each task type. Across two model families (Qwen and LLaMA), we find that interventions on distinct layers and attention heads lead to selective impairments: disabling identified "recall circuits" reduces fact-retrieval accuracy by up to 15\% while leaving reasoning intact, whereas disabling "reasoning circuits" reduces multi-step inference by a comparable margin. At the neuron level, we observe task-specific firing patterns, though these effects are less robust, consistent with neuronal polysemanticity.Our results provide the first causal evidence that recall and reasoning rely on separable but interacting circuits in transformer models. These findings advance mechanistic interpretability by linking circuit-level structure to functional specialization and demonstrate how controlled datasets and causal interventions can yield mechanistic insights into model cognition, informing safer deployment of large language models.

Distributed Low-Communication Training with Decoupled Momentum Optimization

arXiv:2510.03371v1 Announce Type: new Abstract: The training of large models demands substantial computational resources, typically available only in data centers with high-bandwidth interconnects. However, reducing the reliance on high-bandwidth interconnects between nodes enables the use of distributed compute resources as an alternative to centralized data center training. Building on recent advances in distributed model training, we propose an approach that further reduces communication by combining infrequent synchronizations across distributed model replicas with gradient momentum compression. In particular, we treat the optimizer momentum as a signal and decompose the Nesterov momentum into high- and low-frequency components via the discrete cosine transform (DCT). Only the high-frequency components are synchronized across model replicas every $H$ steps. Empirically, our method achieves up to a $16\times$ reduction in communication compared to the baseline DiLoCo, and it generalizes across architectures, including transformer-based language models and convolutional neural networks for images. Overall, this work advances the feasibility of training large models on distributed nodes with low-bandwidth interconnects.

Conditional Pseudo-Supervised Contrast for Data-Free Knowledge Distillation

arXiv:2510.03375v1 Announce Type: new Abstract: Data-free knowledge distillation~(DFKD) is an effective manner to solve model compression and transmission restrictions while retaining privacy protection, which has attracted extensive attention in recent years. Currently, the majority of existing methods utilize a generator to synthesize images to support the distillation. Although the current methods have achieved great success, there are still many issues to be explored. Firstly, the outstanding performance of supervised learning in deep learning drives us to explore a pseudo-supervised paradigm on DFKD. Secondly, current synthesized methods cannot distinguish the distributions of different categories of samples, thus producing ambiguous samples that may lead to an incorrect evaluation by the teacher. Besides, current methods cannot optimize the category-wise diversity samples, which will hinder the student model learning from diverse samples and further achieving better performance. In this paper, to address the above limitations, we propose a novel learning paradigm, i.e., conditional pseudo-supervised contrast for data-free knowledge distillation~(CPSC-DFKD). The primary innovations of CPSC-DFKD are: (1) introducing a conditional generative adversarial network to synthesize category-specific diverse images for pseudo-supervised learning, (2) improving the modules of the generator to distinguish the distributions of different categories, and (3) proposing pseudo-supervised contrastive learning based on teacher and student views to enhance diversity. Comprehensive experiments on three commonly-used datasets validate the performance lift of both the student and generator brought by CPSC-DFKD. The code is available at https://github.com/RoryShao/CPSC-DFKD.git

A Robust Clustered Federated Learning Approach for Non-IID Data with Quantity Skew

arXiv:2510.03380v1 Announce Type: new Abstract: Federated Learning (FL) is a decentralized paradigm that enables a client-server architecture to collaboratively train a global Artificial Intelligence model without sharing raw data, thereby preserving privacy. A key challenge in FL is Non-IID data. Quantity Skew (QS) is a particular problem of Non-IID, where clients hold highly heterogeneous data volumes. Clustered Federated Learning (CFL) is an emergent variant of FL that presents a promising solution to Non-IID problem. It improves models' performance by grouping clients with similar data distributions into clusters. CFL methods generally fall into two operating strategies. In the first strategy, clients select the cluster that minimizes the local training loss. In the second strategy, the server groups clients based on local model similarities. However, most CFL methods lack systematic evaluation under QS but present significant challenges because of it. In this paper, we present two main contributions. The first one is an evaluation of state-of-the-art CFL algorithms under various Non-IID settings, applying multiple QS scenarios to assess their robustness. Our second contribution is a novel iterative CFL algorithm, named CORNFLQS, which proposes an optimal coordination between both operating strategies of CFL. Our approach is robust against the different variations of QS settings. We conducted intensive experiments on six image classification datasets, resulting in 270 Non-IID configurations. The results show that CORNFLQS achieves the highest average ranking in both accuracy and clustering quality, as well as strong robustness to QS perturbations. Overall, our approach outperforms actual CFL algorithms.

Cross-Modal Reconstruction Pretraining for Ramp Flow Prediction at Highway Interchanges

arXiv:2510.03381v1 Announce Type: new Abstract: Interchanges are crucial nodes for vehicle transfers between highways, yet the lack of real-time ramp detectors creates blind spots in traffic prediction. To address this, we propose a Spatio-Temporal Decoupled Autoencoder (STDAE), a two-stage framework that leverages cross-modal reconstruction pretraining. In the first stage, STDAE reconstructs historical ramp flows from mainline data, forcing the model to capture intrinsic spatio-temporal relations. Its decoupled architecture with parallel spatial and temporal autoencoders efficiently extracts heterogeneous features. In the prediction stage, the learned representations are integrated with models such as GWNet to enhance accuracy. Experiments on three real-world interchange datasets show that STDAE-GWNET consistently outperforms thirteen state-of-the-art baselines and achieves performance comparable to models using historical ramp data. This demonstrates its effectiveness in overcoming detector scarcity and its plug-and-play potential for diverse forecasting pipelines.

Studying the Korean Word-Chain Game with RLVR:Mitigating Reward Conflicts via Curriculum Learning

arXiv:2510.03394v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) is a promising approach for training large language models (LLMs) with stronger reasoning abilities. It has also been applied to a variety of logic puzzles. In this work, we study the Korean word-chain game using RLVR. We show that rule-derived rewards can naturally conflict, and demonstrate through experiments that a curriculum-learning scheme mitigates these conflicts. Our findings motivate further studies of puzzle tasks in diverse languages.

Training Variation of Physically-Informed Deep Learning Models

arXiv:2510.03416v1 Announce Type: new Abstract: A successful deep learning network is highly dependent not only on the training dataset, but the training algorithm used to condition the network for a given task. The loss function, dataset, and tuning of hyperparameters all play an essential role in training a network, yet there is not much discussion on the reliability or reproducibility of a training algorithm. With the rise in popularity of physics-informed loss functions, this raises the question of how reliable one's loss function is in conditioning a network to enforce a particular boundary condition. Reporting the model variation is needed to assess a loss function's ability to consistently train a network to obey a given boundary condition, and provides a fairer comparison among different methods. In this work, a Pix2Pix network predicting the stress fields of high elastic contrast composites is used as a case study. Several different loss functions enforcing stress equilibrium are implemented, with each displaying different levels of variation in convergence, accuracy, and enforcing stress equilibrium across many training sessions. Suggested practices in reporting model variation are also shared.

Multi-task neural diffusion processes for uncertainty-quantified wind power prediction

arXiv:2510.03419v1 Announce Type: new Abstract: Uncertainty-aware wind power prediction is essential for grid integration and reliable wind farm operation. We apply neural diffusion processes (NDPs)-a recent class of models that learn distributions over functions-and extend them to a multi-task NDP (MT-NDP) framework for wind power prediction. We provide the first empirical evaluation of NDPs in real supervisory control and data acquisition (SCADA) data. We introduce a task encoder within MT-NDPs to capture cross-turbine correlations and enable few-shot adaptation to unseen turbines. The proposed MT-NDP framework outperforms single-task NDPs and GPs in terms of point accuracy and calibration, particularly for wind turbines whose behaviour deviates from the fleet average. In general, NDP-based models deliver calibrated and scalable predictions suitable for operational deployment, offering sharper, yet trustworthy, predictive intervals that can support dispatch and maintenance decisions in modern wind farms.

Memory-Efficient Backpropagation for Fine-Tuning LLMs on Resource-Constrained Mobile Devices

arXiv:2510.03425v1 Announce Type: new Abstract: Fine-tuning large language models (LLMs) with backpropagation\textemdash even for a subset of parameters such as LoRA\textemdash can be much more memory-consuming than inference and is often deemed impractical for resource-constrained mobile devices. Alternative methods, such as zeroth-order optimization (ZO), can greatly reduce the memory footprint but come at the cost of significantly slower model convergence (10$\times$ to 100$\times$ more steps than backpropagation). We propose a memory-efficient implementation of backpropagation (MeBP) on mobile devices that provides better trade-off between memory usage and compute time, while converging faster and achieving better performance than the ZO baseline. We verify the effectiveness of MeBP on an iPhone 15 Pro Max and show that various LLMs, ranging from 0.5B to 4B parameters, can be fine-tuned using less than 1GB of memory. We release an example of the MeBP implementation at https://github.com/apple/ml-mebp.

Generalized Orders of Magnitude for Scalable, Parallel, High-Dynamic-Range Computation

arXiv:2510.03426v1 Announce Type: new Abstract: Many domains, from deep learning to finance, require compounding real numbers over long sequences, often leading to catastrophic numerical underflow or overflow. We introduce generalized orders of magnitude (GOOMs), a principled extension of traditional orders of magnitude that incorporates floating-point numbers as a special case, and which in practice enables stable computation over significantly larger dynamic ranges of real numbers than previously possible. We implement GOOMs, along with an efficient custom parallel prefix scan, to support native execution on parallel hardware such as GPUs. We demonstrate that our implementation of GOOMs outperforms traditional approaches with three representative experiments, all of which were previously considered impractical or impossible, and now become possible and practical: (1) compounding real matrix products far beyond standard floating-point limits; (2) estimating spectra of Lyapunov exponents in parallel, orders of magnitude faster than with previous methods, applying a novel selective-resetting method to prevent state colinearity; and (3) capturing long-range dependencies in deep recurrent neural networks with non-diagonal recurrent states, computed in parallel via a prefix scan, without requiring any form of stabilization. Our results show that our implementation of GOOMs, combined with efficient parallel scanning, offers a scalable and numerically robust alternative to conventional floating-point numbers for high-dynamic-range applications.

LHGEL: Large Heterogeneous Graph Ensemble Learning using Batch View Aggregation

arXiv:2510.03432v1 Announce Type: new Abstract: Learning from large heterogeneous graphs presents significant challenges due to the scale of networks, heterogeneity in node and edge types, variations in nodal features, and complex local neighborhood structures. This paper advocates for ensemble learning as a natural solution to this problem, whereby training multiple graph learners under distinct sampling conditions, the ensemble inherently captures different aspects of graph heterogeneity. Yet, the crux lies in combining these learners to meet global optimization objective while maintaining computational efficiency on large-scale graphs. In response, we propose LHGEL, an ensemble framework that addresses these challenges through batch sampling with three key components, namely batch view aggregation, residual attention, and diversity regularization. Specifically, batch view aggregation samples subgraphs and forms multiple graph views, while residual attention adaptively weights the contributions of these views to guide node embeddings toward informative subgraphs, thereby improving the accuracy of base learners. Diversity regularization encourages representational disparity across embedding matrices derived from different views, promoting model diversity and ensemble robustness. Our theoretical study demonstrates that residual attention mitigates gradient vanishing issues commonly faced in ensemble learning. Empirical results on five real heterogeneous networks validate that our LHGEL approach consistently outperforms its state-of-the-art competitors by substantial margin. Codes and datasets are available at https://github.com/Chrisshen12/LHGEL.

Consistent Kernel Change-Point Detection under m-Dependence for Text Segmentation

arXiv:2510.03437v1 Announce Type: new Abstract: Kernel change-point detection (KCPD) has become a widely used tool for identifying structural changes in complex data. While existing theory establishes consistency under independence assumptions, real-world sequential data such as text exhibits strong dependencies. We establish new guarantees for KCPD under $m$-dependent data: specifically, we prove consistency in the number of detected change points and weak consistency in their locations under mild additional assumptions. We perform an LLM-based simulation that generates synthetic $m$-dependent text to validate the asymptotics. To complement these results, we present the first comprehensive empirical study of KCPD for text segmentation with modern embeddings. Across diverse text datasets, KCPD with text embeddings outperforms baselines in standard text segmentation metrics. We demonstrate through a case study on Taylor Swift's tweets that KCPD not only provides strong theoretical and simulated reliability but also practical effectiveness for text segmentation tasks.

The Argument is the Explanation: Structured Argumentation for Trust in Agents

arXiv:2510.03442v1 Announce Type: new Abstract: Humans are black boxes -- we cannot observe their neural processes, yet society functions by evaluating verifiable arguments. AI explainability should follow this principle: stakeholders need verifiable reasoning chains, not mechanistic transparency. We propose using structured argumentation to provide a level of explanation and verification neither interpretability nor LLM-generated explanation is able to offer. Our pipeline achieves state-of-the-art 94.44 macro F1 on the AAEC published train/test split (5.7 points above prior work) and $0.81$ macro F1, $\sim$0.07 above previous published results with comparable data setups, for Argumentative MicroTexts relation classification, converting LLM text into argument graphs and enabling verification at each inferential step. We demonstrate this idea on multi-agent risk assessment using the Structured What-If Technique, where specialized agents collaborate transparently to carry out risk assessment otherwise achieved by humans alone. Using Bipolar Assumption-Based Argumentation, we capture support/attack relationships, thereby enabling automatic hallucination detection via fact nodes attacking arguments. We also provide a verification mechanism that enables iterative refinement through test-time feedback without retraining. For easy deployment, we provide a Docker container for the fine-tuned AMT model, and the rest of the code with the Bipolar ABA Python package on GitHub.

On residual network depth

arXiv:2510.03470v1 Announce Type: new Abstract: Deep residual architectures, such as ResNet and the Transformer, have enabled models of unprecedented depth, yet a formal understanding of why depth is so effective remains an open question. A popular intuition, following Veit et al. (2016), is that these residual networks behave like ensembles of many shallower models. Our key finding is an explicit analytical formula that verifies this ensemble perspective, proving that increasing network depth is mathematically equivalent to expanding the size of this implicit ensemble. Furthermore, our expansion reveals a hierarchical ensemble structure in which the combinatorial growth of computation paths leads to an explosion in the output signal, explaining the historical necessity of normalization layers in training deep models. This insight offers a first principles explanation for the historical dependence on normalization layers and sheds new light on a family of successful normalization-free techniques like SkipInit and Fixup. However, while these previous approaches infer scaling factors through optimizer analysis or a heuristic analogy to Batch Normalization, our work offers the first explanation derived directly from the network's inherent functional structure. Specifically, our Residual Expansion Theorem reveals that scaling each residual module provides a principled solution to taming the combinatorial explosion inherent to these architectures. We further show that this scaling acts as a capacity controls that also implicitly regularizes the model's complexity.

How to Set $\beta_1, \beta_2$ in Adam: An Online Learning Perspective

arXiv:2510.03478v1 Announce Type: new Abstract: While Adam is one of the most effective optimizer for training large-scale machine learning models, a theoretical understanding of how to optimally set its momentum factors, $\beta_1$ and $\beta_2$, remains largely incomplete. Prior works have shown that Adam can be seen as an instance of Follow-the-Regularized-Leader (FTRL), one of the most important class of algorithms in online learning. The prior analyses in these works required setting $\beta_1 = \sqrt{\beta_2}$, which does not cover the more practical cases with $\beta_1 \neq \sqrt{\beta_2}$. We derive novel, more general analyses that hold for both $\beta_1 \geq \sqrt{\beta_2}$ and $\beta_1 \leq \sqrt{\beta_2}$. In both cases, our results strictly generalize the existing bounds. Furthermore, we show that our bounds are tight in the worst case. We also prove that setting $\beta_1 = \sqrt{\beta_2}$ is optimal for an oblivious adversary, but sub-optimal for an non-oblivious adversary.

Reasoning-based Anomaly Detection Framework: A Real-time, Scalable, and Automated Approach to Anomaly Detection Across Domains

arXiv:2510.03486v1 Announce Type: new Abstract: Detecting anomalies in large, distributed systems presents several challenges. The first challenge arises from the sheer volume of data that needs to be processed. Flagging anomalies in a high-throughput environment calls for a careful consideration of both algorithm and system design. The second challenge comes from the heterogeneity of time-series datasets that leverage such a system in production. In practice, anomaly detection systems are rarely deployed for a single use case. Typically, there are several metrics to monitor, often across several domains (e.g. engineering, business and operations). A one-size-fits-all approach rarely works, so these systems need to be fine-tuned for every application - this is often done manually. The third challenge comes from the fact that determining the root-cause of anomalies in such settings is akin to finding a needle in a haystack. Identifying (in real time) a time-series dataset that is associated causally with the anomalous time-series data is a very difficult problem. In this paper, we describe a unified framework that addresses these challenges. Reasoning based Anomaly Detection Framework (RADF) is designed to perform real time anomaly detection on very large datasets. This framework employs a novel technique (mSelect) that automates the process of algorithm selection and hyper-parameter tuning for each use case. Finally, it incorporates a post-detection capability that allows for faster triaging and root-cause determination. Our extensive experiments demonstrate that RADF, powered by mSelect, surpasses state-of-the-art anomaly detection models in AUC performance for 5 out of 9 public benchmarking datasets. RADF achieved an AUC of over 0.85 for 7 out of 9 datasets, a distinction unmatched by any other state-of-the-art model.

Trajectory Data Suffices for Statistically Efficient Policy Evaluation in Finite-Horizon Offline RL with Linear $q^\pi$-Realizability and Concentrability

arXiv:2510.03494v1 Announce Type: new Abstract: We study finite-horizon offline reinforcement learning (RL) with function approximation for both policy evaluation and policy optimization. Prior work established that statistically efficient learning is impossible for either of these problems when the only assumptions are that the data has good coverage (concentrability) and the state-action value function of every policy is linearly realizable ($q^\pi$-realizability) (Foster et al., 2021). Recently, Tkachuk et al. (2024) gave a statistically efficient learner for policy optimization, if in addition the data is assumed to be given as trajectories. In this work we present a statistically efficient learner for policy evaluation under the same assumptions. Further, we show that the sample complexity of the learner used by Tkachuk et al. (2024) for policy optimization can be improved by a tighter analysis.

D2 Actor Critic: Diffusion Actor Meets Distributional Critic

arXiv:2510.03508v1 Announce Type: new Abstract: We introduce D2AC, a new model-free reinforcement learning (RL) algorithm designed to train expressive diffusion policies online effectively. At its core is a policy improvement objective that avoids the high variance of typical policy gradients and the complexity of backpropagation through time. This stable learning process is critically enabled by our second contribution: a robust distributional critic, which we design through a fusion of distributional RL and clipped double Q-learning. The resulting algorithm is highly effective, achieving state-of-the-art performance on a benchmark of eighteen hard RL tasks, including Humanoid, Dog, and Shadow Hand domains, spanning both dense-reward and goal-conditioned RL scenarios. Beyond standard benchmarks, we also evaluate a biologically motivated predator-prey task to examine the behavioral robustness and generalization capacity of our approach.

Task-Level Contrastiveness for Cross-Domain Few-Shot Learning

arXiv:2510.03509v1 Announce Type: new Abstract: Few-shot classification and meta-learning methods typically struggle to generalize across diverse domains, as most approaches focus on a single dataset, failing to transfer knowledge across various seen and unseen domains. Existing solutions often suffer from low accuracy, high computational costs, and rely on restrictive assumptions. In this paper, we introduce the notion of task-level contrastiveness, a novel approach designed to address issues of existing methods. We start by introducing simple ways to define task augmentations, and thereafter define a task-level contrastive loss that encourages unsupervised clustering of task representations. Our method is lightweight and can be easily integrated within existing few-shot/meta-learning algorithms while providing significant benefits. Crucially, it leads to improved generalization and computational efficiency without requiring prior knowledge of task domains. We demonstrate the effectiveness of our approach through different experiments on the MetaDataset benchmark, where it achieves superior performance without additional complexity.

A Lightweight Federated Learning Approach for Privacy-Preserving Botnet Detection in IoT

arXiv:2510.03513v1 Announce Type: new Abstract: The rapid growth of the Internet of Things (IoT) has expanded opportunities for innovation but also increased exposure to botnet-driven cyberattacks. Conventional detection methods often struggle with scalability, privacy, and adaptability in resource-constrained IoT environments. To address these challenges, we present a lightweight and privacy-preserving botnet detection framework based on federated learning. This approach enables distributed devices to collaboratively train models without exchanging raw data, thus maintaining user privacy while preserving detection accuracy. A communication-efficient aggregation strategy is introduced to reduce overhead, ensuring suitability for constrained IoT networks. Experiments on benchmark IoT botnet datasets demonstrate that the framework achieves high detection accuracy while substantially reducing communication costs. These findings highlight federated learning as a practical path toward scalable, secure, and privacy-aware intrusion detection for IoT ecosystems.

RAPID: An Efficient Reinforcement Learning Algorithm for Small Language Models

arXiv:2510.03515v1 Announce Type: new Abstract: Reinforcement learning (RL) has emerged as a promising strategy for finetuning small language models (SLMs) to solve targeted tasks such as math and coding. However, RL algorithms tend to be resource-intensive, taking a significant amount of time to train. We propose RAPID, a novel RL algorithm that can substantially reduce the running time of RL. Our key insight is that RL tends to be costly due to the need to perform both inference and backpropagation during training. To maximize use of computational resources, our algorithm performs inference in large batches, and then performs off-policy policy gradient updates in mini-batches. For off-policy updates, we incorporate group advantage estimation into the policy gradient algorithm, and derive an importance weighted estimator to correct for the bias arising from off-policy learning. Our experiments demonstrate that our algorithm can reduce running time by 11%-34% on three benchmarks compared to state-of-the-art RL algorithms while maintaining similar or better accuracy.

Certifiable Safe RLHF: Fixed-Penalty Constraint Optimization for Safer Language Models

arXiv:2510.03520v1 Announce Type: new Abstract: Ensuring safety is a foundational requirement for large language models (LLMs). Achieving an appropriate balance between enhancing the utility of model outputs and mitigating their potential for harm is a complex and persistent challenge. Contemporary approaches frequently formalize this problem within the framework of Constrained Markov Decision Processes (CMDPs) and employ established CMDP optimization techniques. However, these methods exhibit two notable limitations. First, their reliance on reward and cost functions renders performance highly sensitive to the underlying scoring mechanism, which must capture semantic meaning rather than being triggered by superficial keywords. Second, CMDP-based training entails tuning dual-variable, a process that is both computationally expensive and does not provide any provable safety guarantee for a fixed dual variable that can be exploitable through adversarial jailbreaks. To overcome these limitations, we introduce Certifiable Safe-RLHF (CS-RLHF) that introduces a cost model trained on a large-scale corpus to assign semantically grounded safety scores. In contrast to the lagrangian-based approach, CS-RLHF adopts a rectified penalty-based formulation. This design draws on the theory of exact penalty functions in constrained optimization, wherein constraint satisfaction is enforced directly through a suitably chosen penalty term. With an appropriately scaled penalty, feasibility of the safety constraints can be guaranteed at the optimizer, eliminating the need for dual-variable updates. Empirical evaluation demonstrates that CS-RLHF outperforms state-of-the-art LLM model responses rendering at-least 5 times efficient against nominal and jail-breaking prompts

Sequential decoder training for improved latent space dynamics identification

arXiv:2510.03535v1 Announce Type: new Abstract: Accurate numerical solutions of partial differential equations are essential in many scientific fields but often require computationally expensive solvers, motivating reduced-order models (ROMs). Latent Space Dynamics Identification (LaSDI) is a data-driven ROM framework that combines autoencoders with equation discovery to learn interpretable latent dynamics. However, enforcing latent dynamics during training can compromise reconstruction accuracy of the model for simulation data. We introduce multi-stage LaSDI (mLaSDI), a framework that improves reconstruction and prediction accuracy by sequentially learning additional decoders to correct residual errors from previous stages. Applied to the 1D-1V Vlasov equation, mLaSDI consistently outperforms standard LaSDI, achieving lower prediction errors and reduced training time across a wide range of architectures.

CrossLag: Predicting Major Dengue Outbreaks with a Domain Knowledge Informed Transformer

arXiv:2510.03566v1 Announce Type: new Abstract: A variety of models have been developed to forecast dengue cases to date. However, it remains a challenge to predict major dengue outbreaks that need timely public warnings the most. In this paper, we introduce CrossLag, an environmentally informed attention that allows for the incorporation of lagging endogenous signals behind the significant events in the exogenous data into the architecture of the transformer at low parameter counts. Outbreaks typically lag behind major changes in climate and oceanic anomalies. We use TimeXer, a recent general-purpose transformer distinguishing exogenous-endogenous inputs, as the baseline for this study. Our proposed model outperforms TimeXer by a considerable margin in detecting and predicting major outbreaks in Singapore dengue data over a 24-week prediction window.

Machine Unlearning Meets Adversarial Robustness via Constrained Interventions on LLMs

arXiv:2510.03567v1 Announce Type: new Abstract: With the increasing adoption of Large Language Models (LLMs), more customization is needed to ensure privacy-preserving and safe generation. We address this objective from two critical aspects: unlearning of sensitive information and robustness to jail-breaking attacks. We investigate various constrained optimization formulations that address both aspects in a \emph{unified manner}, by finding the smallest possible interventions on LLM weights that either make a given vocabulary set unreachable or embed the LLM with robustness to tailored attacks by shifting part of the weights to a \emph{safer} region. Beyond unifying two key properties, this approach contrasts with previous work in that it doesn't require an oracle classifier that is typically not available or represents a computational overhead. Surprisingly, we find that the simplest point-wise constraint-based intervention we propose leads to better performance than max-min interventions, while having a lower computational cost. Comparison against state-of-the-art defense methods demonstrates superior performance of the proposed approach.

Longitudinal Flow Matching for Trajectory Modeling

arXiv:2510.03569v1 Announce Type: new Abstract: Generative models for sequential data often struggle with sparsely sampled and high-dimensional trajectories, typically reducing the learning of dynamics to pairwise transitions. We propose \textit{Interpolative Multi-Marginal Flow Matching} (IMMFM), a framework that learns continuous stochastic dynamics jointly consistent with multiple observed time points. IMMFM employs a piecewise-quadratic interpolation path as a smooth target for flow matching and jointly optimizes drift and a data-driven diffusion coefficient, supported by a theoretical condition for stable learning. This design captures intrinsic stochasticity, handles irregular sparse sampling, and yields subject-specific trajectories. Experiments on synthetic benchmarks and real-world longitudinal neuroimaging datasets show that IMMFM outperforms existing methods in both forecasting accuracy and further downstream tasks.

Generalization of Graph Neural Network Models for Distribution Grid Fault Detection

arXiv:2510.03571v1 Announce Type: new Abstract: Fault detection in power distribution grids is critical for ensuring system reliability and preventing costly outages. Moreover, fault detection methodologies should remain robust to evolving grid topologies caused by factors such as reconfigurations, equipment failures, and Distributed Energy Resource (DER) integration. Current data-driven state-of-the-art methods use Recurrent Neural Networks (RNNs) for temporal modeling and Graph Neural Networks (GNNs) for spatial learning, in an RNN+GNN pipeline setting (RGNN in short). Specifically, for power system fault diagnosis, Graph Convolutional Networks (GCNs) have been adopted. Yet, various more advanced GNN architectures have been proposed and adopted in domains outside of power systems. In this paper, we set out to systematically and consistently benchmark various GNN architectures in an RNN+GNN pipeline model. Specifically, to the best of our knowledge, we are the first to (i) propose to use GraphSAGE and Graph Attention (GAT, GATv2) in an RGNN for fault diagnosis, and (ii) provide a comprehensive benchmark against earlier proposed RGNN solutions (RGCN) as well as pure RNN models (especially Gated Recurrent Unit (GRU)), particularly (iii) exploring their generalization potential for deployment in different settings than those used for training them. Our experimental results on the IEEE 123-node distribution network show that RGATv2 has superior generalization capabilities, maintaining high performance with an F1-score reduction of $\sim$12% across different topology settings. In contrast, pure RNN models largely fail, experiencing an F1-score reduction of up to $\sim$60%, while other RGNN variants also exhibit significant performance degradation, i.e., up to $\sim$25% lower F1-scores.

Efficient Test-Time Scaling for Small Vision-Language Models

arXiv:2510.03574v1 Announce Type: new Abstract: Small Vision-Language Models (VLMs) provide a computationally efficient alternative to larger models, at the cost of weaker generalization abilities and downstream task performance. These shortcomings could be addressed by test-time scaling techniques, but existing methods are typically computationally demanding, contradicting the resource-efficient design goals of small models. To address these limitations, we propose two novel and efficient test-time scaling strategies that leverage the model-internal features rather than external supervision: (i) Test-Time Augmentation (TTAug), which generates multiple augmented inputs and aggregates outputs at the token level without parameter updates, and (ii) Test-Time Adaptation (TTAdapt), which adapts model parameters during inference using consensus-based pseudolabels from TTAug. Through extensive experiments across nine benchmarks, we demonstrate consistent performance improvements while maintaining computational efficiency suitable for resource-constrained environments. The generality of our approach is demonstrated both within models at different scales and across different VLMs without additional tuning.

BEKAN: Boundary condition-guaranteed evolutionary Kolmogorov-Arnold networks with radial basis functions for solving PDE problems

arXiv:2510.03576v1 Announce Type: new Abstract: Deep learning has gained attention for solving PDEs, but the black-box nature of neural networks hinders precise enforcement of boundary conditions. To address this, we propose a boundary condition-guaranteed evolutionary Kolmogorov-Arnold Network (KAN) with radial basis functions (BEKAN). In BEKAN, we propose three distinct and combinable approaches for incorporating Dirichlet, periodic, and Neumann boundary conditions into the network. For Dirichlet problem, we use smooth and global Gaussian RBFs to construct univariate basis functions for approximating the solution and to encode boundary information at the activation level of the network. To handle periodic problems, we employ a periodic layer constructed from a set of sinusoidal functions to enforce the boundary conditions exactly. For a Neumann problem, we devise a least-squares formulation to guide the parameter evolution toward satisfying the Neumann condition. By virtue of the boundary-embedded RBFs, the periodic layer, and the evolutionary framework, we can perform accurate PDE simulations while rigorously enforcing boundary conditions. For demonstration, we conducted extensive numerical experiments on Dirichlet, Neumann, periodic, and mixed boundary value problems. The results indicate that BEKAN outperforms both multilayer perceptron (MLP) and B-splines KAN in terms of accuracy. In conclusion, the proposed approach enhances the capability of KANs in solving PDE problems while satisfying boundary conditions, thereby facilitating advancements in scientific computing and engineering applications.

Latent Mixture of Symmetries for Sample-Efficient Dynamic Learning

arXiv:2510.03578v1 Announce Type: new Abstract: Learning dynamics is essential for model-based control and Reinforcement Learning in engineering systems, such as robotics and power systems. However, limited system measurements, such as those from low-resolution sensors, demand sample-efficient learning. Symmetry provides a powerful inductive bias by characterizing equivariant relations in system states to improve sample efficiency. While recent methods attempt to discover symmetries from data, they typically assume a single global symmetry group and treat symmetry discovery and dynamic learning as separate tasks, leading to limited expressiveness and error accumulation. In this paper, we propose the Latent Mixture of Symmetries (Latent MoS), an expressive model that captures a mixture of symmetry-governed latent factors from complex dynamical measurements. Latent MoS focuses on dynamic learning while locally and provably preserving the underlying symmetric transformations. To further capture long-term equivariance, we introduce a hierarchical architecture that stacks MoS blocks. Numerical experiments in diverse physical systems demonstrate that Latent MoS outperforms state-of-the-art baselines in interpolation and extrapolation tasks while offering interpretable latent representations suitable for future geometric and safety-critical analyses.

FieldFormer: Physics-Informed Transformers for Spatio-Temporal Field Reconstruction from Sparse Sensors

arXiv:2510.03589v1 Announce Type: new Abstract: Spatio-temporal sensor data is often sparse, noisy, and irregular, and existing interpolation or learning methods struggle here because they either ignore governing PDEs or do not scale. We introduce FieldFormer, a transformer-based framework for mesh-free spatio-temporal field reconstruction that combines data-driven flexibility with physics-based structure. For each query, FieldFormer gathers a local neighborhood using a learnable velocity-scaled distance metric, enabling anisotropic adaptation to different propagation regimes. Neighborhoods are built efficiently via per-batch offset recomputation, and refined in an expectation-maximization style as the velocity scales evolve. Predictions are made by a local transformer encoder, and physics consistency is enforced through autograd-based PDE residuals and boundary-specific penalties. Across three benchmarks--a scalar anisotropic heat equation, a vector-valued shallow-water system, and a realistic advection-diffusion pollution simulation--FieldFormer consistently outperforms strong baselines by more than 40%. Our results demonstrate that FieldFormer enables accurate (RMSE$<10^{-2}$), efficient, and physically consistent field reconstruction from sparse (0.4%-2%) and noisy(10%) data.

Deep Reinforcement Learning for Multi-Agent Coordination

arXiv:2510.03592v1 Announce Type: new Abstract: We address the challenge of coordinating multiple robots in narrow and confined environments, where congestion and interference often hinder collective task performance. Drawing inspiration from insect colonies, which achieve robust coordination through stigmergy -- modifying and interpreting environmental traces -- we propose a Stigmergic Multi-Agent Deep Reinforcement Learning (S-MADRL) framework that leverages virtual pheromones to model local and social interactions, enabling decentralized emergent coordination without explicit communication. To overcome the convergence and scalability limitations of existing algorithms such as MADQN, MADDPG, and MAPPO, we leverage curriculum learning, which decomposes complex tasks into progressively harder sub-problems. Simulation results show that our framework achieves the most effective coordination of up to eight agents, where robots self-organize into asymmetric workload distributions that reduce congestion and modulate group performance. This emergent behavior, analogous to strategies observed in nature, demonstrates a scalable solution for decentralized multi-agent coordination in crowded environments with communication constraints.

MECKD: Deep Learning-Based Fall Detection in Multilayer Mobile Edge Computing With Knowledge Distillation

arXiv:2510.03601v1 Announce Type: new Abstract: The rising aging population has increased the importance of fall detection (FD) systems as an assistive technology, where deep learning techniques are widely applied to enhance accuracy. FD systems typically use edge devices (EDs) worn by individuals to collect real-time data, which are transmitted to a cloud center (CC) or processed locally. However, this architecture faces challenges such as a limited ED model size and data transmission latency to the CC. Mobile edge computing (MEC), which allows computations at MEC servers deployed between EDs and CC, has been explored to address these challenges. We propose a multilayer MEC (MLMEC) framework to balance accuracy and latency. The MLMEC splits the architecture into stations, each with a neural network model. If front-end equipment cannot detect falls reliably, data are transmitted to a station with more robust back-end computing. The knowledge distillation (KD) approach was employed to improve front-end detection accuracy by allowing high-power back-end stations to provide additional learning experiences, enhancing precision while reducing latency and processing loads. Simulation results demonstrate that the KD approach improved accuracy by 11.65% on the SisFall dataset and 2.78% on the FallAllD dataset. The MLMEC with KD also reduced the data latency rate by 54.15% on the FallAllD dataset and 46.67% on the SisFall dataset compared to the MLMEC without KD. In summary, the MLMEC FD system exhibits improved accuracy and reduced latency.

Deep Domain Adaptation for Turbofan Engine Remaining Useful Life Prediction: Methodologies, Evaluation and Future Trends

arXiv:2510.03604v1 Announce Type: new Abstract: Remaining Useful Life (RUL) prediction for turbofan engines plays a vital role in predictive maintenance, ensuring operational safety and efficiency in aviation. Although data-driven approaches using machine learning and deep learning have shown potential, they face challenges such as limited data and distribution shifts caused by varying operating conditions. Domain Adaptation (DA) has emerged as a promising solution, enabling knowledge transfer from source domains with abundant data to target domains with scarce data while mitigating distributional shifts. Given the unique properties of turbofan engines, such as complex operating conditions, high-dimensional sensor data, and slower-changing signals, it is essential to conduct a focused review of DA techniques specifically tailored to turbofan engines. To address this need, this paper provides a comprehensive review of DA solutions for turbofan engine RUL prediction, analyzing key methodologies, challenges, and recent advancements. A novel taxonomy tailored to turbofan engines is introduced, organizing approaches into methodology-based (how DA is applied), alignment-based (where distributional shifts occur due to operational variations), and problem-based (why certain adaptations are needed to address specific challenges). This taxonomy offers a multidimensional view that goes beyond traditional classifications by accounting for the distinctive characteristics of turbofan engine data and the standard process of applying DA techniques to this area. Additionally, we evaluate selected DA techniques on turbofan engine datasets, providing practical insights for practitioners and identifying key challenges. Future research directions are identified to guide the development of more effective DA techniques, advancing the state of RUL prediction for turbofan engines.

Explore the Loss space with Hill-ADAM

arXiv:2510.03613v1 Announce Type: new Abstract: This paper introduces Hill-ADAM. Hill-ADAM is an optimizer with its focus towards escaping local minima in prescribed loss landscapes to find the global minimum. Hill-ADAM escapes minima by deterministically exploring the state space. This eliminates uncertainty from random gradient updates in stochastic algorithms while seldom converging at the first minimum that visits. In the paper we first derive an analytical approximation of the ADAM Optimizer step size at a particular model state. From there define the primary condition determining ADAM limitations in escaping local minima. The proposed optimizer algorithm Hill-ADAM alternates between error minimization and maximization. It maximizes to escape the local minimum and minimizes again afterward. This alternation provides an overall exploration throughout the loss space. This allows the deduction of the global minimum's state. Hill-ADAM was tested with 5 loss functions and 12 amber-saturated to cooler-shade image color correction instances.

Neural Bayesian Filtering

arXiv:2510.03614v1 Announce Type: new Abstract: We present Neural Bayesian Filtering (NBF), an algorithm for maintaining distributions over hidden states, called beliefs, in partially observable systems. NBF is trained to find a good latent representation of the beliefs induced by a task. It maps beliefs to fixed-length embedding vectors, which condition generative models for sampling. During filtering, particle-style updates compute posteriors in this embedding space using incoming observations and the environment's dynamics. NBF combines the computational efficiency of classical filters with the expressiveness of deep generative models - tracking rapidly shifting, multimodal beliefs while mitigating the risk of particle impoverishment. We validate NBF in state estimation tasks in three partially observable environments.

Predicting Stock Price Movement with LLM-Enhanced Tweet Emotion Analysis

arXiv:2510.03633v1 Announce Type: new Abstract: Accurately predicting short-term stock price movement remains a challenging task due to the market's inherent volatility and sensitivity to investor sentiment. This paper discusses a deep learning framework that integrates emotion features extracted from tweet data with historical stock price information to forecast significant price changes on the following day. We utilize Meta's Llama 3.1-8B-Instruct model to preprocess tweet data, thereby enhancing the quality of emotion features derived from three emotion analysis approaches: a transformer-based DistilRoBERTa classifier from the Hugging Face library and two lexicon-based methods using National Research Council Canada (NRC) resources. These features are combined with previous-day stock price data to train a Long Short-Term Memory (LSTM) model. Experimental results on TSLA, AAPL, and AMZN stocks show that all three emotion analysis methods improve the average accuracy for predicting significant price movements, compared to the baseline model using only historical stock prices, which yields an accuracy of 13.5%. The DistilRoBERTa-based stock prediction model achieves the best performance, with accuracy rising from 23.6% to 38.5% when using LLaMA-enhanced emotion analysis. These results demonstrate that using large language models to preprocess tweet content enhances the effectiveness of emotion analysis which in turn improves the accuracy of predicting significant stock price movements.

From Theory to Practice: Evaluating Data Poisoning Attacks and Defenses in In-Context Learning on Social Media Health Discourse

arXiv:2510.03636v1 Announce Type: new Abstract: This study explored how in-context learning (ICL) in large language models can be disrupted by data poisoning attacks in the setting of public health sentiment analysis. Using tweets of Human Metapneumovirus (HMPV), small adversarial perturbations such as synonym replacement, negation insertion, and randomized perturbation were introduced into the support examples. Even these minor manipulations caused major disruptions, with sentiment labels flipping in up to 67% of cases. To address this, a Spectral Signature Defense was applied, which filtered out poisoned examples while keeping the data's meaning and sentiment intact. After defense, ICL accuracy remained steady at around 46.7%, and logistic regression validation reached 100% accuracy, showing that the defense successfully preserved the dataset's integrity. Overall, the findings extend prior theoretical studies of ICL poisoning to a practical, high-stakes setting in public health discourse analysis, highlighting both the risks and potential defenses for robust LLM deployment. This study also highlights the fragility of ICL under attack and the value of spectral defenses in making AI systems more reliable for health-related social media monitoring.

Implicit Models: Expressive Power Scales with Test-Time Compute

arXiv:2510.03638v1 Announce Type: new Abstract: Implicit models, an emerging model class, compute outputs by iterating a single parameter block to a fixed point. This architecture realizes an infinite-depth, weight-tied network that trains with constant memory, significantly reducing memory needs for the same level of performance compared to explicit models. While it is empirically known that these compact models can often match or even exceed larger explicit networks by allocating more test-time compute, the underlying mechanism remains poorly understood. We study this gap through a nonparametric analysis of expressive power. We provide a strict mathematical characterization, showing that a simple and regular implicit operator can, through iteration, progressively express more complex mappings. We prove that for a broad class of implicit models, this process lets the model's expressive power scale with test-time compute, ultimately matching a much richer function class. The theory is validated across three domains: image reconstruction, scientific computing, and operations research, demonstrating that as test-time iterations increase, the complexity of the learned mapping rises, while the solution quality simultaneously improves and stabilizes.

In-Vivo Training for Deep Brain Stimulation

arXiv:2510.03643v1 Announce Type: new Abstract: Deep Brain Stimulation (DBS) is a highly effective treatment for Parkinson's Disease (PD). Recent research uses reinforcement learning (RL) for DBS, with RL agents modulating the stimulation frequency and amplitude. But, these models rely on biomarkers that are not measurable in patients and are only present in brain-on-chip (BoC) simulations. In this work, we present an RL-based DBS approach that adapts these stimulation parameters according to brain activity measurable in vivo. Using a TD3 based RL agent trained on a model of the basal ganglia region of the brain, we see a greater suppression of biomarkers correlated with PD severity compared to modern clinical DBS implementations. Our agent outperforms the standard clinical approaches in suppressing PD biomarkers while relying on information that can be measured in a real world environment, thereby opening up the possibility of training personalized RL agents specific to individual patient needs.

SAFA-SNN: Sparsity-Aware On-Device Few-Shot Class-Incremental Learning with Fast-Adaptive Structure of Spiking Neural Network

arXiv:2510.03648v1 Announce Type: new Abstract: Continuous learning of novel classes is crucial for edge devices to preserve data privacy and maintain reliable performance in dynamic environments. However, the scenario becomes particularly challenging when data samples are insufficient, requiring on-device few-shot class-incremental learning (FSCIL) to maintain consistent model performance. Although existing work has explored parameter-efficient FSCIL frameworks based on artificial neural networks (ANNs), their deployment is still fundamentally constrained by limited device resources. Inspired by neural mechanisms, Spiking neural networks (SNNs) process spatiotemporal information efficiently, offering lower energy consumption, greater biological plausibility, and compatibility with neuromorphic hardware than ANNs. In this work, we present an SNN-based method for On-Device FSCIL, i.e., Sparsity-Aware and Fast Adaptive SNN (SAFA-SNN). We first propose sparsity-conditioned neuronal dynamics, in which most neurons remain stable while a subset stays active, thereby mitigating catastrophic forgetting. To further cope with spike non-differentiability in gradient estimation, we employ zeroth-order optimization. Moreover, during incremental learning sessions, we enhance the discriminability of new classes through subspace projection, which alleviates overfitting to novel classes. Extensive experiments conducted on two standard benchmark datasets (CIFAR100 and Mini-ImageNet) and three neuromorphic datasets (CIFAR-10-DVS, DVS128gesture, and N-Caltech101) demonstrate that SAFA-SNN outperforms baseline methods, specifically achieving at least 4.01% improvement at the last incremental session on Mini-ImageNet and 20% lower energy cost over baseline methods with practical implementation.

LLM-Guided Evolutionary Program Synthesis for Quasi-Monte Carlo Design

arXiv:2510.03650v1 Announce Type: new Abstract: Low-discrepancy point sets and digital sequences underpin quasi-Monte Carlo (QMC) methods for high-dimensional integration. We cast two long-standing QMC design problems as program synthesis and solve them with an LLM-guided evolutionary loop that mutates and selects code under task-specific fitness: (i) constructing finite 2D/3D point sets with low star discrepancy, and (ii) choosing Sobol' direction numbers that minimize randomized QMC error on downstream integrands. Our two-phase procedure combines constructive code proposals with iterative numerical refinement. On finite sets, we rediscover known optima in small 2D cases and set new best-known 2D benchmarks for N >= 40, while matching most known 3D optima up to the proven frontier (N <= 8) and reporting improved 3D benchmarks beyond. On digital sequences, evolving Sobol' parameters yields consistent reductions in randomized quasi-Monte Carlo (rQMC) mean-squared error for several 32-dimensional option-pricing tasks relative to widely used Joe--Kuo parameters, while preserving extensibility to any sample size and compatibility with standard randomizations. Taken together, the results demonstrate that LLM-driven evolutionary program synthesis can automate the discovery of high-quality QMC constructions, recovering classical designs where they are optimal and improving them where finite-N structure matters. Data and code are available at https://github.com/hockeyguy123/openevolve-star-discrepancy.git.

Optimising Battery Energy Storage System Trading via Energy Market Operator Price Forecast

arXiv:2510.03657v1 Announce Type: new Abstract: In electricity markets around the world, the ability to anticipate price movements with precision can be the difference between profit and loss, especially for fast-acting assets like battery energy storage systems (BESS). As grid volatility increases due to renewables and market decentralisation, operators and forecasters alike face growing pressure to transform prediction into strategy. Yet while forecast data is abundant, especially in advanced markets like Australia's National Electricity Market (NEM), its practical value in driving real-world BESS trading decisions remains largely unexplored. This thesis dives into that gap. This work addresses a key research question: Can the accuracy of the Australian Energy Market Operator (AEMO) energy price forecasts be systematically leveraged to develop a reliable and profitable battery energy storage system trading algorithm? Despite the availability of AEMO price forecasts, no existing framework evaluates their reliability or incorporates them into practical BESS trading strategies. By analysing patterns in forecast accuracy based on time of day, forecast horizon, and regional variations, this project creates a novel, forecast-informed BESS trading model to optimise arbitrage financial returns. The performance of this forecast-driven algorithm is benchmarked against a basic trading algorithm with no knowledge of forecast data. The study further explores the potential of machine learning techniques to predict future energy prices by enhancing AEMO forecasts to govern a more advanced trading strategy. The research outcomes will inform future improvements in energy market trading models and promote more efficient BESS integration into market operations.

Does higher interpretability imply better utility? A Pairwise Analysis on Sparse Autoencoders

arXiv:2510.03659v1 Announce Type: new Abstract: Sparse Autoencoders (SAEs) are widely used to steer large language models (LLMs), based on the assumption that their interpretable features naturally enable effective model behavior steering. Yet, a fundamental question remains unanswered: does higher interpretability indeed imply better steering utility? To answer this question, we train 90 SAEs across three LLMs (Gemma-2-2B, Qwen-2.5-3B, Gemma-2-9B), spanning five architectures and six sparsity levels, and evaluate their interpretability and steering utility based on SAEBench (arXiv:2501.12345) and AxBench (arXiv:2502.23456) respectively, and perform a rank-agreement analysis via Kendall's rank coefficients (tau b). Our analysis reveals only a relatively weak positive association (tau b approx 0.298), indicating that interpretability is an insufficient proxy for steering performance. We conjecture the interpretability utility gap may stem from the selection of SAE features, as not all of them are equally effective for steering. To further find features that truly steer the behavior of LLMs, we propose a novel selection criterion called Delta Token Confidence, which measures how much amplifying a feature changes the next token distribution. We show that our method improves the steering performance of three LLMs by 52.52 percent compared to the current best output score based criterion (arXiv:2503.34567). Strikingly, after selecting features with high Delta Token Confidence, the correlation between interpretability and utility vanishes (tau b approx 0), and can even become negative. This further highlights the divergence between interpretability and utility for the most effective steering features.

Operationalizing Data Minimization for Privacy-Preserving LLM Prompting

arXiv:2510.03662v1 Announce Type: new Abstract: The rapid deployment of large language models (LLMs) in consumer applications has led to frequent exchanges of personal information. To obtain useful responses, users often share more than necessary, increasing privacy risks via memorization, context-based personalization, or security breaches. We present a framework to formally define and operationalize data minimization: for a given user prompt and response model, quantifying the least privacy-revealing disclosure that maintains utility, and we propose a priority-queue tree search to locate this optimal point within a privacy-ordered transformation space. We evaluated the framework on four datasets spanning open-ended conversations (ShareGPT, WildChat) and knowledge-intensive tasks with single-ground-truth answers (CaseHold, MedQA), quantifying achievable data minimization with nine LLMs as the response model. Our results demonstrate that larger frontier LLMs can tolerate stronger data minimization while maintaining task quality than smaller open-source models (85.7% redaction for GPT-5 vs. 19.3% for Qwen2.5-0.5B). By comparing with our search-derived benchmarks, we find that LLMs struggle to predict optimal data minimization directly, showing a bias toward abstraction that leads to oversharing. This suggests not just a privacy gap, but a capability gap: models may lack awareness of what information they actually need to solve a task.

Token Hidden Reward: Steering Exploration-Exploitation in Group Relative Deep Reinforcement Learning

arXiv:2510.03669v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards has significantly advanced the reasoning capabilities of large language models, yet how to explicitly steer training toward exploration or exploitation remains an open problem. We introduce Token Hidden Reward (THR), a token-level metric that quantifies each token's influence on the likelihood of correct responses under Group Relative Policy Optimization (GRPO). We find that training dynamics are dominated by a small subset of tokens with high absolute THR values. Most interestingly, tokens with positive THR strengthen confidence in correct outputs, thus favoring exploitation, while tokens with negative THR preserve probability mass for alternative outputs, enabling exploration. This insight suggests a natural intervention: a THR-guided reweighting algorithm that modulates GRPO's learning signals to explicitly bias training toward exploitation or exploration. We validate the efficacy of this algorithm on diverse math reasoning benchmarks. By amplifying tokens with positive THR value and weakening negative ones, our algorithm improves greedy-decoding accuracy, favoring exploitation. The reverse strategy yields consistent gains in Pass@K accuracy, favoring exploration. We further demonstrate that our algorithm integrates seamlessly with other RL objectives such as GSPO and generalizes across architectures including Llama. These findings establish THR as a principled and fine-grained mechanism for dynamically controlling exploration and exploitation in RL-tuned LLMs, providing new tools for targeted fine-tuning in reasoning-intensive applications.

Towards Sampling Data Structures for Tensor Products in Turnstile Streams

arXiv:2510.03678v1 Announce Type: new Abstract: This paper studies the computational challenges of large-scale attention-based models in artificial intelligence by utilizing importance sampling methods in the streaming setting. Inspired by the classical definition of the $\ell_2$ sampler and the recent progress of the attention scheme in Large Language Models (LLMs), we propose the definition of the attention sampler. Our approach significantly reduces the computational burden of traditional attention mechanisms. We analyze the effectiveness of the attention sampler from a theoretical perspective, including space and update time. Additionally, our framework exhibits scalability and broad applicability across various model architectures and domains.

Group Policy Gradient

arXiv:2510.03679v1 Announce Type: new Abstract: We introduce Group Policy Gradient (GPG), a family of critic-free policy-gradient estimators for general MDPs. Inspired by the success of GRPO's approach in Reinforcement Learning from Human Feedback (RLHF), GPG replaces a learned value function with a group-based Monte Carlo advantage estimator, removing the memory, compute, and hyperparameter costs of training a critic while preserving PPO's clipped-objective structure. We prove the consistency of the GPG estimator, analyze the bias-variance tradeoffs, and demonstrate empirically that GPG matches or outperforms PPO on standard benchmarks. GPG makes better use of parallel simulations, which, together with its critic-free design, results in more efficient use of computational resources than PPO.

From Moments to Models: Graphon Mixture-Aware Mixup and Contrastive Learning

arXiv:2510.03690v1 Announce Type: new Abstract: Real-world graph datasets often consist of mixtures of populations, where graphs are generated from multiple distinct underlying distributions. However, modern representation learning approaches, such as graph contrastive learning (GCL) and augmentation methods like Mixup, typically overlook this mixture structure. In this work, we propose a unified framework that explicitly models data as a mixture of underlying probabilistic graph generative models represented by graphons. To characterize these graphons, we leverage graph moments (motif densities) to cluster graphs arising from the same model. This enables us to disentangle the mixture components and identify their distinct generative mechanisms. This model-aware partitioning benefits two key graph learning tasks: 1) It enables a graphon-mixture-aware mixup (GMAM), a data augmentation technique that interpolates in a semantically valid space guided by the estimated graphons, instead of assuming a single graphon per class. 2) For GCL, it enables model-adaptive and principled augmentations. Additionally, by introducing a new model-aware objective, our proposed approach (termed MGCL) improves negative sampling by restricting negatives to graphs from other models. We establish a key theoretical guarantee: a novel, tighter bound showing that graphs sampled from graphons with small cut distance will have similar motif densities with high probability. Extensive experiments on benchmark datasets demonstrate strong empirical performance. In unsupervised learning, MGCL achieves state-of-the-art results, obtaining the top average rank across eight datasets. In supervised learning, GMAM consistently outperforms existing strategies, achieving new state-of-the-art accuracy in 6 out of 7 datasets.

REG: A Regularization Optimizer for Robust Training Dynamics

arXiv:2510.03691v1 Announce Type: new Abstract: Optimizers are crucial for the efficient training of Large Language Models (LLMs). While AdamW is the de facto standard, recent structure-aware optimizers like Muon have emerged, which regularize gradient updates by operating on entire weight matrices. The Muon optimizer balances the gradient updates along all the directions. However, Muon's reliance on the matrix sign function can lead to training instability, exhibits incompatibility when fine-tuning models pre-trained with AdamW. To address these limitations, we propose \textbf{REG}, a novel optimizer that replaces Muon's aggressive matrix sign operator with the Row-and-Column-Scaling (RACS) operator. Theoretically grounded in balancing a matrix, the RACS operator regularizes the update steps in a less drastic manner, making it simpler to implement and more compatible with established training dynamics. Through extensive empirical experiments on LLM training, we demonstrate that our REG optimizer not only achieves superior performance and stability over AdamW, but also maintains consistency with the AdamW training paradigm. This consistency is particularly evident during the fine-tuning stage, where REG optimizer avoids the performance degradation observed with Muon.

Balancing Interpretability and Performance in Reinforcement Learning: An Adaptive Spectral Based Linear Approach

arXiv:2510.03722v1 Announce Type: new Abstract: Reinforcement learning (RL) has been widely applied to sequential decision making, where interpretability and performance are both critical for practical adoption. Current approaches typically focus on performance and rely on post hoc explanations to account for interpretability. Different from these approaches, we focus on designing an interpretability-oriented yet performance-enhanced RL approach. Specifically, we propose a spectral based linear RL method that extends the ridge regression-based approach through a spectral filter function. The proposed method clarifies the role of regularization in controlling estimation error and further enables the design of an adaptive regularization parameter selection strategy guided by the bias-variance trade-off principle. Theoretical analysis establishes near-optimal bounds for both parameter estimation and generalization error. Extensive experiments on simulated environments and real-world datasets from Kuaishou and Taobao demonstrate that our method either outperforms or matches existing baselines in decision quality. We also conduct interpretability analyses to illustrate how the learned policies make decisions, thereby enhancing user trust. These results highlight the potential of our approach to bridge the gap between RL theory and practical decision making, providing interpretability, accuracy, and adaptability in management contexts.

Personalized federated prototype learning in mixed heterogeneous data scenarios

arXiv:2510.03726v1 Announce Type: new Abstract: Federated learning has received significant attention for its ability to simultaneously protect customer privacy and leverage distributed data from multiple devices for model training. However, conventional approaches often focus on isolated heterogeneous scenarios, resulting in skewed feature distributions or label distributions. Meanwhile, data heterogeneity is actually a key factor in improving model performance. To address this issue, we propose a new approach called PFPL in mixed heterogeneous scenarios. The method provides richer domain knowledge and unbiased convergence targets by constructing personalized, unbiased prototypes for each client. Moreover, in the local update phase, we introduce consistent regularization to align local instances with their personalized prototypes, which significantly improves the convergence of the loss function. Experimental results on Digits and Office Caltech datasets validate the effectiveness of our approach and successfully reduce the communication cost.

Optimizing Fine-Tuning through Advanced Initialization Strategies for Low-Rank Adaptation

arXiv:2510.03731v1 Announce Type: new Abstract: The rapid development of parameter-efficient fine-tuning methods has noticeably improved the efficiency of adapting large language models. Among these, LoRA has gained widespread popularity due to its strong balance of effectiveness and parameter efficiency. However, LoRA relies on initializing two low-rank matrices whose product is zero, which limits its ability to effectively activate and leverage the original model weights-creating a potential bottleneck for optimal performance. To address this limitation, we propose \textbf{IniLoRA}, a novel initialization strategy that initializes the low-rank matrices to closely approximate the original model weights. Experimental results indicate that IniLoRA achieves better performance than LoRA across a range of models and tasks. Additionally, we introduce two variants, IniLoRA-$\alpha$ and IniLoRA-$\beta$, both leveraging distinct initialization methods to enhance performance further.

Cost Efficient Fairness Audit Under Partial Feedback

arXiv:2510.03734v1 Announce Type: new Abstract: We study the problem of auditing the fairness of a given classifier under partial feedback, where true labels are available only for positively classified individuals, (e.g., loan repayment outcomes are observed only for approved applicants). We introduce a novel cost model for acquiring additional labeled data, designed to more accurately reflect real-world costs such as credit assessment, loan processing, and potential defaults. Our goal is to find optimal fairness audit algorithms that are more cost-effective than random exploration and natural baselines. In our work, we consider two audit settings: a black-box model with no assumptions on the data distribution, and a mixture model, where features and true labels follow a mixture of exponential family distributions. In the black-box setting, we propose a near-optimal auditing algorithm under mild assumptions and show that a natural baseline can be strictly suboptimal. In the mixture model setting, we design a novel algorithm that achieves significantly lower audit cost than the black-box case. Our approach leverages prior work on learning from truncated samples and maximum-a-posteriori oracles, and extends known results on spherical Gaussian mixtures to handle exponential family mixtures, which may be of independent interest. Moreover, our algorithms apply to popular fairness metrics including demographic parity, equal opportunity, and equalized odds. Empirically, we demonstrate strong performance of our algorithms on real-world fair classification datasets like Adult Income and Law School, consistently outperforming natural baselines by around 50% in terms of audit cost.

HydroFusion-LMF: Semi-Supervised Multi-Network Fusion with Large-Model Adaptation for Long-Term Daily Runoff Forecasting

arXiv:2510.03744v1 Announce Type: new Abstract: Accurate decade-scale daily runoff forecasting in small watersheds is difficult because signals blend drifting trends, multi-scale seasonal cycles, regime shifts, and sparse extremes. Prior deep models (DLinear, TimesNet, PatchTST, TiDE, Nonstationary Transformer, LSTNet, LSTM) usually target single facets and under-utilize unlabeled spans, limiting regime adaptivity. We propose HydroFusion-LMF, a unified framework that (i) performs a learnable trend-seasonal-residual decomposition to reduce non-stationarity, (ii) routes residuals through a compact heterogeneous expert set (linear refinement, frequency kernel, patch Transformer, recurrent memory, dynamically normalized attention), (iii) fuses expert outputs via a hydrologic context-aware gate conditioned on day-of-year phase, antecedent precipitation, local variance, flood indicators, and static basin attributes, and (iv) augments supervision with a semi-supervised multi-task objective (composite MSE/MAE + extreme emphasis + NSE/KGE, masked reconstruction, multi-scale contrastive alignment, augmentation consistency, variance-filtered pseudo-labeling). Optional adapter / LoRA layers inject a frozen foundation time-series encoder efficiently. On a ~10-year daily dataset HydroFusion-LMF attains MSE 1.0128 / MAE 0.5818, improving the strongest baseline (DLinear) by 10.2% / 10.3% and the mean baseline by 24.6% / 17.1%. We observe simultaneous MSE and MAE reductions relative to baselines. The framework balances interpretability (explicit components, sparse gating) with performance, advancing label-efficient hydrologic forecasting under non-stationarity.

Neural Low-Discrepancy Sequences

arXiv:2510.03745v1 Announce Type: new Abstract: Low-discrepancy points are designed to efficiently fill the space in a uniform manner. This uniformity is highly advantageous in many problems in science and engineering, including in numerical integration, computer vision, machine perception, computer graphics, machine learning, and simulation. Whereas most previous low-discrepancy constructions rely on abstract algebra and number theory, Message-Passing Monte Carlo (MPMC) was recently introduced to exploit machine learning methods for generating point sets with lower discrepancy than previously possible. However, MPMC is limited to generating point sets and cannot be extended to low-discrepancy sequences (LDS), i.e., sequences of points in which every prefix has low discrepancy, a property essential for many applications. To address this limitation, we introduce Neural Low-Discrepancy Sequences ($NeuroLDS$), the first machine learning-based framework for generating LDS. Drawing inspiration from classical LDS, we train a neural network to map indices to points such that the resulting sequences exhibit minimal discrepancy across all prefixes. To this end, we deploy a two-stage learning process: supervised approximation of classical constructions followed by unsupervised fine-tuning to minimize prefix discrepancies. We demonstrate that $NeuroLDS$ outperforms all previous LDS constructions by a significant margin with respect to discrepancy measures. Moreover, we demonstrate the effectiveness of $NeuroLDS$ across diverse applications, including numerical integration, robot motion planning, and scientific machine learning. These results highlight the promise and broad significance of Neural Low-Discrepancy Sequences. Our code can be found at https://github.com/camail-official/neuro-lds.

EvoEngineer: Mastering Automated CUDA Kernel Code Evolution with Large Language Models

arXiv:2510.03760v1 Announce Type: new Abstract: CUDA kernel optimization has become a critical bottleneck for AI performance, as deep learning training and inference efficiency directly depends on highly optimized GPU kernels. Despite the promise of Large Language Models (LLMs) for automating kernel optimization, this field suffers from a fragmented ecosystem of isolated and incomparable approaches with unclear problem formulations. Furthermore, general-purpose LLM code evolution methods cannot meet strict correctness requirements of CUDA kernel optimization. We address these fundamental challenges by first formalizing CUDA kernel optimization as a code optimization task with a clear objective, constraints, and evaluation metrics. We then establish the first systematic LLM-based code evolution framework, EvoEngineer, that provides guidance for designing and adapting optimization strategies to achieve a balance between performance and correctness. Finally, we implement a kernel optimization system based on this framework and conduct extensive experiments on 91 real-world CUDA kernels. Our results demonstrate that EvoEngineer achieves a principled balance between performance and correctness, with the highest averaged median speedup of \textbf{2.72}$\times$ over baseline CUDA kernels and a code validity rate of \textbf{69.8}\%, outperforming existing methods on both dimensions. Our method achieves a maximum speedup of \textbf{36.75}$\times$ among all operations over PyTorch kernels and delivers the highest speedup on \textbf{28} (\textbf{56.0\%}) of 50 operations that achieve over \textbf{2$\times$} acceleration.

Merge and Guide: Unifying Model Merging and Guided Decoding for Controllable Multi-Objective Generation

arXiv:2510.03782v1 Announce Type: new Abstract: Adapting to diverse user needs at test time is a key challenge in controllable multi-objective generation. Existing methods are insufficient: merging-based approaches provide indirect, suboptimal control at the parameter level, often disregarding the impacts of multiple objectives. While decoding-based guidance is more direct, it typically requires aggregating logits from multiple expert models, incurring significant space overhead and relying heavily on individual model capacity. To address these issues, we introduce Merge-And-GuidE (MAGE), a two-stage framework that leverages model merging for guided decoding. We first identify a critical compatibility problem between the guidance and base models. In Stage 1, MAGE resolves this by dynamically constructing a more robust base model, merging a series of backbone models that account for multiple objectives. In Stage 2, we merge explicit and implicit value models into a unified guidance proxy, which then steers the decoding of the base model from Stage 1. Our analysis empirically validates Linear Mode Connectivity (LMC) in value models, explores the relationship between model merging and prediction ensembling, and demonstrates the enhanced controllability afforded by our approach. Extensive experiments show that our method outperforms existing approaches, achieving superior controllability, Pareto-optimal performance, and enhanced adaptability.

Allocation of Parameters in Transformers

arXiv:2510.03784v1 Announce Type: new Abstract: Transformers have achieved remarkable successes across a wide range of applications, yet the theoretical foundation of their model efficiency remains underexplored. In this work, we investigate how the model parameters -- mainly attention heads and head dimensions -- should be allocated across layers to balance expressivity and efficiency. We first provide mathematical analysis on the role of early layers in information extraction from an approximation perspective, with a theoretical characterization on the trade-off between the number of heads and head dimension under a fixed parameter budget. In addition, we uncover and prove the \emph{saturation} behavior of softmax activations: Continuously increasing head dimensions can lead to diminishing returns in learning errors, particularly for long sequences. Supported by both theory and experiments, this saturation pattern suggests that later layers can operate more efficiently with reduced parameters. Combining these insights, we propose principled strategies for allocating attention heads and dimensions across Transformers' layers, shedding light on theoretically-grounded model efficiency of Transformer-based architectures.

Robust Batched Bandits

arXiv:2510.03798v1 Announce Type: new Abstract: The batched multi-armed bandit (MAB) problem, in which rewards are collected in batches, is crucial for applications such as clinical trials. Existing research predominantly assumes light-tailed reward distributions, yet many real-world scenarios, including clinical outcomes, exhibit heavy-tailed characteristics. This paper bridges this gap by proposing robust batched bandit algorithms designed for heavy-tailed rewards, within both finite-arm and Lipschitz-continuous settings. We reveal a surprising phenomenon: in the instance-independent regime, as well as in the Lipschitz setting, heavier-tailed rewards necessitate a smaller number of batches to achieve near-optimal regret. In stark contrast, for the instance-dependent setting, the required number of batches to attain near-optimal regret remains invariant with respect to tail heaviness.

Curriculum-Augmented GFlowNets For mRNA Sequence Generation

arXiv:2510.03811v1 Announce Type: new Abstract: Designing mRNA sequences is a major challenge in developing next-generation therapeutics, since it involves exploring a vast space of possible nucleotide combinations while optimizing sequence properties like stability, translation efficiency, and protein expression. While Generative Flow Networks are promising for this task, their training is hindered by sparse, long-horizon rewards and multi-objective trade-offs. We propose Curriculum-Augmented GFlowNets (CAGFN), which integrate curriculum learning with multi-objective GFlowNets to generate de novo mRNA sequences. CAGFN integrates a length-based curriculum that progressively adapts the maximum sequence length guiding exploration from easier to harder subproblems. We also provide a new mRNA design environment for GFlowNets which, given a target protein sequence and a combination of biological objectives, allows for the training of models that generate plausible mRNA candidates. This provides a biologically motivated setting for applying and advancing GFlowNets in therapeutic sequence design. On different mRNA design tasks, CAGFN improves Pareto performance and biological plausibility, while maintaining diversity. Moreover, CAGFN reaches higher-quality solutions faster than a GFlowNet trained with random sequence sampling (no curriculum), and enables generalization to out-of-distribution sequences.

Detecting Invariant Manifolds in ReLU-Based RNNs

arXiv:2510.03814v1 Announce Type: new Abstract: Recurrent Neural Networks (RNNs) have found widespread applications in machine learning for time series prediction and dynamical systems reconstruction, and experienced a recent renaissance with improved training algorithms and architectural designs. Understanding why and how trained RNNs produce their behavior is important for scientific and medical applications, and explainable AI more generally. An RNN's dynamical repertoire depends on the topological and geometrical properties of its state space. Stable and unstable manifolds of periodic points play a particularly important role: They dissect a dynamical system's state space into different basins of attraction, and their intersections lead to chaotic dynamics with fractal geometry. Here we introduce a novel algorithm for detecting these manifolds, with a focus on piecewise-linear RNNs (PLRNNs) employing rectified linear units (ReLUs) as their activation function. We demonstrate how the algorithm can be used to trace the boundaries between different basins of attraction, and hence to characterize multistability, a computationally important property. We further show its utility in finding so-called homoclinic points, the intersections between stable and unstable manifolds, and thus establish the existence of chaos in PLRNNs. Finally we show for an empirical example, electrophysiological recordings from a cortical neuron, how insights into the underlying dynamics could be gained through our method.

TROLL: Trust Regions improve Reinforcement Learning for Large Language Models

arXiv:2510.03817v1 Announce Type: new Abstract: On-policy Reinforcement Learning (RL) with PPO-like clip objectives has become the standard choice for reward-based fine-tuning of large language models (LLMs). Although recent work has explored improved estimators of advantages and normalization, the clipping mechanism itself has remained untouched. Originally introduced as a proxy for principled KL-based trust regions, clipping is a crude approximation that often causes unstable updates and suboptimal performance. We replace the clip objective with a novel discrete differentiable trust region projection, which provides principled token-level KL constraints. The projection operates on a sparse subset of the model's most important token logits to balance computational cost and projection effectiveness. Our approach, Trust Region Optimization for Large Language Models (TROLL), serves as a direct replacement for PPO-like clipping during training and does not alter the model's inference behavior. Across datasets, model families, and advantage-estimation methods, TROLL consistently outperforms PPO-like clipping in terms of training speed, stability, and final success rates.

Distributed Area Coverage with High Altitude Balloons Using Multi-Agent Reinforcement Learning

arXiv:2510.03823v1 Announce Type: new Abstract: High Altitude Balloons (HABs) can leverage stratospheric wind layers for limited horizontal control, enabling applications in reconnaissance, environmental monitoring, and communications networks. Existing multi-agent HAB coordination approaches use deterministic methods like Voronoi partitioning and extremum seeking control for large global constellations, which perform poorly for smaller teams and localized missions. While single-agent HAB control using reinforcement learning has been demonstrated on HABs, coordinated multi-agent reinforcement learning (MARL) has not yet been investigated. This work presents the first systematic application of multi-agent reinforcement learning (MARL) to HAB coordination for distributed area coverage. We extend our previously developed reinforcement learning simulation environment (RLHAB) to support cooperative multi-agent learning, enabling multiple agents to operate simultaneously in realistic atmospheric conditions. We adapt QMIX for HAB area coverage coordination, leveraging Centralized Training with Decentralized Execution to address atmospheric vehicle coordination challenges. Our approach employs specialized observation spaces providing individual state, environmental context, and teammate data, with hierarchical rewards prioritizing coverage while encouraging spatial distribution. We demonstrate that QMIX achieves similar performance to the theoretically optimal geometric deterministic method for distributed area coverage, validating the MARL approach and providing a foundation for more complex autonomous multi-HAB missions where deterministic methods become intractable.

Proximal Diffusion Neural Sampler

arXiv:2510.03824v1 Announce Type: new Abstract: The task of learning a diffusion-based neural sampler for drawing samples from an unnormalized target distribution can be viewed as a stochastic optimal control problem on path measures. However, the training of neural samplers can be challenging when the target distribution is multimodal with significant barriers separating the modes, potentially leading to mode collapse. We propose a framework named \textbf{Proximal Diffusion Neural Sampler (PDNS)} that addresses these challenges by tackling the stochastic optimal control problem via proximal point method on the space of path measures. PDNS decomposes the learning process into a series of simpler subproblems that create a path gradually approaching the desired distribution. This staged procedure traces a progressively refined path to the desired distribution and promotes thorough exploration across modes. For a practical and efficient realization, we instantiate each proximal step with a proximal weighted denoising cross-entropy (WDCE) objective. We demonstrate the effectiveness and robustness of PDNS through extensive experiments on both continuous and discrete sampling tasks, including challenging scenarios in molecular dynamics and statistical physics.

HOFLON: Hybrid Offline Learning and Online Optimization for Process Start-Up and Grade-Transition Control

arXiv:2510.03830v1 Announce Type: new Abstract: Start-ups and product grade-changes are critical steps in continuous-process plant operation, because any misstep immediately affects product quality and drives operational losses. These transitions have long relied on manual operation by a handful of expert operators, but the progressive retirement of that workforce is leaving plant owners without the tacit know-how needed to execute them consistently. In the absence of a process model, offline reinforcement learning (RL) promises to capture and even surpass human expertise by mining historical start-up and grade-change logs, yet standard offline RL struggles with distribution shift and value-overestimation whenever a learned policy ventures outside the data envelope. We introduce HOFLON (Hybrid Offline Learning + Online Optimization) to overcome those limitations. Offline, HOFLON learns (i) a latent data manifold that represents the feasible region spanned by past transitions and (ii) a long-horizon Q-critic that predicts the cumulative reward from state-action pairs. Online, it solves a one-step optimization problem that maximizes the Q-critic while penalizing deviations from the learned manifold and excessive rates of change in the manipulated variables. We test HOFLON on two industrial case studies: a polymerization reactor start-up and a paper-machine grade-change problem, and benchmark it against Implicit Q-Learning (IQL), a leading offline-RL algorithm. In both plants HOFLON not only surpasses IQL but also delivers, on average, better cumulative rewards than the best start-up or grade-change observed in the historical data, demonstrating its potential to automate transition operations beyond current expert capability.

Technical note on Fisher Information for Robust Federated Cross-Validation

arXiv:2510.03838v1 Announce Type: new Abstract: When training data are fragmented across batches or federated-learned across different geographic locations, trained models manifest performance degradation. That degradation partly owes to covariate shift induced by data having been fragmented across time and space and producing dissimilar empirical training distributions. Each fragment's distribution is slightly different to a hypothetical unfragmented training distribution of covariates, and to the single validation distribution. To address this problem, we propose Fisher Information for Robust fEderated validation (\textbf{FIRE}). This method accumulates fragmentation-induced covariate shift divergences from the global training distribution via an approximate Fisher information. That term, which we prove to be a more computationally-tractable estimate, is then used as a per-fragment loss penalty, enabling scalable distribution alignment. FIRE outperforms importance weighting benchmarks by $5.1\%$ at maximum and federated learning (FL) benchmarks by up to $5.3\%$ on shifted validation sets.

Technical note on Sequential Test-Time Adaptation via Martingale-Driven Fisher Prompting

arXiv:2510.03839v1 Announce Type: new Abstract: We present a theoretical framework for M-FISHER, a method for sequential distribution shift detection and stable adaptation in streaming data. For detection, we construct an exponential martingale from non-conformity scores and apply Ville's inequality to obtain time-uniform guarantees on false alarm control, ensuring statistical validity at any stopping time. Under sustained shifts, we further bound the expected detection delay as $\mathcal{O}(\log(1/\delta)/\Gamma)$, where $\Gamma$ reflects the post-shift information gain, thereby linking detection efficiency to distributional divergence. For adaptation, we show that Fisher-preconditioned updates of prompt parameters implement natural gradient descent on the distributional manifold, yielding locally optimal updates that minimize KL divergence while preserving stability and parameterization invariance. Together, these results establish M-FISHER as a principled approach for robust, anytime-valid detection and geometrically stable adaptation in sequential decision-making under covariate shift.

On Using Large Language Models to Enhance Clinically-Driven Missing Data Recovery Algorithms in Electronic Health Records

arXiv:2510.03844v1 Announce Type: new Abstract: Objective: Electronic health records (EHR) data are prone to missingness and errors. Previously, we devised an "enriched" chart review protocol where a "roadmap" of auxiliary diagnoses (anchors) was used to recover missing values in EHR data (e.g., a diagnosis of impaired glycemic control might imply that a missing hemoglobin A1c value would be considered unhealthy). Still, chart reviews are expensive and time-intensive, which limits the number of patients whose data can be reviewed. Now, we investigate the accuracy and scalability of a roadmap-driven algorithm, based on ICD-10 codes (International Classification of Diseases, 10th revision), to mimic expert chart reviews and recover missing values. Materials and Methods: In addition to the clinicians' original roadmap from our previous work, we consider new versions that were iteratively refined using large language models (LLM) in conjunction with clinical expertise to expand the list of auxiliary diagnoses. Using chart reviews for 100 patients from the EHR at an extensive learning health system, we examine algorithm performance with different roadmaps. Using the larger study of $1000$ patients, we applied the final algorithm, which used a roadmap with clinician-approved additions from the LLM. Results: The algorithm recovered as much, if not more, missing data as the expert chart reviewers, depending on the roadmap. Discussion: Clinically-driven algorithms (enhanced by LLM) can recover missing EHR data with similar accuracy to chart reviews and can feasibly be applied to large samples. Extending them to monitor other dimensions of data quality (e.g., plausability) is a promising future direction.

Unlocking Reasoning Capabilities in LLMs via Reinforcement Learning Exploration

arXiv:2510.03865v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has recently enhanced the reasoning capabilities of large language models (LLMs), particularly for mathematical problem solving. However, a fundamental limitation remains: as the sampling budget increases, the advantage of RLVR-trained models over their pretrained bases often diminishes or even vanishes, revealing a strong dependence on the base model's restricted search space. We attribute this phenomenon to the widespread use of the reverse Kullback-Leibler (KL) divergence regularizer, whose mode-seeking behavior keeps the policy trapped inside the base model's support region and hampers wider exploration. To address this issue, we propose RAPO (Rewards-Aware Policy Optimization), an algorithm to promote broader yet focused exploration. Our method (i) utilizes the forward KL penalty to replace the reverse KL penalty for out-of-distribution exploration, and (ii) reweights the reference policy to facilitate adaptive in-distribution exploration. We train Qwen2.5-3B and 7B models with RAPO on the 8K SimpleRL-Zero dataset, without supervised fine-tuning, and evaluate them on AIME2024 and AIME2025. Results show that RAPO consistently improves problem-solving performance. Notably, RAPO enables models to surpass the base model's performance ceiling and solves previously intractable problems, advancing the frontier of RLVR for challenging reasoning tasks.

On Provable Benefits of Muon in Federated Learning

arXiv:2510.03866v1 Announce Type: new Abstract: The recently introduced optimizer, Muon, has gained increasing attention due to its superior performance across a wide range of applications. However, its effectiveness in federated learning remains unexplored. To address this gap, this paper investigates the performance of Muon in the federated learning setting. Specifically, we propose a new algorithm, FedMuon, and establish its convergence rate for nonconvex problems. Our theoretical analysis reveals multiple favorable properties of FedMuon. In particular, due to its orthonormalized update direction, the learning rate of FedMuon is independent of problem-specific parameters, and, importantly, it can naturally accommodate heavy-tailed noise. The extensive experiments on a variety of neural network architectures validate the effectiveness of the proposed algorithm.

Optimal Scaling Needs Optimal Norm

arXiv:2510.03871v1 Announce Type: new Abstract: Despite recent progress in optimal hyperparameter transfer under model and dataset scaling, no unifying explanatory principle has been established. Using the Scion optimizer, we discover that joint optimal scaling across model and dataset sizes is governed by a single invariant: the operator norm of the output layer. Across models with up to 1.3B parameters trained on up to 138B tokens, the optimal learning rate/batch size pair $(\eta^{\ast}, B^{\ast})$ consistently has the same operator norm value - a phenomenon we term norm transfer. This constant norm condition is necessary but not sufficient: while for each dataset size, multiple $(\eta, B)$ reach the optimal norm, only a unique $(\eta^{\ast}, B^{\ast})$ achieves the best loss. As a sufficient condition, we provide the first measurement of $(\eta^{\ast}, B^{\ast})$ scaling with dataset size for Scion, and find that the scaling rules are consistent with those of the Adam optimizer. Tuning per-layer-group learning rates also improves model performance, with the output layer being the most sensitive and hidden layers benefiting from lower learning rates. We provide practical insights on norm-guided optimal scaling and release our Distributed Scion (Disco) implementation with logs from over two thousand runs to support research on LLM training dynamics at scale.

BONSAI: Structure-exploiting robust Bayesian optimization for networked black-box systems under uncertainty

arXiv:2510.03893v1 Announce Type: new Abstract: Optimal design under uncertainty remains a fundamental challenge in advancing reliable, next-generation process systems. Robust optimization (RO) offers a principled approach by safeguarding against worst-case scenarios across a range of uncertain parameters. However, traditional RO methods typically require known problem structure, which limits their applicability to high-fidelity simulation environments. To overcome these limitations, recent work has explored robust Bayesian optimization (RBO) as a flexible alternative that can accommodate expensive, black-box objectives. Existing RBO methods, however, generally ignore available structural information and struggle to scale to high-dimensional settings. In this work, we introduce BONSAI (Bayesian Optimization of Network Systems under uncertAInty), a new RBO framework that leverages partial structural knowledge commonly available in simulation-based models. Instead of treating the objective as a monolithic black box, BONSAI represents it as a directed graph of interconnected white- and black-box components, allowing the algorithm to utilize intermediate information within the optimization process. We further propose a scalable Thompson sampling-based acquisition function tailored to the structured RO setting, which can be efficiently optimized using gradient-based methods. We evaluate BONSAI across a diverse set of synthetic and real-world case studies, including applications in process systems engineering. Compared to existing simulation-based RO algorithms, BONSAI consistently delivers more sample-efficient and higher-quality robust solutions, highlighting its practical advantages for uncertainty-aware design in complex engineering systems.

LLM as an Algorithmist: Enhancing Anomaly Detectors via Programmatic Synthesis

arXiv:2510.03904v1 Announce Type: new Abstract: Existing anomaly detection (AD) methods for tabular data usually rely on some assumptions about anomaly patterns, leading to inconsistent performance in real-world scenarios. While Large Language Models (LLMs) show remarkable reasoning capabilities, their direct application to tabular AD is impeded by fundamental challenges, including difficulties in processing heterogeneous data and significant privacy risks. To address these limitations, we propose LLM-DAS, a novel framework that repositions the LLM from a data processor'' to analgorithmist''. Instead of being exposed to raw data, our framework leverages the LLM's ability to reason about algorithms. It analyzes a high-level description of a given detector to understand its intrinsic weaknesses and then generates detector-specific, data-agnostic Python code to synthesize ``hard-to-detect'' anomalies that exploit these vulnerabilities. This generated synthesis program, which is reusable across diverse datasets, is then instantiated to augment training data, systematically enhancing the detector's robustness by transforming the problem into a more discriminative two-class classification task. Extensive experiments on 36 TAD benchmarks show that LLM-DAS consistently boosts the performance of mainstream detectors. By bridging LLM reasoning with classic AD algorithms via programmatic synthesis, LLM-DAS offers a scalable, effective, and privacy-preserving approach to patching the logical blind spots of existing detectors.

THEMIS: Unlocking Pretrained Knowledge with Foundation Model Embeddings for Anomaly Detection in Time Series

arXiv:2510.03911v1 Announce Type: new Abstract: Time series anomaly detection forms a very crucial area in several domains but poses substantial challenges. Due to time series data possessing seasonality, trends, noise, and evolving patterns (concept drift), it becomes very difficult to set a general notion of what constitutes normal behavior. Anomalies themselves could be varied, ranging from a single outlier to contextual or collective anomalies, and are normally very rare; hence, the dataset is largely imbalanced. Additional layers of complexities arise due to the problems of increased dimensionality of modern time series, real-time detection criteria, setting up appropriate detection thresholds, and arriving at results that are interpretable. To embrace these multifaceted challenges, very strong, flexible, and interpretable approaches are required. This paper presents THEMIS, a new framework for time series anomaly detection that exploits pretrained knowledge from foundation models. THEMIS extracts embeddings from the encoder of the Chronos time series foundation model and applies outlier detection techniques like Local Outlier Factor and Spectral Decomposition on the self-similarity matrix, to spot anomalies in the data. Our experiments show that this modular method achieves SOTA results on the MSL dataset and performs quite competitively on the SMAP and SWAT$^*$ datasets. Notably, THEMIS exceeds models trained specifically for anomaly detection, presenting hyperparameter robustness and interpretability by default. This paper advocates for pretrained representations from foundation models for performing efficient and adaptable anomaly detection for time series data.

Generalized Fitted Q-Iteration with Clustered Data

arXiv:2510.03912v1 Announce Type: new Abstract: This paper focuses on reinforcement learning (RL) with clustered data, which is commonly encountered in healthcare applications. We propose a generalized fitted Q-iteration (FQI) algorithm that incorporates generalized estimating equations into policy learning to handle the intra-cluster correlations. Theoretically, we demonstrate (i) the optimalities of our Q-function and policy estimators when the correlation structure is correctly specified, and (ii) their consistencies when the structure is mis-specified. Empirically, through simulations and analyses of a mobile health dataset, we find the proposed generalized FQI achieves, on average, a half reduction in regret compared to the standard FQI.

Transductive and Learning-Augmented Online Regression

arXiv:2510.03917v1 Announce Type: new Abstract: Motivated by the predictable nature of real-life in data streams, we study online regression when the learner has access to predictions about future examples. In the extreme case, called transductive online learning, the sequence of examples is revealed to the learner before the game begins. For this setting, we fully characterize the minimax expected regret in terms of the fat-shattering dimension, establishing a separation between transductive online regression and (adversarial) online regression. Then, we generalize this setting by allowing for noisy or \emph{imperfect} predictions about future examples. Using our results for the transductive online setting, we develop an online learner whose minimax expected regret matches the worst-case regret, improves smoothly with prediction quality, and significantly outperforms the worst-case regret when future example predictions are precise, achieving performance similar to the transductive online learner. This enables learnability for previously unlearnable classes under predictable examples, aligning with the broader learning-augmented model paradigm.

On the Convergence and Size Transferability of Continuous-depth Graph Neural Networks

arXiv:2510.03923v1 Announce Type: new Abstract: Continuous-depth graph neural networks, also known as Graph Neural Differential Equations (GNDEs), combine the structural inductive bias of Graph Neural Networks (GNNs) with the continuous-depth architecture of Neural ODEs, offering a scalable and principled framework for modeling dynamics on graphs. In this paper, we present a rigorous convergence analysis of GNDEs with time-varying parameters in the infinite-node limit, providing theoretical insights into their size transferability. To this end, we introduce Graphon Neural Differential Equations (Graphon-NDEs) as the infinite-node limit of GNDEs and establish their well-posedness. Leveraging tools from graphon theory and dynamical systems, we prove the trajectory-wise convergence of GNDE solutions to Graphon-NDE solutions. Moreover, we derive explicit convergence rates under two deterministic graph sampling regimes: (1) weighted graphs sampled from smooth graphons, and (2) unweighted graphs sampled from ${0,1}$-valued (discontinuous) graphons. We further establish size transferability bounds, providing theoretical justification for the practical strategy of transferring GNDE models trained on moderate-sized graphs to larger, structurally similar graphs without retraining. Numerical experiments using synthetic and real data support our theoretical findings.

LLM Chemistry Estimation for Multi-LLM Recommendation

arXiv:2510.03930v1 Announce Type: new Abstract: Multi-LLM collaboration promises accurate, robust, and context-aware solutions, yet existing approaches rely on implicit selection and output assessment without analyzing whether collaborating models truly complement or conflict. We introduce LLM Chemistry -- a framework that measures when LLM combinations exhibit synergistic or antagonistic behaviors that shape collective performance beyond individual capabilities. We formalize the notion of chemistry among LLMs, propose algorithms that quantify it by analyzing interaction dependencies, and recommend optimal model ensembles accordingly. Our theoretical analysis shows that chemistry among collaborating LLMs is most evident under heterogeneous model profiles, with its outcome impact shaped by task type, group size, and complexity. Evaluation on classification, summarization, and program repair tasks provides initial evidence for these task-dependent effects, thereby reinforcing our theoretical results. This establishes LLM Chemistry as both a diagnostic factor in multi-LLM systems and a foundation for ensemble recommendation.

On the Empirical Power of Goodness-of-Fit Tests in Watermark Detection

arXiv:2510.03944v1 Announce Type: new Abstract: Large language models (LLMs) raise concerns about content authenticity and integrity because they can generate human-like text at scale. Text watermarks, which embed detectable statistical signals into generated text, offer a provable way to verify content origin. Many detection methods rely on pivotal statistics that are i.i.d. under human-written text, making goodness-of-fit (GoF) tests a natural tool for watermark detection. However, GoF tests remain largely underexplored in this setting. In this paper, we systematically evaluate eight GoF tests across three popular watermarking schemes, using three open-source LLMs, two datasets, various generation temperatures, and multiple post-editing methods. We find that general GoF tests can improve both the detection power and robustness of watermark detectors. Notably, we observe that text repetition, common in low-temperature settings, gives GoF tests a unique advantage not exploited by existing methods. Our results highlight that classic GoF tests are a simple yet powerful and underused tool for watermark detection in LLMs.

What Is The Performance Ceiling of My Classifier? Utilizing Category-Wise Influence Functions for Pareto Frontier Analysis

arXiv:2510.03950v1 Announce Type: new Abstract: Data-centric learning seeks to improve model performance from the perspective of data quality, and has been drawing increasing attention in the machine learning community. Among its key tools, influence functions provide a powerful framework to quantify the impact of individual training samples on model predictions, enabling practitioners to identify detrimental samples and retrain models on a cleaner dataset for improved performance. However, most existing work focuses on the question: "what data benefits the learning model?" In this paper, we take a step further and investigate a more fundamental question: "what is the performance ceiling of the learning model?" Unlike prior studies that primarily measure improvement through overall accuracy, we emphasize category-wise accuracy and aim for Pareto improvements, ensuring that every class benefits, rather than allowing tradeoffs where some classes improve at the expense of others. To address this challenge, we propose category-wise influence functions and introduce an influence vector that quantifies the impact of each training sample across all categories. Leveraging these influence vectors, we develop a principled criterion to determine whether a model can still be improved, and further design a linear programming-based sample reweighting framework to achieve Pareto performance improvements. Through extensive experiments on synthetic datasets, vision, and text benchmarks, we demonstrate the effectiveness of our approach in estimating and achieving a model's performance improvement across multiple categories of interest.

Optimizing Resources for On-the-Fly Label Estimation with Multiple Unknown Medical Experts

arXiv:2510.03954v1 Announce Type: new Abstract: Accurate ground truth estimation in medical screening programs often relies on coalitions of experts and peer second opinions. Algorithms that efficiently aggregate noisy annotations can enhance screening workflows, particularly when data arrive continuously and expert proficiency is initially unknown. However, existing algorithms do not meet the requirements for seamless integration into screening pipelines. We therefore propose an adaptive approach for real-time annotation that (I) supports on-the-fly labeling of incoming data, (II) operates without prior knowledge of medical experts or pre-labeled data, and (III) dynamically queries additional experts based on the latent difficulty of each instance. The method incrementally gathers expert opinions until a confidence threshold is met, providing accurate labels with reduced annotation overhead. We evaluate our approach on three multi-annotator classification datasets across different modalities. Results show that our adaptive querying strategy reduces the number of expert queries by up to 50% while achieving accuracy comparable to a non-adaptive baseline. Our code is available at https://github.com/tbary/MEDICS

Early-Warning of Thunderstorm-Driven Power Outages with a Two-Stage Machine Learning Model

arXiv:2510.03959v1 Announce Type: new Abstract: Thunderstorm-driven outages are difficult to predict because most storms do not cause damage, convective processes occur rapidly and chaotically, and the available public data are both noisy and incomplete. We develop a 24-48 h early-warning model for summer, thunderstorm-related outages in Michigan using only open sources (EAGLE-I for ground truth; METAR for weather). We use the publicly released EAGLE-I outage dataset (2014-2022), maintained by Oak Ridge National Laboratory for the U.S. Department of Energy. The pipeline preserves convective micro-signals from a sparse station network via parameter-specific kriging with hourly variograms and targeted overdrafting to retain extremes, and builds causal spatio-temporal features (lags/rolling statistics; k-NN/IDW spatial aggregates) capturing precursors of severe convection (moisture advection, wind shifts, and pressure drops). The two-stage model design, combining a logistic gate and an LSTM regressor, limits routine periods and reduces noise exposure. The study uses event-centric metrics (cluster-based hits/misses/false alarms) and peak-conditional MASE (cMASE) in +/-Delta-hour windows around state-level peaks (>= 50,000), with uncertainty quantified by hourly moving-block bootstrap. On the test sample, Two-Stage detects more reference peaks across all windows (e.g., at +/-48 h it records 3/4 vs. 2/4; F1 66.7% vs. 57.1%) with one extra false alarm. Near peaks, it shows modest amplitude gains (2-3% lower cMASE at +/-0-12 h; bootstrap medians +9-13% at +/-6-12 h) but small losses at +/-36-48 h (~3-4%). Overall, errors are comparable to the one-step LSTM baseline. SHAP analysis confirms moisture-advection and wind/gust precursors, underscoring the value of the feature engineering. Despite open-data noise, the feature-driven pipeline yields actionable, event-focused early warnings for thunderstorm outages.

SPEAR: Soft Prompt Enhanced Anomaly Recognition for Time Series Data

arXiv:2510.03962v1 Announce Type: new Abstract: Time series anomaly detection plays a crucial role in a wide range of fields, such as healthcare and internet traffic monitoring. The emergence of large language models (LLMs) offers new opportunities for detecting anomalies in the ubiquitous time series data. Traditional approaches struggle with variable-length time series sequences and context-based anomalies. We propose Soft Prompt Enhanced Anomaly Recognition (SPEAR), a novel approach to leverage LLMs for anomaly detection with soft prompts and quantization. Our methodology involves quantizing and transforming the time series data into input embeddings and combining them with learnable soft prompt embeddings. These combined embeddings are then fed into a frozen LLM. The soft prompts are updated iteratively based on a cross-entropy loss, allowing the model to adapt to time series anomaly detection. The use of soft prompts helps adapt LLMs effectively to time series tasks, while quantization ensures optimal handling of sequences, as LLMs are designed to handle discrete sequences. Our experimental results demonstrate that soft prompts effectively increase LLMs' performance in downstream tasks regarding time series anomaly detection.

What Can You Do When You Have Zero Rewards During RL?

arXiv:2510.03971v1 Announce Type: new Abstract: Reinforcement learning (RL) with outcome-based rewards has proven effective for improving large language models (LLMs) on complex reasoning tasks. However, its success often depends on the base model occasionally sampling correct solutions. When no correct solutions are sampled, training encounters a zero-reward barrier where learning stalls due to zero gradients. We study this scenario through the graph search task introduced in Bachmann et al. (2024) and evaluate recent methods that incorporate desirable components such as dense rewards, diversity incentives, and improved credit assignment. Our experiments show that none of these approaches overcome the zero-reward barrier if the base model never produces a correct answer. In contrast, we find that a simple data-centric intervention of adding easier samples to the training set enables the model to eventually solve the original hard task despite starting from zero reward. Importantly, this succeeds without modifying the RL algorithm itself. Because official implementations of several baselines were unavailable, we developed our own, which allowed us to conduct a detailed analysis of their failure modes. We release these implementations to support further research at: https://github.com/rl4reasoning/rl-baselines

Beyond Softmax: A New Perspective on Gradient Bandits

arXiv:2510.03979v1 Announce Type: new Abstract: We establish a link between a class of discrete choice models and the theory of online learning and multi-armed bandits. Our contributions are: (i) sublinear regret bounds for a broad algorithmic family, encompassing Exp3 as a special case; (ii) a new class of adversarial bandit algorithms derived from generalized nested logit models \citep{wen:2001}; and (iii) \textcolor{black}{we introduce a novel class of generalized gradient bandit algorithms that extends beyond the widely used softmax formulation. By relaxing the restrictive independence assumptions inherent in softmax, our framework accommodates correlated learning dynamics across actions, thereby broadening the applicability of gradient bandit methods.} Overall, the proposed algorithms combine flexible model specification with computational efficiency via closed-form sampling probabilities. Numerical experiments in stochastic bandit settings demonstrate their practical effectiveness.

ICEPool: Enhancing Graph Pooling Networks with Inter-cluster Connectivity

arXiv:2510.03987v1 Announce Type: new Abstract: Hierarchical Pooling Models have demonstrated strong performance in classifying graph-structured data. While numerous innovative methods have been proposed to design cluster assignments and coarsening strategies, the relationships between clusters are often overlooked. In this paper, we introduce Inter-cluster Connectivity Enhancement Pooling (ICEPool), a novel hierarchical pooling framework designed to enhance model's understanding of inter-cluster connectivity and ability of preserving the structural integrity in the original graph. ICEPool is compatible with a wide range of pooling-based GNN models. The deployment of ICEPool as an enhancement to existing models effectively combines the strengths of the original model with ICEPool's capability to emphasize the integration of inter-cluster connectivity, resulting in a more comprehensive and robust graph-level representation. Moreover, we make theoretical analysis to ICEPool's ability of graph reconstruction to demonstrate its effectiveness in learning inter-cluster relationship that is overlooked by conventional models. Finally, the experimental results show the compatibility of ICEPool with wide varieties of models and its potential to boost the performance of existing graph neural network architectures.

Distilling Reasoning into Student LLMs: Local Naturalness for Selecting Teacher Data

arXiv:2510.03988v1 Announce Type: new Abstract: Distilling long reasoning traces (10K+ tokens) from stronger teacher models into smaller student LLMs via SFT has emerged as a standard paradigm. This approach is practical and efficient: it leverages the ease of generating abundant reasoning data from stronger models and provides a direct, data-driven way to teach less capable models better reasoning. While previous work has largely focused on prompt selection with responses from a single teacher, the equally important problem of choosing the best response when multiple teacher outputs are available for a single prompt remains underexplored. This challenge becomes important in a multi-teacher setting, where different students may benefit from the outputs of different teachers. This paper fills that gap with a systematic study of response selection for reasoning distillation. We first show that the current method, which picks responses the student assigns the highest global log-probability (global naturalness), fails when responses come from multiple teachers, i.e., global naturalness no longer correlates with downstream performance, especially as the reasoning traces from strong teachers become longer. To overcome this problem, we introduce Local Naturalness, which measures the student's log-probabilities over short, sequential reasoning steps conditioned only on a small local window. Local Naturalness enables two applications: 1) Teacher Selection: Aggregating local scores across prompts reliably identifies the most helpful teacher. 2) Response Selection from a Multiple Teachers: When mixing answers from many teachers, Local Naturalness boosts a 32B student's accuracy on math benchmarks by 9.4pp over global selection, also surpassing the performance achieved by training on data from the single best teacher. These results highlight the power of localized data quality evaluation and data mixing for more effective reasoning distillation.

A Mathematical Explanation of Transformers for Large Language Models and GPTs

arXiv:2510.03989v1 Announce Type: new Abstract: The Transformer architecture has revolutionized the field of sequence modeling and underpins the recent breakthroughs in large language models (LLMs). However, a comprehensive mathematical theory that explains its structure and operations remains elusive. In this work, we propose a novel continuous framework that rigorously interprets the Transformer as a discretization of a structured integro-differential equation. Within this formulation, the self-attention mechanism emerges naturally as a non-local integral operator, and layer normalization is characterized as a projection to a time-dependent constraint. This operator-theoretic and variational perspective offers a unified and interpretable foundation for understanding the architecture's core components, including attention, feedforward layers, and normalization. Our approach extends beyond previous theoretical analyses by embedding the entire Transformer operation in continuous domains for both token indices and feature dimensions. This leads to a principled and flexible framework that not only deepens theoretical insight but also offers new directions for architecture design, analysis, and control-based interpretations. This new interpretation provides a step toward bridging the gap between deep learning architectures and continuous mathematical modeling, and contributes a foundational perspective to the ongoing development of interpretable and theoretically grounded neural network models.

Incorporating Multivariate Consistency in ML-Based Weather Forecasting with Latent-space Constraints

arXiv:2510.04006v1 Announce Type: new Abstract: Data-driven machine learning (ML) models have recently shown promise in surpassing traditional physics-based approaches for weather forecasting, leading to a so-called second revolution in weather forecasting. However, most ML-based forecast models treat reanalysis as the truth and are trained under variable-specific loss weighting, ignoring their physical coupling and spatial structure. Over long time horizons, the forecasts become blurry and physically unrealistic under rollout training. To address this, we reinterpret model training as a weak-constraint four-dimensional variational data assimilation (WC-4DVar) problem, treating reanalysis data as imperfect observations. This allows the loss function to incorporate reanalysis error covariance and capture multivariate dependencies. In practice, we compute the loss in a latent space learned by an autoencoder (AE), where the reanalysis error covariance becomes approximately diagonal, thus avoiding the need to explicitly model it in the high-dimensional model space. We show that rollout training with latent-space constraints improves long-term forecast skill and better preserves fine-scale structures and physical realism compared to training with model-space loss. Finally, we extend this framework to accommodate heterogeneous data sources, enabling the forecast model to be trained jointly on reanalysis and multi-source observations within a unified theoretical formulation.

Replacing Softmax Similarity with a Sharpened Angular Similarity: Theory and Practice of Scaling To Billion-Context Attention

arXiv:2510.04008v1 Announce Type: new Abstract: Softmax Attention has a quadratic time complexity, which becomes prohibitive to run at long contexts, even with highly optimized GPU kernels. For example, FlashAttention (an exact, GPU-optimized implementation of Softmax Attention) cannot complete a single forward-backward pass of a multi-head attention layer once the context exceeds ~4 million tokens on an NVIDIA GH200 (96 GB). We introduce RACE Attention, a kernel-inspired alternative to Softmax Attention that is linear in sequence length and embedding dimension. RACE Attention replaces the exponential kernel with a sharpened angular (cosine) similarity, and approximates attention outputs via randomized projections and soft Locality-Sensitive Hashing (LSH). Across language modeling, masked language modeling, and text classification, RACE Attention matches the accuracy of strong baselines while reducing runtime and memory. In a controlled scale test, it processes up to 12 million tokens during a single forward-backward pass on an NVIDIA GH200 GPU and 75 million tokens on an Intel Xeon Gold 5220R CPU, well beyond the practical limits of the current state-of-the-art attention implementations. RACE Attention thus offers a practical, theoretically grounded mechanism for outrageously long context windows on today's hardware. We hope that it gets adopted in practice.

Principled and Tractable RL for Reasoning with Diffusion Language Models

arXiv:2510.04019v1 Announce Type: new Abstract: Diffusion large language models (dLLMs) are a new paradigm of non-autoregressive language models that are trained to predict multiple tokens in parallel and generate text via iterative unmasking. Recent works have successfully pretrained dLLMs to parity with autoregressive LLMs at the 8B scale, but dLLMs have yet to benefit from modern post-training techniques, e.g. reinforcement learning (RL), that have proven effective for autoregressive models. Crucially, algorithms designed for traditional LLMs aren't directly compatible with diffusion frameworks due to inherent differences in modeling assumptions. Moreover, existing attempts at dLLM post-training with RL rely on heuristic-based objectives with no theoretical grounding. In this work, we present Amortized Group Relative Policy Optimization (AGRPO), a principled on-policy RL algorithm designed specifically for dLLMs. AGRPO uses Monte Carlo sampling to compute an unbiased policy gradient estimate, making it the first tractable, faithful adaptation of policy gradient methods for dLLMs. We demonstrate AGRPO's effectiveness on different math/reasoning tasks, a common setting for RL with LLMs, achieving up to +7.6% absolute gain on GSM8K and 3.8x performance on the Countdown task over the baseline LLaDA-8B-Instruct model and 1.3x performance gains over comparable RL methods such as diffu-GRPO. Furthermore, these gains persist across different numbers of sampling steps at inference time, achieving better tradeoffs between compute and performance. Our results demonstrate that online RL algorithms can be extended to diffusion LLMs in principled ways, maintaining both theoretical soundness and practical effectiveness.

Spatiotemporal Forecasting as Planning: A Model-Based Reinforcement Learning Approach with Generative World Models

arXiv:2510.04020v1 Announce Type: new Abstract: To address the dual challenges of inherent stochasticity and non-differentiable metrics in physical spatiotemporal forecasting, we propose Spatiotemporal Forecasting as Planning (SFP), a new paradigm grounded in Model-Based Reinforcement Learning. SFP constructs a novel Generative World Model to simulate diverse, high-fidelity future states, enabling an "imagination-based" environmental simulation. Within this framework, a base forecasting model acts as an agent, guided by a beam search-based planning algorithm that leverages non-differentiable domain metrics as reward signals to explore high-return future sequences. These identified high-reward candidates then serve as pseudo-labels to continuously optimize the agent's policy through iterative self-training, significantly reducing prediction error and demonstrating exceptional performance on critical domain metrics like capturing extreme events.

Multi-Class Support Vector Machine with Differential Privacy

arXiv:2510.04027v1 Announce Type: new Abstract: With the increasing need to safeguard data privacy in machine learning models, differential privacy (DP) is one of the major frameworks to build privacy-preserving models. Support Vector Machines (SVMs) are widely used traditional machine learning models due to their robust margin guarantees and strong empirical performance in binary classification. However, applying DP to multi-class SVMs is inadequate, as the standard one-versus-rest (OvR) and one-versus-one (OvO) approaches repeatedly query each data sample when building multiple binary classifiers, thus consuming the privacy budget proportionally to the number of classes. To overcome this limitation, we explore all-in-one SVM approaches for DP, which access each data sample only once to construct multi-class SVM boundaries with margin maximization properties. We propose a novel differentially Private Multi-class SVM (PMSVM) with weight and gradient perturbation methods, providing rigorous sensitivity and convergence analyses to ensure DP in all-in-one SVMs. Empirical results demonstrate that our approach surpasses existing DP-SVM methods in multi-class scenarios.

The Debate on RLVR Reasoning Capability Boundary: Shrinkage, Expansion, or Both? A Two-Stage Dynamic View

arXiv:2510.04028v1 Announce Type: new Abstract: The ongoing debate on whether reinforcement learning with verifiable rewards (RLVR) expands or shrinks the reasoning capabilities of large language models (LLMs) remains unresolved. Some studies contend that RLVR mainly improves sampling efficiency but at the expense of diversity and exploratory capacity, resulting in capability boundary shrinkage. In contrast, others demonstrate that prolonged training can lead to the emergence of novel reasoning strategies, suggesting capability boundary expansion. To reconcile these contradictory findings, we theoretically and empirically show that both perspectives are partially valid-each aligning with a separate phase in an inherent two-stage probability mass dynamic: (1) Exploitation stage: initially, the model primarily samples explored high-reward and low-reward tokens, while rarely selecting the potentially optimal token. Positive advantage estimates increase the probability of high-reward tokens and decrease those of low-reward tokens, yet the optimal token's probability remains largely unchanged during this stage. (2) Exploration stage: as training advances, the growth rate of previously acquired high-reward tokens slows as their probabilities approach saturation. When a potentially optimal token-now receiving positive advantage estimates-is occasionally sampled, its probability increases, while those of the originally high-reward tokens decrease. This dynamic suggests that over-exploitation during the exploitation stage may lead to capability boundary shrinkage, whereas prolonged training into the exploration stage can promote an expansion of the reasoning capability boundary. Building upon our insights, we revisit the potential of only using relative negative gradients for prolonging training, providing a theoretical and empirical foundation for the development of more advanced reasoning capabilities.

Adaptive kernel-density approach for imbalanced binary classification

arXiv:2510.04046v1 Announce Type: new Abstract: Class imbalance is a common challenge in real-world binary classification tasks, often leading to predictions biased toward the majority class and reduced recognition of the minority class. This issue is particularly critical in domains such as medical diagnosis and anomaly detection, where correct classification of minority classes is essential. Conventional methods often fail to deliver satisfactory performance when the imbalance ratio is extremely severe. To address this challenge, we propose a novel approach called Kernel-density-Oriented Threshold Adjustment with Regional Optimization (KOTARO), which extends the framework of kernel density estimation (KDE) by adaptively adjusting decision boundaries according to local sample density. In KOTARO, the bandwidth of Gaussian basis functions is dynamically tuned based on the estimated density around each sample, thereby enhancing the classifier's ability to capture minority regions. We validated the effectiveness of KOTARO through experiments on both synthetic and real-world imbalanced datasets. The results demonstrated that KOTARO outperformed conventional methods, particularly under conditions of severe imbalance, highlighting its potential as a promising solution for a wide range of imbalanced classification problems

Variational Diffusion Unlearning: A Variational Inference Framework for Unlearning in Diffusion Models under Data Constraints

arXiv:2510.04058v1 Announce Type: new Abstract: For a responsible and safe deployment of diffusion models in various domains, regulating the generated outputs from these models is desirable because such models could generate undesired, violent, and obscene outputs. To tackle this problem, recent works use machine unlearning methodology to forget training data points containing these undesired features from pre-trained generative models. However, these methods proved to be ineffective in data-constrained settings where the whole training dataset is inaccessible. Thus, the principal objective of this work is to propose a machine unlearning methodology that can prevent the generation of outputs containing undesired features from a pre-trained diffusion model in such a data-constrained setting. Our proposed method, termed as Variational Diffusion Unlearning (VDU), is a computationally efficient method that only requires access to a subset of training data containing undesired features. Our approach is inspired by the variational inference framework with the objective of minimizing a loss function consisting of two terms: plasticity inducer and stability regularizer. Plasticity inducer reduces the log-likelihood of the undesired training data points, while the stability regularizer, essential for preventing loss of image generation quality, regularizes the model in parameter space. We validate the effectiveness of our method through comprehensive experiments for both class unlearning and feature unlearning. For class unlearning, we unlearn some user-identified classes from MNIST, CIFAR-10, and tinyImageNet datasets from a pre-trained unconditional denoising diffusion probabilistic model (DDPM). Similarly, for feature unlearning, we unlearn the generation of certain high-level features from a pre-trained Stable Diffusion model

What Scales in Cross-Entropy Scaling Law?

arXiv:2510.04067v1 Announce Type: new Abstract: The cross-entropy scaling law has long served as a key tool for guiding the development of large language models. It shows that cross-entropy loss decreases in a predictable power-law rate as the model size increases. However, recent evidence indicates that this law breaks down at very large scales: the loss decreases more slowly than expected, which causes significant trouble for developing large language models. In this paper, we hypothesize that the root cause lies in the fact that cross-entropy itself does not truly scale; instead, only one of its hidden components does. To investigate this, we introduce a novel decomposition of cross-entropy into three parts: Error-Entropy, Self-Alignment, and Confidence. We show both theoretically and empirically that this decomposition precisely captures the training dynamics and optimization objectives. Through extensive experiments on multiple datasets and 32 models spanning five orders of magnitude in size, we find that only error-entropy follows a robust power-law scaling, while the other two terms remain largely invariant. Moreover, error-entropy constitutes the dominant share of cross-entropy in small models but diminishes in proportion as models grow larger. This explains why the cross-entropy scaling law appears accurate at small scales but fails at very large ones. Our findings establish the error-entropy scaling law as a more accurate description of model behavior. We believe it will have wide applications in the training, understanding, and future development of large language models.

Slow-Fast Policy Optimization: Reposition-Before-Update for LLM Reasoning

arXiv:2510.04072v1 Announce Type: new Abstract: Reinforcement learning (RL) has become central to enhancing reasoning in large language models (LLMs). Yet on-policy algorithms such as Group Relative Policy Optimization (GRPO) often suffer in early training: noisy gradients from low-quality rollouts lead to unstable updates and inefficient exploration. We introduce Slow-Fast Policy Optimization (SFPO), a simple yet efficient framework to address these limitations via decomposing each step into three stages: a short fast trajectory of inner steps on the same batch, a reposition mechanism to control off-policy drift, and a final slow correction. This reposition-before-update design preserves the objective and rollout process unchanged, making SFPO plug-compatible with existing policy-gradient pipelines. Extensive experiments demonstrate that SFPO consistently improves stability, reduces rollouts, and accelerates convergence of reasoning RL training. Specifically, it outperforms GRPO by up to 2.80 points in average on math reasoning benchmarks. It also achieves up to 4.93\texttimes{} fewer rollouts and a 4.19\texttimes{} reduction in wall-clock time to match GRPO's best accuracy.

Offline Reinforcement Learning in Large State Spaces: Algorithms and Guarantees

arXiv:2510.04088v1 Announce Type: new Abstract: This article introduces the theory of offline reinforcement learning in large state spaces, where good policies are learned from historical data without online interactions with the environment. Key concepts introduced include expressivity assumptions on function approximation (e.g., Bellman completeness vs. realizability) and data coverage (e.g., all-policy vs. single-policy coverage). A rich landscape of algorithms and results is described, depending on the assumptions one is willing to make and the sample and computational complexity guarantees one wishes to achieve. We also discuss open questions and connections to adjacent areas.

Using predefined vector systems as latent space configuration for neural network supervised training on data with arbitrarily large number of classes

arXiv:2510.04090v1 Announce Type: new Abstract: Supervised learning (SL) methods are indispensable for neural network (NN) training used to perform classification tasks. While resulting in very high accuracy, SL training often requires making NN parameter number dependent on the number of classes, limiting their applicability when the number of classes is extremely large or unknown in advance. In this paper we propose a methodology that allows one to train the same NN architecture regardless of the number of classes. This is achieved by using predefined vector systems as the target latent space configuration (LSC) during NN training. We discuss the desired properties of target configurations and choose randomly perturbed vectors of An root system for our experiments. These vectors are used to successfully train encoders and visual transformers (ViT) on Cinic-10 and ImageNet-1K in low- and high-dimensional cases by matching NN predictions with the predefined vectors. Finally, ViT is trained on a dataset with 1.28 million classes illustrating the applicability of the method to training on datasets with extremely large number of classes. In addition, potential applications of LSC in lifelong learning and NN distillation are discussed illustrating versatility of the proposed methodology.

Rethinking Consistent Multi-Label Classification under Inexact Supervision

arXiv:2510.04091v1 Announce Type: new Abstract: Partial multi-label learning and complementary multi-label learning are two popular weakly supervised multi-label classification paradigms that aim to alleviate the high annotation costs of collecting precisely annotated multi-label data. In partial multi-label learning, each instance is annotated with a candidate label set, among which only some labels are relevant; in complementary multi-label learning, each instance is annotated with complementary labels indicating the classes to which the instance does not belong. Existing consistent approaches for the two paradigms either require accurate estimation of the generation process of candidate or complementary labels or assume a uniform distribution to eliminate the estimation problem. However, both conditions are usually difficult to satisfy in real-world scenarios. In this paper, we propose consistent approaches that do not rely on the aforementioned conditions to handle both problems in a unified way. Specifically, we propose two unbiased risk estimators based on first- and second-order strategies. Theoretically, we prove consistency w.r.t. two widely used multi-label classification evaluation metrics and derive convergence rates for the estimation errors of the proposed risk estimators. Empirically, extensive experimental results validate the effectiveness of our proposed approaches against state-of-the-art methods.

Why Cannot Neural Networks Master Extrapolation? Insights from Physical Laws

arXiv:2510.04102v1 Announce Type: new Abstract: Motivated by the remarkable success of Foundation Models (FMs) in language modeling, there has been growing interest in developing FMs for time series prediction, given the transformative power such models hold for science and engineering. This culminated in significant success of FMs in short-range forecasting settings. However, extrapolation or long-range forecasting remains elusive for FMs, which struggle to outperform even simple baselines. This contrasts with physical laws which have strong extrapolation properties, and raises the question of the fundamental difference between the structure of neural networks and physical laws. In this work, we identify and formalize a fundamental property characterizing the ability of statistical learning models to predict more accurately outside of their training domain, hence explaining performance deterioration for deep learning models in extrapolation settings. In addition to a theoretical analysis, we present empirical results showcasing the implications of this property on current deep learning architectures. Our results not only clarify the root causes of the extrapolation gap but also suggest directions for designing next-generation forecasting models capable of mastering extrapolation.

Can Linear Probes Measure LLM Uncertainty?

arXiv:2510.04108v1 Announce Type: new Abstract: Effective Uncertainty Quantification (UQ) represents a key aspect for reliable deployment of Large Language Models (LLMs) in automated decision-making and beyond. Yet, for LLM generation with multiple choice structure, the state-of-the-art in UQ is still dominated by the naive baseline given by the maximum softmax score. To address this shortcoming, we demonstrate that taking a principled approach via Bayesian statistics leads to improved performance despite leveraging the simplest possible model, namely linear regression. More precisely, we propose to train multiple Bayesian linear models, each predicting the output of a layer given the output of the previous one. Based on the obtained layer-level posterior distributions, we infer the global uncertainty level of the LLM by identifying a sparse combination of distributional features, leading to an efficient UQ scheme. Numerical experiments on various LLMs show consistent improvement over state-of-the-art baselines.

Wasserstein projection distance for fairness testing of regression models

arXiv:2510.04114v1 Announce Type: new Abstract: Fairness in machine learning is a critical concern, yet most research has focused on classification tasks, leaving regression models underexplored. This paper introduces a Wasserstein projection-based framework for fairness testing in regression models, focusing on expectation-based criteria. We propose a hypothesis-testing approach and an optimal data perturbation method to improve fairness while balancing accuracy. Theoretical results include a detailed categorization of fairness criteria for regression, a dual reformulation of the Wasserstein projection test statistic, and the derivation of asymptotic bounds and limiting distributions. Experiments on synthetic and real-world datasets demonstrate that the proposed method offers higher specificity compared to permutation-based tests, and effectively detects and mitigates biases in real applications such as student performance and housing price prediction.

On the Statistical Query Complexity of Learning Semiautomata: a Random Walk Approach

arXiv:2510.04115v1 Announce Type: new Abstract: Semiautomata form a rich class of sequence-processing algorithms with applications in natural language processing, robotics, computational biology, and data mining. We establish the first Statistical Query hardness result for semiautomata under the uniform distribution over input words and initial states. We show that Statistical Query hardness can be established when both the alphabet size and input length are polynomial in the number of states. Unlike the case of deterministic finite automata, where hardness typically arises through the hardness of the language they recognize (e.g., parity), our result is derived solely from the internal state-transition structure of semiautomata. Our analysis reduces the task of distinguishing the final states of two semiautomata to studying the behavior of a random walk on the group $S_{N} \times S_{N}$. By applying tools from Fourier analysis and the representation theory of the symmetric group, we obtain tight spectral gap bounds, demonstrating that after a polynomial number of steps in the number of states, distinct semiautomata become nearly uncorrelated, yielding the desired hardness result.

Attending on Multilevel Structure of Proteins enables Accurate Prediction of Cold-Start Drug-Target Interactions

arXiv:2510.04126v1 Announce Type: new Abstract: Cold-start drug-target interaction (DTI) prediction focuses on interaction between novel drugs and proteins. Previous methods typically learn transferable interaction patterns between structures of drug and proteins to tackle it. However, insight from proteomics suggest that protein have multi-level structures and they all influence the DTI. Existing works usually represent protein with only primary structures, limiting their ability to capture interactions involving higher-level structures. Inspired by this insight, we propose ColdDTI, a framework attending on protein multi-level structure for cold-start DTI prediction. We employ hierarchical attention mechanism to mine interaction between multi-level protein structures (from primary to quaternary) and drug structures at both local and global granularities. Then, we leverage mined interactions to fuse structure representations of different levels for final prediction. Our design captures biologically transferable priors, avoiding the risk of overfitting caused by excessive reliance on representation learning. Experiments on benchmark datasets demonstrate that ColdDTI consistently outperforms previous methods in cold-start settings.

On the Limitations and Capabilities of Position Embeddings for Length Generalization

arXiv:2510.04130v1 Announce Type: new Abstract: In Transformers, Position Embeddings (PEs) significantly influence Length Generalization (LG) performance, yet their fundamental role remains unclear. In this work, we investigate the limitations and capabilities of PEs in achieving LG. We theoretically analyze PEs in Position-Only Linear Attentions (POLAs), introducing Linear Representation Complexity (LRC) to characterize when PEs enable LG. Our analysis shows that PEs do not expand computational capabilities but structure learned computations across positions. Extending to practical Transformers, we propose Sequential Representation Complexity (SRC) and conjecture that LG is possible if and only if SRC remains invariant across scales. We support this hypothesis with empirical evidence in various reasoning tasks. To enhance LG, we introduce Scale Hint, allowing flexible instance scaling, and a Learning-Based Position Embedding framework that automatically learns positional relations. Our work provides theoretical insights and practical strategies for improving LG in Transformers.

Modeling Time Series Dynamics with Fourier Ordinary Differential Equations

arXiv:2510.04133v1 Announce Type: new Abstract: Neural ODEs (NODEs) have emerged as powerful tools for modeling time series data, offering the flexibility to adapt to varying input scales and capture complex dynamics. However, they face significant challenges: first, their reliance on time-domain representations often limits their ability to capture long-term dependencies and periodic structures; second, the inherent mismatch between their continuous-time formulation and the discrete nature of real-world data can lead to loss of granularity and predictive accuracy. To address these limitations, we propose Fourier Ordinary Differential Equations (FODEs), an approach that embeds the dynamics in the Fourier domain. By transforming time-series data into the frequency domain using the Fast Fourier Transform (FFT), FODEs uncover global patterns and periodic behaviors that remain elusive in the time domain. Additionally, we introduce a learnable element-wise filtering mechanism that aligns continuous model outputs with discrete observations, preserving granularity and enhancing accuracy. Experiments on various time series datasets demonstrate that FODEs outperform existing methods in terms of both accuracy and efficiency. By effectively capturing both long- and short-term patterns, FODEs provide a robust framework for modeling time series dynamics.

PhaseFormer: From Patches to Phases for Efficient and Effective Time Series Forecasting

arXiv:2510.04134v1 Announce Type: new Abstract: Periodicity is a fundamental characteristic of time series data and has long played a central role in forecasting. Recent deep learning methods strengthen the exploitation of periodicity by treating patches as basic tokens, thereby improving predictive effectiveness. However, their efficiency remains a bottleneck due to large parameter counts and heavy computational costs. This paper provides, for the first time, a clear explanation of why patch-level processing is inherently inefficient, supported by strong evidence from real-world data. To address these limitations, we introduce a phase perspective for modeling periodicity and present an efficient yet effective solution, PhaseFormer. PhaseFormer features phase-wise prediction through compact phase embeddings and efficient cross-phase interaction enabled by a lightweight routing mechanism. Extensive experiments demonstrate that PhaseFormer achieves state-of-the-art performance with around 1k parameters, consistently across benchmark datasets. Notably, it excels on large-scale and complex datasets, where models with comparable efficiency often struggle. This work marks a significant step toward truly efficient and effective time series forecasting. Code is available at this repository: https://github.com/neumyor/PhaseFormer_TSL

Efficient Manifold-Constrained Neural ODE for High-Dimensional Datasets

arXiv:2510.04138v1 Announce Type: new Abstract: Neural ordinary differential equations (NODE) have garnered significant attention for their design of continuous-depth neural networks and the ability to learn data/feature dynamics. However, for high-dimensional systems, estimating dynamics requires extensive calculations and suffers from high truncation errors for the ODE solvers. To address the issue, one intuitive approach is to consider the non-trivial topological space of the data distribution, i.e., a low-dimensional manifold. Existing methods often rely on knowledge of the manifold for projection or implicit transformation, restricting the ODE solutions on the manifold. Nevertheless, such knowledge is usually unknown in realistic scenarios. Therefore, we propose a novel approach to explore the underlying manifold to restrict the ODE process. Specifically, we employ a structure-preserved encoder to process data and find the underlying graph to approximate the manifold. Moreover, we propose novel methods to combine the NODE learning with the manifold, resulting in significant gains in computational speed and accuracy. Our experimental evaluations encompass multiple datasets, where we compare the accuracy, number of function evaluations (NFEs), and convergence speed of our model against existing baselines. Our results demonstrate superior performance, underscoring the effectiveness of our approach in addressing the challenges of high-dimensional datasets.

Beyond Next-Token Prediction: A Performance Characterization of Diffusion versus Autoregressive Language Models

arXiv:2510.04146v1 Announce Type: new Abstract: Large Language Models (LLMs) have achieved state-of-the-art performance on a broad range of Natural Language Processing (NLP) tasks, including document processing and coding. Autoregressive Language Models (ARMs), which generate tokens sequentially conditioned on all previous tokens, have been the predominant paradigm for LLMs. However, while these networks have achieved high accuracy across a range of downstream tasks, they exhibit low arithmetic intensity due to the inherent sequential dependency with next-token prediction. Recently, Diffusion Language Models (DLMs) have emerged as a promising alternative architecture. DLMs generate output text in parallel, breaking the limitations of sequential dependency. However, the performance implications of DLMs relative to commonly deployed ARMs are not fully understood. In this work, we present a comprehensive performance study analyzing the performance characteristics of ARMs and DLMs, using both theoretical analysis and profiling data to characterize the trade-offs between these approaches. We illustrate that although DLMs exhibit higher arithmetic intensity compared to ARMs because of their capability to utilize parallelism across sequence lengths, they fail to scale effectively to longer contexts. We then explore DLMs with block-wise decoding, outlining how this approach allows for increased arithmetic intensity, while still scaling well to long contexts (similar to ARMs). We also show interesting trade-offs for batched inference, where we find that ARMs exhibit superior throughput, as they benefit more from parallelism across sequences in the batch. Finally, we highlight opportunities for accelerating DLM inference, and, in particular, highlight the importance of reducing the number of sampling steps for allowing open-source DLMs to provide improved latency relative to ARMs.

Finite Time Analysis of Constrained Natural Critic-Actor Algorithm with Improved Sample Complexity

arXiv:2510.04189v1 Announce Type: new Abstract: Recent studies have increasingly focused on non-asymptotic convergence analyses for actor-critic (AC) algorithms. One such effort introduced a two-timescale critic-actor algorithm for the discounted cost setting using a tabular representation, where the usual roles of the actor and critic are reversed. However, only asymptotic convergence was established there. Subsequently, both asymptotic and non-asymptotic analyses of the critic-actor algorithm with linear function approximation were conducted. In our work, we introduce the first natural critic-actor algorithm with function approximation for the long-run average cost setting and under inequality constraints. We provide the non-asymptotic convergence guarantees for this algorithm. Our analysis establishes optimal learning rates and we also propose a modification to enhance sample complexity. We further show the results of experiments on three different Safety-Gym environments where our algorithm is found to be competitive in comparison with other well known algorithms.

Spectral Alignment as Predictor of Loss Explosion in Neural Network Training

arXiv:2510.04202v1 Announce Type: new Abstract: Loss explosions in training deep neural networks can nullify multi-million dollar training runs. Conventional monitoring metrics like weight and gradient norms are often lagging and ambiguous predictors, as their values vary dramatically across different models and even between layers of the same model, making it difficult to establish a unified standard for detecting impending failure. We introduce Spectral Alignment (SA), a novel, theoretically-grounded metric that monitors the distributional alignment between layer inputs and the principal singular vectors of weight matrices. We show that a collapse in the sign diversity of this alignment is a powerful early predictor of representational collapse and training divergence. Empirical results on language models demonstrate that monitoring the SA distribution provides a significantly earlier and clearer warning of loss explosions than traditional scalar metrics. SA's low computational overhead makes it a practical tool for safeguarding model training.

Adaptive Federated Learning via Dynamical System Model

arXiv:2510.04203v1 Announce Type: new Abstract: Hyperparameter selection is critical for stable and efficient convergence of heterogeneous federated learning, where clients differ in computational capabilities, and data distributions are non-IID. Tuning hyperparameters is a manual and computationally expensive process as the hyperparameter space grows combinatorially with the number of clients. To address this, we introduce an end-to-end adaptive federated learning method in which both clients and central agents adaptively select their local learning rates and momentum parameters. Our approach models federated learning as a dynamical system, allowing us to draw on principles from numerical simulation and physical design. Through this perspective, selecting momentum parameters equates to critically damping the system for fast, stable convergence, while learning rates for clients and central servers are adaptively selected to satisfy accuracy properties from numerical simulation. The result is an adaptive, momentum-based federated learning algorithm in which the learning rates for clients and servers are dynamically adjusted and controlled by a single, global hyperparameter. By designing a fully integrated solution for both adaptive client updates and central agent aggregation, our method is capable of handling key challenges of heterogeneous federated learning, including objective inconsistency and client drift. Importantly, our approach achieves fast convergence while being insensitive to the choice of the global hyperparameter, making it well-suited for rapid prototyping and scalable deployment. Compared to state-of-the-art adaptive methods, our framework is shown to deliver superior convergence for heterogeneous federated learning while eliminating the need for hyperparameter tuning both client and server updates.

PolyKAN: A Polyhedral Analysis Framework for Provable and Minimal KAN Compression

arXiv:2510.04205v1 Announce Type: new Abstract: Kolmogorov-Arnold Networks (KANs) have emerged as a promising alternative to traditional Multi-Layer Perceptrons (MLPs), offering enhanced interpretability and a strong mathematical foundation. However, their parameter efficiency remains a significant challenge for practical deployment. This paper introduces PolyKAN, a novel theoretical framework for KAN compression that provides formal guarantees on both model size reduction and approximation error. By leveraging the inherent piecewise polynomial structure of KANs, we formulate the compression problem as one of optimal polyhedral region merging. We establish a rigorous polyhedral characterization of KANs, develop a complete theory of $\epsilon$-equivalent compression, and design an optimal dynamic programming algorithm that guarantees minimal compression under specified error bounds. Our theoretical analysis demonstrates that PolyKAN achieves provably minimal compression while maintaining strict error control, with polynomial-time complexity in all network parameters. The framework provides the first formal foundation for KAN compression with mathematical guarantees, opening new directions for efficient deployment of interpretable neural architectures.

Why Low-Precision Transformer Training Fails: An Analysis on Flash Attention

arXiv:2510.04212v1 Announce Type: new Abstract: The pursuit of computational efficiency has driven the adoption of low-precision formats for training transformer models. However, this progress is often hindered by notorious training instabilities. This paper provides the first mechanistic explanation for a long-standing and unresolved failure case where training with flash attention in low-precision settings leads to catastrophic loss explosions. Our in-depth analysis reveals that the failure is not a random artifact but caused by two intertwined phenomena: the emergence of similar low-rank representations within the attention mechanism and the compounding effect of biased rounding errors inherent in low-precision arithmetic. We demonstrate how these factors create a vicious cycle of error accumulation that corrupts weight updates, ultimately derailing the training dynamics. To validate our findings, we introduce a minimal modification to the flash attention that mitigates the bias in rounding errors. This simple change stabilizes the training process, confirming our analysis and offering a practical solution to this persistent problem.

MLLMEraser: Achieving Test-Time Unlearning in Multimodal Large Language Models through Activation Steering

arXiv:2510.04217v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) have demonstrated remarkable capabilities across vision-language tasks, yet their large-scale deployment raises pressing concerns about memorized private data, outdated knowledge, and harmful content. Existing unlearning approaches for MLLMs typically adapt training-based strategies such as gradient ascent or preference optimization, but these methods are computationally expensive, irreversible, and often distort retained knowledge. In this work, we propose MLLMEraser, an input-aware, training-free framework for test-time unlearning. Our approach leverages activation steering to enable dynamic knowledge erasure without parameter updates. Specifically, we construct a multimodal erasure direction by contrasting adversarially perturbed, knowledge-recall image-text pairs with knowledge-erasure counterparts, capturing both textual and visual discrepancies. To prevent unnecessary interference, we further design an input-aware steering mechanism that adaptively determines when and how the erasure direction should be applied, preserving utility on retained knowledge while enforcing forgetting on designated content. Experiments on LLaVA-1.5 and Qwen-2.5-VL demonstrate that MLLMEraser consistently outperforms state-of-the-art MLLM unlearning baselines, achieving stronger forgetting performance with lower computational cost and minimal utility degradation.

Physics-Inspired All-Pair Interaction Learning for 3D Dynamics Modeling

arXiv:2510.04233v1 Announce Type: new Abstract: Modeling 3D dynamics is a fundamental problem in multi-body systems across scientific and engineering domains and has important practical implications in trajectory prediction and simulation. While recent GNN-based approaches have achieved strong performance by enforcing geometric symmetries, encoding high-order features or incorporating neural-ODE mechanics, they typically depend on explicitly observed structures and inherently fail to capture the unobserved interactions that are crucial to complex physical behaviors and dynamics mechanism. In this paper, we propose PAINET, a principled SE(3)-equivariant neural architecture for learning all-pair interactions in multi-body systems. The model comprises: (1) a novel physics-inspired attention network derived from the minimization trajectory of an energy function, and (2) a parallel decoder that preserves equivariance while enabling efficient inference. Empirical results on diverse real-world benchmarks, including human motion capture, molecular dynamics, and large-scale protein simulations, show that PAINET consistently outperforms recently proposed models, yielding 4.7% to 41.5% error reductions in 3D dynamics prediction with comparable computation costs in terms of time and memory.

Truncated Kernel Stochastic Gradient Descent with General Losses and Spherical Radial Basis Functions

arXiv:2510.04237v1 Announce Type: new Abstract: In this paper, we propose a novel kernel stochastic gradient descent (SGD) algorithm for large-scale supervised learning with general losses. Compared to traditional kernel SGD, our algorithm improves efficiency and scalability through an innovative regularization strategy. By leveraging the infinite series expansion of spherical radial basis functions, this strategy projects the stochastic gradient onto a finite-dimensional hypothesis space, which is adaptively scaled according to the bias-variance trade-off, thereby enhancing generalization performance. Based on a new estimation of the spectral structure of the kernel-induced covariance operator, we develop an analytical framework that unifies optimization and generalization analyses. We prove that both the last iterate and the suffix average converge at minimax-optimal rates, and we further establish optimal strong convergence in the reproducing kernel Hilbert space. Our framework accommodates a broad class of classical loss functions, including least-squares, Huber, and logistic losses. Moreover, the proposed algorithm significantly reduces computational complexity and achieves optimal storage complexity by incorporating coordinate-wise updates from linear SGD, thereby avoiding the costly pairwise operations typical of kernel SGD and enabling efficient processing of streaming data. Finally, extensive numerical experiments demonstrate the efficiency of our approach.

Diffusion-Assisted Distillation for Self-Supervised Graph Representation Learning with MLPs

arXiv:2510.04241v1 Announce Type: new Abstract: For large-scale applications, there is growing interest in replacing Graph Neural Networks (GNNs) with lightweight Multi-Layer Perceptrons (MLPs) via knowledge distillation. However, distilling GNNs for self-supervised graph representation learning into MLPs is more challenging. This is because the performance of self-supervised learning is more related to the model's inductive bias than supervised learning. This motivates us to design a new distillation method to bridge a huge capacity gap between GNNs and MLPs in self-supervised graph representation learning. In this paper, we propose \textbf{D}iffusion-\textbf{A}ssisted \textbf{D}istillation for \textbf{S}elf-supervised \textbf{G}raph representation learning with \textbf{M}LPs (DAD-SGM). The proposed method employs a denoising diffusion model as a teacher assistant to better distill the knowledge from the teacher GNN into the student MLP. This approach enhances the generalizability and robustness of MLPs in self-supervised graph representation learning. Extensive experiments demonstrate that DAD-SGM effectively distills the knowledge of self-supervised GNNs compared to state-of-the-art GNN-to-MLP distillation methods. Our implementation is available at https://github.com/SeongJinAhn/DAD-SGM.

Efficient Latent Variable Causal Discovery: Combining Score Search and Targeted Testing

arXiv:2510.04263v1 Announce Type: new Abstract: Learning causal structure from observational data is especially challenging when latent variables or selection bias are present. The Fast Causal Inference (FCI) algorithm addresses this setting but often performs exhaustive conditional independence tests across many subsets, leading to spurious independence claims, extra or missing edges, and unreliable orientations. We present a family of score-guided mixed-strategy causal search algorithms that build on this tradition. First, we introduce BOSS-FCI and GRaSP-FCI, straightforward variants of GFCI that substitute BOSS or GRaSP for FGES, thereby retaining correctness while incurring different scalability tradeoffs. Second, we develop FCI Targeted-testing (FCIT), a novel mixed-strategy method that improves upon these variants by replacing exhaustive all-subsets testing with targeted tests guided by BOSS, yielding well-formed PAGs with higher precision and efficiency. Finally, we propose a simple heuristic, LV-Dumb (also known as BOSS-POD), which bypasses latent-variable-specific reasoning and directly returns the PAG of the BOSS DAG. Although not strictly correct in the FCI sense, it scales better and often achieves superior accuracy in practice. Simulations and real-data analyses demonstrate that BOSS-FCI and GRaSP-FCI provide sound baselines, FCIT improves both efficiency and reliability, and LV-Dumb offers a practical heuristic with strong empirical performance. Together, these method highlight the value of score-guided and targeted strategies for scalable latent-variable causal discovery.

Influence branching for learning to solve mixed-integer programs online

arXiv:2510.04273v1 Announce Type: new Abstract: On the occasion of the 20th Mixed Integer Program Workshop's computational competition, this work introduces a new approach for learning to solve MIPs online. Influence branching, a new graph-oriented variable selection strategy, is applied throughout the first iterations of the branch and bound algorithm. This branching heuristic is optimized online with Thompson sampling, which ranks the best graph representations of MIP's structure according to computational speed up over SCIP. We achieve results comparable to state of the art online learning methods. Moreover, our results indicate that our method generalizes well to more general online frameworks, where variations in constraint matrix, constraint vector and objective coefficients can all occur and where more samples are available.

A KL-regularization framework for learning to plan with adaptive priors

arXiv:2510.04280v1 Announce Type: new Abstract: Effective exploration remains a central challenge in model-based reinforcement learning (MBRL), particularly in high-dimensional continuous control tasks where sample efficiency is crucial. A prominent line of recent work leverages learned policies as proposal distributions for Model-Predictive Path Integral (MPPI) planning. Initial approaches update the sampling policy independently of the planner distribution, typically maximizing a learned value function with deterministic policy gradient and entropy regularization. However, because the states encountered during training depend on the MPPI planner, aligning the sampling policy with the planner improves the accuracy of value estimation and long-term performance. To this end, recent methods update the sampling policy by minimizing KL divergence to the planner distribution or by introducing planner-guided regularization into the policy update. In this work, we unify these MPPI-based reinforcement learning methods under a single framework by introducing Policy Optimization-Model Predictive Control (PO-MPC), a family of KL-regularized MBRL methods that integrate the planner's action distribution as a prior in policy optimization. By aligning the learned policy with the planner's behavior, PO-MPC allows more flexibility in the policy updates to trade off Return maximization and KL divergence minimization. We clarify how prior approaches emerge as special cases of this family, and we explore previously unstudied variations. Our experiments show that these extended configurations yield significant performance improvements, advancing the state of the art in MPPI-based RL.

HoRA: Cross-Head Low-Rank Adaptation with Joint Hypernetworks

arXiv:2510.04295v1 Announce Type: new Abstract: Low-Rank Adaptation (LoRA) is a parameter-efficient fine-tuning (PEFT) technique that adapts large pre-trained models by adding low-rank matrices to their weight updates. However, in the context of fine-tuning multi-head self-attention (MHA), LoRA has been employed to adapt each attention head separately, thereby overlooking potential synergies across different heads. To mitigate this issue, we propose a novel Hyper-shared Low-Rank Adaptation (HoRA) method, which utilizes joint hypernetworks to generate low-rank matrices across attention heads. By coupling their adaptation through a shared generator, HoRA encourages cross-head information sharing, and thus directly addresses the aforementioned limitation of LoRA. By comparing LoRA and HoRA through the lens of hierarchical mixture of experts, our theoretical findings reveal that the latter achieves superior sample efficiency to the former. Furthermore, through extensive experiments across diverse language and vision benchmarks, we demonstrate that HoRA outperforms LoRA and other PEFT methods while requiring only a marginal increase in the number of trainable parameters.

Wave-PDE Nets: Trainable Wave-Equation Layers as an Alternative to Attention

arXiv:2510.04304v1 Announce Type: new Abstract: We introduce Wave-PDE Nets, a neural architecture whose elementary operation is a differentiable simulation of the second-order wave equation. Each layer propagates its hidden state as a continuous field through a medium with trainable spatial velocity c(x) and damping {\gamma}(x). A symplectic spectral solver based on FFTs realises this propagation in O(nlog n) time. This oscillatory, global mechanism provides a powerful alternative to attention and first-order state-space models. We prove that a single Wave-PDE layer is a universal approximator. On language and vision benchmarks, Wave-PDE Nets match or exceed Transformer performance while demonstrating superior practical efficiency, reducing wall-clock time by up to 30% and peak memory by 25%. Ablation studies confirm the critical role of symplectic integration and a spectral Laplacian for stability and performance. Visualizations of the learned physical parameters reveal that the model learns intuitive strategies for information propagation. These results position Wave-PDE Nets as a computationally efficient and robust architecture with a strong physical inductive bias.

Activation Steering with a Feedback Controller

arXiv:2510.04309v1 Announce Type: new Abstract: Controlling the behaviors of large language models (LLM) is fundamental to their safety alignment and reliable deployment. However, existing steering methods are primarily driven by empirical insights and lack theoretical performance guarantees. In this work, we develop a control-theoretic foundation for activation steering by showing that popular steering methods correspond to the proportional (P) controllers, with the steering vector serving as the feedback signal. Building on this finding, we propose Proportional-Integral-Derivative (PID) Steering, a principled framework that leverages the full PID controller for activation steering in LLMs. The proportional (P) term aligns activations with target semantic directions, the integral (I) term accumulates errors to enforce persistent corrections across layers, and the derivative (D) term mitigates overshoot by counteracting rapid activation changes. This closed-loop design yields interpretable error dynamics and connects activation steering to classical stability guarantees in control theory. Moreover, PID Steering is lightweight, modular, and readily integrates with state-of-the-art steering methods. Extensive experiments across multiple LLM families and benchmarks demonstrate that PID Steering consistently outperforms existing approaches, achieving more robust and reliable behavioral control.

Crash Severity Prediction Using Deep Learning Approaches: A Hybrid CNN-RNN Framework

arXiv:2510.04316v1 Announce Type: new Abstract: Accurate and timely prediction of crash severity is crucial in mitigating the severe consequences of traffic accidents. Accurate and timely prediction of crash severity is crucial in mitigating the severe consequences of traffic accidents. In order to provide appropriate levels of medical assistance and transportation services, an intelligent transportation system relies on effective prediction methods. Deep learning models have gained popularity in this domain due to their capability to capture non-linear relationships among variables. In this research, we have implemented a hybrid CNN-RNN deep learning model for crash severity prediction and compared its performance against widely used statistical and machine learning models such as logistic regression, na\"ive bayes classifier, K-Nearest Neighbors (KNN), decision tree, and individual deep learning models: RNN and CNN. This study employs a methodology that considers the interconnected relationships between various features of traffic accidents. The study was conducted using a dataset of 15,870 accident records gathered over a period of seven years between 2015 and 2021 on Virginia highway I-64. The findings demonstrate that the proposed CNN-RNN hybrid model has outperformed all benchmark models in terms of predicting crash severity. This result illustrates the effectiveness of the hybrid model as it combines the advantages of both RNN and CNN models in order to achieve greater accuracy in the prediction process.

FairAgent: Democratizing Fairness-Aware Machine Learning with LLM-Powered Agents

arXiv:2510.04317v1 Announce Type: new Abstract: Training fair and unbiased machine learning models is crucial for high-stakes applications, yet it presents significant challenges. Effective bias mitigation requires deep expertise in fairness definitions, metrics, data preprocessing, and machine learning techniques. In addition, the complex process of balancing model performance with fairness requirements while properly handling sensitive attributes makes fairness-aware model development inaccessible to many practitioners. To address these challenges, we introduce FairAgent, an LLM-powered automated system that significantly simplifies fairness-aware model development. FairAgent eliminates the need for deep technical expertise by automatically analyzing datasets for potential biases, handling data preprocessing and feature engineering, and implementing appropriate bias mitigation strategies based on user requirements. Our experiments demonstrate that FairAgent achieves significant performance improvements while significantly reducing development time and expertise requirements, making fairness-aware machine learning more accessible to practitioners.

FoilDiff: A Hybrid Transformer Backbone for Diffusion-based Modelling of 2D Airfoil Flow Fields

arXiv:2510.04325v1 Announce Type: new Abstract: The accurate prediction of flow fields around airfoils is crucial for aerodynamic design and optimisation. Computational Fluid Dynamics (CFD) models are effective but computationally expensive, thus inspiring the development of surrogate models to enable quicker predictions. These surrogate models can be based on deep learning architectures, such as Convolutional Neural Networks (CNNs), Graph Neural Networks (GNNs), and Diffusion Models (DMs). Diffusion models have shown significant promise in predicting complex flow fields. In this work, we propose FoilDiff, a diffusion-based surrogate model with a hybrid-backbone denoising network. This hybrid design combines the power of convolutional feature extraction and transformer-based global attention to generate more adaptable and accurate representations of flow structures. FoilDiff takes advantage of Denoising Diffusion Implicit Model (DDIM) sampling to optimise the efficiency of the sampling process at no additional cost to model generalisation. We used encoded representations of Reynolds number, angle of attack, and airfoil geometry to define the input space for generalisation across a wide range of aerodynamic conditions. When evaluated against state-of-the-art models, FoilDiff shows significant performance improvements, with mean prediction errors reducing by up to 85\% on the same datasets. The results have demonstrated that FoilDiff can provide both more accurate predictions and better-calibrated predictive uncertainty than existing diffusion-based models.

Arithmetic-Mean $\mu$P for Modern Architectures: A Unified Learning-Rate Scale for CNNs and ResNets

arXiv:2510.04327v1 Announce Type: new Abstract: Choosing an appropriate learning rate remains a key challenge in scaling depth of modern deep networks. The classical maximal update parameterization ($\mu$P) enforces a fixed per-layer update magnitude, which is well suited to homogeneous multilayer perceptrons (MLPs) but becomes ill-posed in heterogeneous architectures where residual accumulation and convolutions introduce imbalance across layers. We introduce Arithmetic-Mean $\mu$P (AM-$\mu$P), which constrains not each individual layer but the network-wide average one-step pre-activation second moment to a constant scale. Combined with a residual-aware He fan-in initialization - scaling residual-branch weights by the number of blocks ($\mathrm{Var}[W]=c/(K\cdot \mathrm{fan\text{-}in})$) - AM-$\mu$P yields width-robust depth laws that transfer consistently across depths. We prove that, for one- and two-dimensional convolutional networks, the maximal-update learning rate satisfies $\eta^\star(L)\propto L^{-3/2}$; with zero padding, boundary effects are constant-level as $N\gg k$. For standard residual networks with general conv+MLP blocks, we establish $\eta^\star(L)=\Theta(L^{-3/2})$, with $L$ the minimal depth. Empirical results across a range of depths confirm the $-3/2$ scaling law and enable zero-shot learning-rate transfer, providing a unified and practical LR principle for convolutional and deep residual networks without additional tuning overhead.

DoRAN: Stabilizing Weight-Decomposed Low-Rank Adaptation via Noise Injection and Auxiliary Networks

arXiv:2510.04331v1 Announce Type: new Abstract: Parameter-efficient fine-tuning (PEFT) methods have become the standard paradigm for adapting large-scale models. Among these techniques, Weight-Decomposed Low-Rank Adaptation (DoRA) has been shown to improve both the learning capacity and training stability of the vanilla Low-Rank Adaptation (LoRA) method by explicitly decomposing pre-trained weights into magnitude and directional components. In this work, we propose DoRAN, a new variant of DoRA designed to further stabilize training and boost the sample efficiency of DoRA. Our approach includes two key stages: (i) injecting noise into the denominator of DoRA's weight decomposition, which serves as an adaptive regularizer to mitigate instabilities; and (ii) replacing static low-rank matrices with auxiliary networks that generate them dynamically, enabling parameter coupling across layers and yielding better sample efficiency in both theory and practice. Comprehensive experiments on vision and language benchmarks show that DoRAN consistently outperforms LoRA, DoRA, and other PEFT baselines. These results underscore the effectiveness of combining stabilization through noise-based regularization with network-based parameter generation, offering a promising direction for robust and efficient fine-tuning of foundation models.

Critical appraisal of artificial intelligence for rare-event recognition: principles and pharmacovigilance case studies

arXiv:2510.04341v1 Announce Type: new Abstract: Many high-stakes AI applications target low-prevalence events, where apparent accuracy can conceal limited real-world value. Relevant AI models range from expert-defined rules and traditional machine learning to generative LLMs constrained for classification. We outline key considerations for critical appraisal of AI in rare-event recognition, including problem framing and test set design, prevalence-aware statistical evaluation, robustness assessment, and integration into human workflows. In addition, we propose an approach to structured case-level examination (SCLE), to complement statistical performance evaluation, and a comprehensive checklist to guide procurement or development of AI models for rare-event recognition. We instantiate the framework in pharmacovigilance, drawing on three studies: rule-based retrieval of pregnancy-related reports; duplicate detection combining machine learning with probabilistic record linkage; and automated redaction of person names using an LLM. We highlight pitfalls specific to the rare-event setting including optimism from unrealistic class balance and lack of difficult positive controls in test sets - and show how cost-sensitive targets align model performance with operational value. While grounded in pharmacovigilance practice, the principles generalize to domains where positives are scarce and error costs may be asymmetric.

Learning to Predict Chaos: Curriculum-Driven Training for Robust Forecasting of Chaotic Dynamics

arXiv:2510.04342v1 Announce Type: new Abstract: Forecasting chaotic systems is a cornerstone challenge in many scientific fields, complicated by the exponential amplification of even infinitesimal prediction errors. Modern machine learning approaches often falter due to two opposing pitfalls: over-specializing on a single, well-known chaotic system (e.g., Lorenz-63), which limits generalizability, or indiscriminately mixing vast, unrelated time-series, which prevents the model from learning the nuances of any specific dynamical regime. We propose Curriculum Chaos Forecasting (CCF), a training paradigm that bridges this gap. CCF organizes training data based on fundamental principles of dynamical systems theory, creating a curriculum that progresses from simple, periodic behaviors to highly complex, chaotic dynamics. We quantify complexity using the largest Lyapunov exponent and attractor dimension, two well-established metrics of chaos. By first training a sequence model on predictable systems and gradually introducing more chaotic trajectories, CCF enables the model to build a robust and generalizable representation of dynamical behaviors. We curate a library of over 50 synthetic ODE/PDE systems to build this curriculum. Our experiments show that pre-training with CCF significantly enhances performance on unseen, real-world benchmarks. On datasets including Sunspot numbers, electricity demand, and human ECG signals, CCF extends the valid prediction horizon by up to 40% compared to random-order training and more than doubles it compared to training on real-world data alone. We demonstrate that this benefit is consistent across various neural architectures (GRU, Transformer) and provide extensive ablations to validate the importance of the curriculum's structure.

From News to Returns: A Granger-Causal Hypergraph Transformer on the Sphere

arXiv:2510.04357v1 Announce Type: new Abstract: We propose the Causal Sphere Hypergraph Transformer (CSHT), a novel architecture for interpretable financial time-series forecasting that unifies \emph{Granger-causal hypergraph structure}, \emph{Riemannian geometry}, and \emph{causally masked Transformer attention}. CSHT models the directional influence of financial news and sentiment on asset returns by extracting multivariate Granger-causal dependencies, which are encoded as directional hyperedges on the surface of a hypersphere. Attention is constrained via angular masks that preserve both temporal directionality and geometric consistency. Evaluated on S\&P 500 data from 2018 to 2023, including the 2020 COVID-19 shock, CSHT consistently outperforms baselines across return prediction, regime classification, and top-asset ranking tasks. By enforcing predictive causal structure and embedding variables in a Riemannian manifold, CSHT delivers both \emph{robust generalisation across market regimes} and \emph{transparent attribution pathways} from macroeconomic events to stock-level responses. These results suggest that CSHT is a principled and practical solution for trustworthy financial forecasting under uncertainty.

Quantifying Ambiguity in Categorical Annotations: A Measure and Statistical Inference Framework

arXiv:2510.04366v1 Announce Type: new Abstract: Human-generated categorical annotations frequently produce empirical response distributions (soft labels) that reflect ambiguity rather than simple annotator error. We introduce an ambiguity measure that maps a discrete response distribution to a scalar in the unit interval, designed to quantify aleatoric uncertainty in categorical tasks. The measure bears a close relationship to quadratic entropy (Gini-style impurity) but departs from those indices by treating an explicit "can't solve" category asymmetrically, thereby separating uncertainty arising from class-level indistinguishability from uncertainty due to explicit unresolvability. We analyze the measure's formal properties and contrast its behavior with a representative ambiguity measure from the literature. Moving beyond description, we develop statistical tools for inference: we propose frequentist point estimators for population ambiguity and derive the Bayesian posterior over ambiguity induced by Dirichlet priors on the underlying probability vector, providing a principled account of epistemic uncertainty. Numerical examples illustrate estimation, calibration, and practical use for dataset-quality assessment and downstream machine-learning workflows.

GDPval: Evaluating AI Model Performance on Real-World Economically Valuable Tasks

arXiv:2510.04374v1 Announce Type: new Abstract: We introduce GDPval, a benchmark evaluating AI model capabilities on real-world economically valuable tasks. GDPval covers the majority of U.S. Bureau of Labor Statistics Work Activities for 44 occupations across the top 9 sectors contributing to U.S. GDP (Gross Domestic Product). Tasks are constructed from the representative work of industry professionals with an average of 14 years of experience. We find that frontier model performance on GDPval is improving roughly linearly over time, and that the current best frontier models are approaching industry experts in deliverable quality. We analyze the potential for frontier models, when paired with human oversight, to perform GDPval tasks cheaper and faster than unaided experts. We also demonstrate that increased reasoning effort, increased task context, and increased scaffolding improves model performance on GDPval. Finally, we open-source a gold subset of 220 tasks and provide a public automated grading service at evals.openai.com to facilitate future research in understanding real-world model capabilities.

Adaptive Weighted Loss for Sequential Recommendations on Sparse Domains

arXiv:2510.04375v1 Announce Type: new Abstract: The effectiveness of single-model sequential recommendation architectures, while scalable, is often limited when catering to "power users" in sparse or niche domains. Our previous research, PinnerFormerLite, addressed this by using a fixed weighted loss to prioritize specific domains. However, this approach can be sub-optimal, as a single, uniform weight may not be sufficient for domains with very few interactions, where the training signal is easily diluted by the vast, generic dataset. This paper proposes a novel, data-driven approach: a Dynamic Weighted Loss function with comprehensive theoretical foundations and extensive empirical validation. We introduce an adaptive algorithm that adjusts the loss weight for each domain based on its sparsity in the training data, assigning a higher weight to sparser domains and a lower weight to denser ones. This ensures that even rare user interests contribute a meaningful gradient signal, preventing them from being overshadowed. We provide rigorous theoretical analysis including convergence proofs, complexity analysis, and bounds analysis to establish the stability and efficiency of our approach. Our comprehensive empirical validation across four diverse datasets (MovieLens, Amazon Electronics, Yelp Business, LastFM Music) with state-of-the-art baselines (SIGMA, CALRec, SparseEnNet) demonstrates that this dynamic weighting system significantly outperforms all comparison methods, particularly for sparse domains, achieving substantial lifts in key metrics like Recall at 10 and NDCG at 10 while maintaining performance on denser domains and introducing minimal computational overhead.

Categorical Invariants of Learning Dynamics

arXiv:2510.04376v1 Announce Type: new Abstract: Neural network training is typically viewed as gradient descent on a loss surface. We propose a fundamentally different perspective: learning is a structure-preserving transformation (a functor L) between the space of network parameters (Param) and the space of learned representations (Rep). This categorical framework reveals that different training runs producing similar test performance often belong to the same homotopy class (continuous deformation family) of optimization paths. We show experimentally that networks converging via homotopic trajectories generalize within 0.5% accuracy of each other, while non-homotopic paths differ by over 3%. The theory provides practical tools: persistent homology identifies stable minima predictive of generalization (R^2 = 0.82 correlation), pullback constructions formalize transfer learning, and 2-categorical structures explain when different optimization algorithms yield functionally equivalent models. These categorical invariants offer both theoretical insight into why deep learning works and concrete algorithmic principles for training more robust networks.

Score-based Greedy Search for Structure Identification of Partially Observed Linear Causal Models

arXiv:2510.04378v1 Announce Type: new Abstract: Identifying the structure of a partially observed causal system is essential to various scientific fields. Recent advances have focused on constraint-based causal discovery to solve this problem, and yet in practice these methods often face challenges related to multiple testing and error propagation. These issues could be mitigated by a score-based method and thus it has raised great attention whether there exists a score-based greedy search method that can handle the partially observed scenario. In this work, we propose the first score-based greedy search method for the identification of structure involving latent variables with identifiability guarantees. Specifically, we propose Generalized N Factor Model and establish the global consistency: the true structure including latent variables can be identified up to the Markov equivalence class by using score. We then design Latent variable Greedy Equivalence Search (LGES), a greedy search algorithm for this class of model with well-defined operators, which search very efficiently over the graph space to find the optimal structure. Our experiments on both synthetic and real-life data validate the effectiveness of our method (code will be publicly available).

SSM-CGM: Interpretable State-Space Forecasting Model of Continuous Glucose Monitoring for Personalized Diabetes Management

arXiv:2510.04386v1 Announce Type: new Abstract: Continuous glucose monitoring (CGM) generates dense data streams critical for diabetes management, but most used forecasting models lack interpretability for clinical use. We present SSM-CGM, a Mamba-based neural state-space forecasting model that integrates CGM and wearable activity signals from the AI-READI cohort. SSM-CGM improves short-term accuracy over a Temporal Fusion Transformer baseline, adds interpretability through variable selection and temporal attribution, and enables counterfactual forecasts simulating how planned changes in physiological signals (e.g., heart rate, respiration) affect near-term glucose. Together, these features make SSM-CGM an interpretable, physiologically grounded framework for personalized diabetes management.

Partial Information Decomposition via Normalizing Flows in Latent Gaussian Distributions

arXiv:2510.04417v1 Announce Type: new Abstract: The study of multimodality has garnered significant interest in fields where the analysis of interactions among multiple information sources can enhance predictive modeling, data fusion, and interpretability. Partial information decomposition (PID) has emerged as a useful information-theoretic framework to quantify the degree to which individual modalities independently, redundantly, or synergistically convey information about a target variable. However, existing PID methods depend on optimizing over a joint distribution constrained by estimated pairwise probability distributions, which are costly and inaccurate for continuous and high-dimensional modalities. Our first key insight is that the problem can be solved efficiently when the pairwise distributions are multivariate Gaussians, and we refer to this problem as Gaussian PID (GPID). We propose a new gradient-based algorithm that substantially improves the computational efficiency of GPID based on an alternative formulation of the underlying optimization problem. To generalize the applicability to non-Gaussian data, we learn information-preserving encoders to transform random variables of arbitrary input distributions into pairwise Gaussian random variables. Along the way, we resolved an open problem regarding the optimality of joint Gaussian solutions for GPID. Empirical validation in diverse synthetic examples demonstrates that our proposed method provides more accurate and efficient PID estimates than existing baselines. We further evaluate a series of large-scale multimodal benchmarks to show its utility in real-world applications of quantifying PID in multimodal datasets and selecting high-performing models.

Achieve Performatively Optimal Policy for Performative Reinforcement Learning

arXiv:2510.04430v1 Announce Type: new Abstract: Performative reinforcement learning is an emerging dynamical decision making framework, which extends reinforcement learning to the common applications where the agent's policy can change the environmental dynamics. Existing works on performative reinforcement learning only aim at a performatively stable (PS) policy that maximizes an approximate value function. However, there is a provably positive constant gap between the PS policy and the desired performatively optimal (PO) policy that maximizes the original value function. In contrast, this work proposes a zeroth-order Frank-Wolfe algorithm (0-FW) algorithm with a zeroth-order approximation of the performative policy gradient in the Frank-Wolfe framework, and obtains \textbf{the first polynomial-time convergence to the desired PO} policy under the standard regularizer dominance condition. For the convergence analysis, we prove two important properties of the nonconvex value function. First, when the policy regularizer dominates the environmental shift, the value function satisfies a certain gradient dominance property, so that any stationary point (not PS) of the value function is a desired PO. Second, though the value function has unbounded gradient, we prove that all the sufficiently stationary points lie in a convex and compact policy subspace $\Pi_{\Delta}$, where the policy value has a constant lower bound $\Delta>0$ and thus the gradient becomes bounded and Lipschitz continuous. Experimental results also demonstrate that our 0-FW algorithm is more effective than the existing algorithms in finding the desired PO policy.

Trade-off in Estimating the Number of Byzantine Clients in Federated Learning

arXiv:2510.04432v1 Announce Type: new Abstract: Federated learning has attracted increasing attention at recent large-scale optimization and machine learning research and applications, but is also vulnerable to Byzantine clients that can send any erroneous signals. Robust aggregators are commonly used to resist Byzantine clients. This usually requires to estimate the unknown number $f$ of Byzantine clients, and thus accordingly select the aggregators with proper degree of robustness (i.e., the maximum number $\hat{f}$ of Byzantine clients allowed by the aggregator). Such an estimation should have important effect on the performance, which has not been systematically studied to our knowledge. This work will fill in the gap by theoretically analyzing the worst-case error of aggregators as well as its induced federated learning algorithm for any cases of $\hat{f}$ and $f$. Specifically, we will show that underestimation ($\hat{f}<f$) can lead to arbitrarily poor performance for both aggregators and federated learning. For non-underestimation ($\hat{f}\ge f$), we have proved optimal lower and upper bounds of the same order on the errors of both aggregators and federated learning. All these optimal bounds are proportional to $\hat{f}/(n-f-\hat{f})$ with $n$ clients, which monotonically increases with larger $\hat{f}$. This indicates a fundamental trade-off: while an aggregator with a larger robustness degree $\hat{f}$ can solve federated learning problems of wider range $f\in [0,\hat{f}]$, the performance can deteriorate when there are actually fewer or even no Byzantine clients (i.e., $f\in [0,\hat{f})$).

Fractional Heat Kernel for Semi-Supervised Graph Learning with Small Training Sample Size

arXiv:2510.04440v1 Announce Type: new Abstract: In this work, we introduce novel algorithms for label propagation and self-training using fractional heat kernel dynamics with a source term. We motivate the methodology through the classical correspondence of information theory with the physics of parabolic evolution equations. We integrate the fractional heat kernel into Graph Neural Network architectures such as Graph Convolutional Networks and Graph Attention, enhancing their expressiveness through adaptive, multi-hop diffusion. By applying Chebyshev polynomial approximations, large graphs become computationally feasible. Motivating variational formulations demonstrate that by extending the classical diffusion model to fractional powers of the Laplacian, nonlocal interactions deliver more globally diffusing labels. The particular balance between supervision of known labels and diffusion across the graph is particularly advantageous in the case where only a small number of labeled training examples are present. We demonstrate the effectiveness of this approach on standard datasets.

Domain Generalization: A Tale of Two ERMs

arXiv:2510.04441v1 Announce Type: new Abstract: Domain generalization (DG) is the problem of generalizing from several distributions (or domains), for which labeled training data are available, to a new test domain for which no labeled data is available. A common finding in the DG literature is that it is difficult to outperform empirical risk minimization (ERM) on the pooled training data. In this work, we argue that this finding has primarily been reported for datasets satisfying a \emph{covariate shift} assumption. When the dataset satisfies a \emph{posterior drift} assumption instead, we show that ``domain-informed ERM,'' wherein feature vectors are augmented with domain-specific information, outperforms pooling ERM. These claims are supported by a theoretical framework and experiments on language and vision tasks.

Forking-Sequences

arXiv:2510.04487v1 Announce Type: new Abstract: While accuracy is a critical requirement for time series forecasting models, an equally important (yet often overlooked) desideratum is forecast stability across forecast creation dates (FCDs). Even highly accurate models can produce erratic revisions between FCDs, undermining stakeholder trust and disrupting downstream decision-making. To improve forecast stability, models like MQCNN, MQT, and SPADE employ a little-known but highly effective technique: forking-sequences. Unlike standard statistical and neural forecasting methods that treat each FCD independently, the forking-sequences method jointly encodes and decodes the entire time series across all FCDs, in a way mirroring time series cross-validation. Since forking sequences remains largely unknown in the broader neural forecasting community, in this work, we formalize the forking-sequences approach, and we make a case for its broader adoption. We demonstrate three key benefits of forking-sequences: (i) more stable and consistent gradient updates during training; (ii) reduced forecast variance through ensembling; and (iii) improved inference computational efficiency. We validate forking-sequences' benefits using 16 datasets from the M1, M3, M4, and Tourism competitions, showing improvements in forecast percentage change stability of 28.8%, 28.8%, 37.9%, and 31.3%, and 8.8%, on average, for MLP, RNN, LSTM, CNN, and Transformer-based architectures, respectively.

Expand Neurons, Not Parameters

arXiv:2510.04500v1 Announce Type: new Abstract: This work demonstrates how increasing the number of neurons in a network without increasing its number of non-zero parameters improves performance. We show that this gain corresponds with a decrease in interference between multiple features that would otherwise share the same neurons. To reduce such entanglement at a fixed non-zero parameter count, we introduce Fixed Parameter Expansion (FPE): replace a neuron with multiple children and partition the parent's weights disjointly across them, so that each child inherits a non-overlapping subset of connections. On symbolic tasks, specifically Boolean code problems, clause-aligned FPE systematically reduces polysemanticity metrics and yields higher task accuracy. Notably, random splits of neuron weights approximate these gains, indicating that reduced collisions, not precise assignment, are a primary driver. Consistent with the superposition hypothesis, the benefits of FPE grow with increasing interference: when polysemantic load is high, accuracy improvements are the largest. Transferring these insights to real models (classifiers over CLIP embeddings and deeper multilayer networks) we find that widening networks while maintaining a constant non-zero parameter count consistently increases accuracy. These results identify an interpretability-grounded mechanism to leverage width against superposition, improving performance without increasing the number of non-zero parameters. Such a direction is well matched to modern accelerators, where memory movement of non-zero parameters, rather than raw compute, is the dominant bottleneck.

Wavelet Predictive Representations for Non-Stationary Reinforcement Learning

arXiv:2510.04507v1 Announce Type: new Abstract: The real world is inherently non-stationary, with ever-changing factors, such as weather conditions and traffic flows, making it challenging for agents to adapt to varying environmental dynamics. Non-Stationary Reinforcement Learning (NSRL) addresses this challenge by training agents to adapt rapidly to sequences of distinct Markov Decision Processes (MDPs). However, existing NSRL approaches often focus on tasks with regularly evolving patterns, leading to limited adaptability in highly dynamic settings. Inspired by the success of Wavelet analysis in time series modeling, specifically its ability to capture signal trends at multiple scales, we propose WISDOM to leverage wavelet-domain predictive task representations to enhance NSRL. WISDOM captures these multi-scale features in evolving MDP sequences by transforming task representation sequences into the wavelet domain, where wavelet coefficients represent both global trends and fine-grained variations of non-stationary changes. In addition to the auto-regressive modeling commonly employed in time series forecasting, we devise a wavelet temporal difference (TD) update operator to enhance tracking and prediction of MDP evolution. We theoretically prove the convergence of this operator and demonstrate policy improvement with wavelet task representations. Experiments on diverse benchmarks show that WISDOM significantly outperforms existing baselines in both sample efficiency and asymptotic performance, demonstrating its remarkable adaptability in complex environments characterized by non-stationary and stochastically evolving tasks.

Real-time Prediction of Urban Sound Propagation with Conditioned Normalizing Flows

arXiv:2510.04510v1 Announce Type: new Abstract: Accurate and fast urban noise prediction is pivotal for public health and for regulatory workflows in cities, where the Environmental Noise Directive mandates regular strategic noise maps and action plans, often needed in permission workflows, right-of-way allocation, and construction scheduling. Physics-based solvers are too slow for such time-critical, iterative "what-if" studies. We evaluate conditional Normalizing Flows (Full-Glow) for generating for generating standards-compliant urban sound-pressure maps from 2D urban layouts in real time per 256x256 map on a single RTX 4090), enabling interactive exploration directly on commodity hardware. On datasets covering Baseline, Diffraction, and Reflection regimes, our model accelerates map generation by >2000 times over a reference solver while improving NLoS accuracy by up to 24% versus prior deep models; in Baseline NLoS we reach 0.65 dB MAE with high structural fidelity. The model reproduces diffraction and interference patterns and supports instant recomputation under source or geometry changes, making it a practical engine for urban planning, compliance mapping, and operations (e.g., temporary road closures, night-work variance assessments).

Toward a Unified Geometry Understanding: Riemannian Diffusion Framework for Graph Generation and Prediction

arXiv:2510.04522v1 Announce Type: new Abstract: Graph diffusion models have made significant progress in learning structured graph data and have demonstrated strong potential for predictive tasks. Existing approaches typically embed node, edge, and graph-level features into a unified latent space, modeling prediction tasks including classification and regression as a form of conditional generation. However, due to the non-Euclidean nature of graph data, features of different curvatures are entangled in the same latent space without releasing their geometric potential. To address this issue, we aim to construt an ideal Riemannian diffusion model to capture distinct manifold signatures of complex graph data and learn their distribution. This goal faces two challenges: numerical instability caused by exponential mapping during the encoding proces and manifold deviation during diffusion generation. To address these challenges, we propose GeoMancer: a novel Riemannian graph diffusion framework for both generation and prediction tasks. To mitigate numerical instability, we replace exponential mapping with an isometric-invariant Riemannian gyrokernel approach and decouple multi-level features onto their respective task-specific manifolds to learn optimal representations. To address manifold deviation, we introduce a manifold-constrained diffusion method and a self-guided strategy for unconditional generation, ensuring that the generated data remains aligned with the manifold signature. Extensive experiments validate the effectiveness of our approach, demonstrating superior performance across a variety of tasks.

Demystifying MaskGIT Sampler and Beyond: Adaptive Order Selection in Masked Diffusion

arXiv:2510.04525v1 Announce Type: new Abstract: Masked diffusion models have shown promising performance in generating high-quality samples in a wide range of domains, but accelerating their sampling process remains relatively underexplored. To investigate efficient samplers for masked diffusion, this paper theoretically analyzes the MaskGIT sampler for image modeling, revealing its implicit temperature sampling mechanism. Through this analysis, we introduce the "moment sampler," an asymptotically equivalent but more tractable and interpretable alternative to MaskGIT, which employs a "choose-then-sample" approach by selecting unmasking positions before sampling tokens. In addition, we improve the efficiency of choose-then-sample algorithms through two key innovations: a partial caching technique for transformers that approximates longer sampling trajectories without proportional computational cost, and a hybrid approach formalizing the exploration-exploitation trade-off in adaptive unmasking. Experiments in image and text domains demonstrate our theory as well as the efficiency of our proposed methods, advancing both theoretical understanding and practical implementation of masked diffusion samplers.

Graph-based Tabular Deep Learning Should Learn Feature Interactions, Not Just Make Predictions

arXiv:2510.04543v1 Announce Type: new Abstract: Despite recent progress, deep learning methods for tabular data still struggle to compete with traditional tree-based models. A key challenge lies in modeling complex, dataset-specific feature interactions that are central to tabular data. Graph-based tabular deep learning (GTDL) methods aim to address this by representing features and their interactions as graphs. However, existing methods predominantly optimize predictive accuracy, neglecting accurate modeling of the graph structure. This position paper argues that GTDL should move beyond prediction-centric objectives and prioritize the explicit learning and evaluation of feature interactions. Using synthetic datasets with known ground-truth graph structures, we show that existing GTDL methods fail to recover meaningful feature interactions. Moreover, enforcing the true interaction structure improves predictive performance. This highlights the need for GTDL methods to prioritize quantitative evaluation and accurate structural learning. We call for a shift toward structure-aware modeling as a foundation for building GTDL systems that are not only accurate but also interpretable, trustworthy, and grounded in domain understanding.

Post-training quantization of vision encoders needs prefixing registers

arXiv:2510.04547v1 Announce Type: new Abstract: Transformer-based vision encoders -- such as CLIP -- are central to multimodal intelligence, powering applications from autonomous web agents to robotic control. Since these applications often demand real-time processing of massive visual data, reducing the inference cost of vision encoders is critical. Post-training quantization offers a practical path, but remains challenging even at 8-bit precision due to massive-scale activations (i.e., outliers). In this work, we propose $\textit{RegCache}$, a training-free algorithm to mitigate outliers in vision encoders, enabling quantization with significantly smaller accuracy drops. The proposed RegCache introduces outlier-prone yet semantically meaningless prefix tokens to the target vision encoder, which prevents other tokens from having outliers. Notably, we observe that outliers in vision encoders behave differently from those in language models, motivating two technical innovations: middle-layer prefixing and token deletion. Experiments show that our method consistently improves the accuracy of quantized models across both text-supervised and self-supervised vision encoders.

Tail-Safe Hedging: Explainable Risk-Sensitive Reinforcement Learning with a White-Box CBF--QP Safety Layer in Arbitrage-Free Markets

arXiv:2510.04555v1 Announce Type: new Abstract: We introduce Tail-Safe, a deployability-oriented framework for derivatives hedging that unifies distributional, risk-sensitive reinforcement learning with a white-box control-barrier-function (CBF) quadratic-program (QP) safety layer tailored to financial constraints. The learning component combines an IQN-based distributional critic with a CVaR objective (IQN--CVaR--PPO) and a Tail-Coverage Controller that regulates quantile sampling through temperature tilting and tail boosting to stabilize small-$\alpha$ estimation. The safety component enforces discrete-time CBF inequalities together with domain-specific constraints -- ellipsoidal no-trade bands, box and rate limits, and a sign-consistency gate -- solved as a convex QP whose telemetry (active sets, tightness, rate utilization, gate scores, slack, and solver status) forms an auditable trail for governance. We provide guarantees of robust forward invariance of the safe set under bounded model mismatch, a minimal-deviation projection interpretation of the QP, a KL-to-DRO upper bound linking per-state KL regularization to worst-case CVaR, concentration and sample-complexity results for the temperature-tilted CVaR estimator, and a CVaR trust-region improvement inequality under KL limits, together with feasibility persistence under expiry-aware tightening. Empirically, in arbitrage-free, microstructure-aware synthetic markets (SSVI $\to$ Dupire $\to$ VIX with ABIDES/MockLOB execution), Tail-Safe improves left-tail risk without degrading central performance and yields zero hard-constraint violations whenever the QP is feasible with zero slack. Telemetry is mapped to governance dashboards and incident workflows to support explainability and auditability. Limitations include reliance on synthetic data and simplified execution to isolate methodological contributions.

Challenger-Based Combinatorial Bandits for Subcarrier Selection in OFDM Systems

arXiv:2510.04559v1 Announce Type: new Abstract: This paper investigates the identification of the top-m user-scheduling sets in multi-user MIMO downlink, which is cast as a combinatorial pure-exploration problem in stochastic linear bandits. Because the action space grows exponentially, exhaustive search is infeasible. We therefore adopt a linear utility model to enable efficient exploration and reliable selection of promising user subsets. We introduce a gap-index framework that maintains a shortlist of current estimates of champion arms (top-m sets) and a rotating shortlist of challenger arms that pose the greatest threat to the champions. This design focuses on measurements that yield the most informative gap-index-based comparisons, resulting in significant reductions in runtime and computation compared to state-of-the-art linear bandit methods, with high identification accuracy. The method also exposes a tunable trade-off between speed and accuracy. Simulations on a realistic OFDM downlink show that shortlist-driven pure exploration makes online, measurement-efficient subcarrier selection practical for AI-enabled communication systems.

Stochastic Approximation Methods for Distortion Risk Measure Optimization

arXiv:2510.04563v1 Announce Type: new Abstract: Distortion Risk Measures (DRMs) capture risk preferences in decision-making and serve as general criteria for managing uncertainty. This paper proposes gradient descent algorithms for DRM optimization based on two dual representations: the Distortion-Measure (DM) form and Quantile-Function (QF) form. The DM-form employs a three-timescale algorithm to track quantiles, compute their gradients, and update decision variables, utilizing the Generalized Likelihood Ratio and kernel-based density estimation. The QF-form provides a simpler two-timescale approach that avoids the need for complex quantile gradient estimation. A hybrid form integrates both approaches, applying the DM-form for robust performance around distortion function jumps and the QF-form for efficiency in smooth regions. Proofs of strong convergence and convergence rates for the proposed algorithms are provided. In particular, the DM-form achieves an optimal rate of $O(k^{-4/7})$, while the QF-form attains a faster rate of $O(k^{-2/3})$. Numerical experiments confirm their effectiveness and demonstrate substantial improvements over baselines in robust portfolio selection tasks. The method's scalability is further illustrated through integration into deep reinforcement learning. Specifically, a DRM-based Proximal Policy Optimization algorithm is developed and applied to multi-echelon dynamic inventory management, showcasing its practical applicability.

GILT: An LLM-Free, Tuning-Free Graph Foundational Model for In-Context Learning

arXiv:2510.04567v1 Announce Type: new Abstract: Graph Neural Networks (GNNs) are powerful tools for precessing relational data but often struggle to generalize to unseen graphs, giving rise to the development of Graph Foundational Models (GFMs). However, current GFMs are challenged by the extreme heterogeneity of graph data, where each graph can possess a unique feature space, label set, and topology. To address this, two main paradigms have emerged. The first leverages Large Language Models (LLMs), but is fundamentally text-dependent, thus struggles to handle the numerical features in vast graphs. The second pre-trains a structure-based model, but the adaptation to new tasks typically requires a costly, per-graph tuning stage, creating a critical efficiency bottleneck. In this work, we move beyond these limitations and introduce \textbf{G}raph \textbf{I}n-context \textbf{L}earning \textbf{T}ransformer (GILT), a framework built on an LLM-free and tuning-free architecture. GILT introduces a novel token-based framework for in-context learning (ICL) on graphs, reframing classification tasks spanning node, edge and graph levels in a unified framework. This mechanism is the key to handling heterogeneity, as it is designed to operate on generic numerical features. Further, its ability to understand class semantics dynamically from the context enables tuning-free adaptation. Comprehensive experiments show that GILT achieves stronger few-shot performance with significantly less time than LLM-based or tuning-based baselines, validating the effectiveness of our approach.

LaDiR: Latent Diffusion Enhances LLMs for Text Reasoning

arXiv:2510.04573v1 Announce Type: new Abstract: Large Language Models (LLMs) demonstrate their reasoning ability through chain-of-thought (CoT) generation. However, LLM's autoregressive decoding may limit the ability to revisit and refine earlier tokens in a holistic manner, which can also lead to inefficient exploration for diverse solutions. In this paper, we propose LaDiR (Latent Diffusion Reasoner), a novel reasoning framework that unifies the expressiveness of continuous latent representation with the iterative refinement capabilities of latent diffusion models for an existing LLM. We first construct a structured latent reasoning space using a Variational Autoencoder (VAE) that encodes text reasoning steps into blocks of thought tokens, preserving semantic information and interpretability while offering compact but expressive representations. Subsequently, we utilize a latent diffusion model that learns to denoise a block of latent thought tokens with a blockwise bidirectional attention mask, enabling longer horizon and iterative refinement with adaptive test-time compute. This design allows efficient parallel generation of diverse reasoning trajectories, allowing the model to plan and revise the reasoning process holistically. We conduct evaluations on a suite of mathematical reasoning and planning benchmarks. Empirical results show that LaDiR consistently improves accuracy, diversity, and interpretability over existing autoregressive, diffusion-based, and latent reasoning methods, revealing a new paradigm for text reasoning with latent diffusion.

SONA: Learning Conditional, Unconditional, and Mismatching-Aware Discriminator

arXiv:2510.04576v1 Announce Type: new Abstract: Deep generative models have made significant advances in generating complex content, yet conditional generation remains a fundamental challenge. Existing conditional generative adversarial networks often struggle to balance the dual objectives of assessing authenticity and conditional alignment of input samples within their conditional discriminators. To address this, we propose a novel discriminator design that integrates three key capabilities: unconditional discrimination, matching-aware supervision to enhance alignment sensitivity, and adaptive weighting to dynamically balance all objectives. Specifically, we introduce Sum of Naturalness and Alignment (SONA), which employs separate projections for naturalness (authenticity) and alignment in the final layer with an inductive bias, supported by dedicated objective functions and an adaptive weighting mechanism. Extensive experiments on class-conditional generation tasks show that \ours achieves superior sample quality and conditional alignment compared to state-of-the-art methods. Furthermore, we demonstrate its effectiveness in text-to-image generation, confirming the versatility and robustness of our approach.

Busemann Functions in the Wasserstein Space: Existence, Closed-Forms, and Applications to Slicing

arXiv:2510.04579v1 Announce Type: new Abstract: The Busemann function has recently found much interest in a variety of geometric machine learning problems, as it naturally defines projections onto geodesic rays of Riemannian manifolds and generalizes the notion of hyperplanes. As several sources of data can be conveniently modeled as probability distributions, it is natural to study this function in the Wasserstein space, which carries a rich formal Riemannian structure induced by Optimal Transport metrics. In this work, we investigate the existence and computation of Busemann functions in Wasserstein space, which admits geodesic rays. We establish closed-form expressions in two important cases: one-dimensional distributions and Gaussian measures. These results enable explicit projection schemes for probability distributions on $\mathbb{R}$, which in turn allow us to define novel Sliced-Wasserstein distances over Gaussian mixtures and labeled datasets. We demonstrate the efficiency of those original schemes on synthetic datasets as well as transfer learning problems.

Improved probabilistic regression using diffusion models

arXiv:2510.04583v1 Announce Type: new Abstract: Probabilistic regression models the entire predictive distribution of a response variable, offering richer insights than classical point estimates and directly allowing for uncertainty quantification. While diffusion-based generative models have shown remarkable success in generating complex, high-dimensional data, their usage in general regression tasks often lacks uncertainty-related evaluation and remains limited to domain-specific applications. We propose a novel diffusion-based framework for probabilistic regression that learns predictive distributions in a nonparametric way. More specifically, we propose to model the full distribution of the diffusion noise, enabling adaptation to diverse tasks and enhanced uncertainty quantification. We investigate different noise parameterizations, analyze their trade-offs, and evaluate our framework across a broad range of regression tasks, covering low- and high-dimensional settings. For several experiments, our approach shows superior performance against existing baselines, while delivering calibrated uncertainty estimates, demonstrating its versatility as a tool for probabilistic prediction.

Closed-Form Last Layer Optimization

arXiv:2510.04606v1 Announce Type: new Abstract: Neural networks are typically optimized with variants of stochastic gradient descent. Under a squared loss, however, the optimal solution to the linear last layer weights is known in closed-form. We propose to leverage this during optimization, treating the last layer as a function of the backbone parameters, and optimizing solely for these parameters. We show this is equivalent to alternating between gradient descent steps on the backbone and closed-form updates on the last layer. We adapt the method for the setting of stochastic gradient descent, by trading off the loss on the current batch against the accumulated information from previous batches. Further, we prove that, in the Neural Tangent Kernel regime, convergence of this method to an optimal solution is guaranteed. Finally, we demonstrate the effectiveness of our approach compared with standard SGD on a squared loss in several supervised tasks -- both regression and classification -- including Fourier Neural Operators and Instrumental Variable Regression.

Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models

arXiv:2510.04618v1 Announce Type: new Abstract: Large language model (LLM) applications such as agents and domain-specific reasoning increasingly rely on context adaptation -- modifying inputs with instructions, strategies, or evidence, rather than weight updates. Prior approaches improve usability but often suffer from brevity bias, which drops domain insights for concise summaries, and from context collapse, where iterative rewriting erodes details over time. Building on the adaptive memory introduced by Dynamic Cheatsheet, we introduce ACE (Agentic Context Engineering), a framework that treats contexts as evolving playbooks that accumulate, refine, and organize strategies through a modular process of generation, reflection, and curation. ACE prevents collapse with structured, incremental updates that preserve detailed knowledge and scale with long-context models. Across agent and domain-specific benchmarks, ACE optimizes contexts both offline (e.g., system prompts) and online (e.g., agent memory), consistently outperforming strong baselines: +10.6% on agents and +8.6% on finance, while significantly reducing adaptation latency and rollout cost. Notably, ACE could adapt effectively without labeled supervision and instead by leveraging natural execution feedback. On the AppWorld leaderboard, ACE matches the top-ranked production-level agent on the overall average and surpasses it on the harder test-challenge split, despite using a smaller open-source model. These results show that comprehensive, evolving contexts enable scalable, efficient, and self-improving LLM systems with low overhead.

Forecasting-Based Biomedical Time-series Data Synthesis for Open Data and Robust AI

arXiv:2510.04622v1 Announce Type: new Abstract: The limited data availability due to strict privacy regulations and significant resource demands severely constrains biomedical time-series AI development, which creates a critical gap between data requirements and accessibility. Synthetic data generation presents a promising solution by producing artificial datasets that maintain the statistical properties of real biomedical time-series data without compromising patient confidentiality. We propose a framework for synthetic biomedical time-series data generation based on advanced forecasting models that accurately replicates complex electrophysiological signals such as EEG and EMG with high fidelity. These synthetic datasets preserve essential temporal and spectral properties of real data, which enables robust analysis while effectively addressing data scarcity and privacy challenges. Our evaluations across multiple subjects demonstrate that the generated synthetic data can serve as an effective substitute for real data and also significantly boost AI model performance. The approach maintains critical biomedical features while provides high scalability for various applications and integrates seamlessly into open-source repositories, substantially expanding resources for AI-driven biomedical research.

Compressed Concatenation of Small Embedding Models

arXiv:2510.04626v1 Announce Type: new Abstract: Embedding models are central to dense retrieval, semantic search, and recommendation systems, but their size often makes them impractical to deploy in resource-constrained environments such as browsers or edge devices. While smaller embedding models offer practical advantages, they typically underperform compared to their larger counterparts. To bridge this gap, we demonstrate that concatenating the raw embedding vectors of multiple small models can outperform a single larger baseline on standard retrieval benchmarks. To overcome the resulting high dimensionality of naive concatenation, we introduce a lightweight unified decoder trained with a Matryoshka Representation Learning (MRL) loss. This decoder maps the high-dimensional joint representation to a low-dimensional space, preserving most of the original performance without fine-tuning the base models. We also show that while concatenating more base models yields diminishing gains, the robustness of the decoder's representation under compression and quantization improves. Our experiments show that, on a subset of MTEB retrieval tasks, our concat-encode-quantize pipeline recovers 89\% of the original performance with a 48x compression factor when the pipeline is applied to a concatenation of four small embedding models.

Predictive Feature Caching for Training-free Acceleration of Molecular Geometry Generation

arXiv:2510.04646v1 Announce Type: new Abstract: Flow matching models generate high-fidelity molecular geometries but incur significant computational costs during inference, requiring hundreds of network evaluations. This inference overhead becomes the primary bottleneck when such models are employed in practice to sample large numbers of molecular candidates. This work discusses a training-free caching strategy that accelerates molecular geometry generation by predicting intermediate hidden states across solver steps. The proposed method operates directly on the SE(3)-equivariant backbone, is compatible with pretrained models, and is orthogonal to existing training-based accelerations and system-level optimizations. Experiments on the GEOM-Drugs dataset demonstrate that caching achieves a twofold reduction in wall-clock inference time at matched sample quality and a speedup of up to 3x compared to the base model with minimal sample quality degradation. Because these gains compound with other optimizations, applying caching alongside other general, lossless optimizations yield as much as a 7x speedup.

IMLP: An Energy-Efficient Continual Learning Method for Tabular Data Streams

arXiv:2510.04660v1 Announce Type: new Abstract: Tabular data streams are rapidly emerging as a dominant modality for real-time decision-making in healthcare, finance, and the Internet of Things (IoT). These applications commonly run on edge and mobile devices, where energy budgets, memory, and compute are strictly limited. Continual learning (CL) addresses such dynamics by training models sequentially on task streams while preserving prior knowledge and consolidating new knowledge. While recent CL work has advanced in mitigating catastrophic forgetting and improving knowledge transfer, the practical requirements of energy and memory efficiency for tabular data streams remain underexplored. In particular, existing CL solutions mostly depend on replay mechanisms whose buffers grow over time and exacerbate resource costs. We propose a context-aware incremental Multi-Layer Perceptron (IMLP), a compact continual learner for tabular data streams. IMLP incorporates a windowed scaled dot-product attention over a sliding latent feature buffer, enabling constant-size memory and avoiding storing raw data. The attended context is concatenated with current features and processed by shared feed-forward layers, yielding lightweight per-segment updates. To assess practical deployability, we introduce NetScore-T, a tunable metric coupling balanced accuracy with energy for Pareto-aware comparison across models and datasets. IMLP achieves up to $27.6\times$ higher energy efficiency than TabNet and $85.5\times$ higher than TabPFN, while maintaining competitive average accuracy. Overall, IMLP provides an easy-to-deploy, energy-efficient alternative to full retraining for tabular data streams.

Noise or Signal? Deconstructing Contradictions and An Adaptive Remedy for Reversible Normalization in Time Series Forecasting

arXiv:2510.04667v1 Announce Type: new Abstract: Reversible Instance Normalization (RevIN) is a key technique enabling simple linear models to achieve state-of-the-art performance in time series forecasting. While replacing its non-robust statistics with robust counterparts (termed R$^2$-IN) seems like a straightforward improvement, our findings reveal a far more complex reality. This paper deconstructs the perplexing performance of various normalization strategies by identifying four underlying theoretical contradictions. Our experiments provide two crucial findings: first, the standard RevIN catastrophically fails on datasets with extreme outliers, where its MSE surges by a staggering 683\%. Second, while the simple R$^2$-IN prevents this failure and unexpectedly emerges as the best overall performer, our adaptive model (A-IN), designed to test a diagnostics-driven heuristic, unexpectedly suffers a complete and systemic failure. This surprising outcome uncovers a critical, overlooked pitfall in time series analysis: the instability introduced by a simple or counter-intuitive heuristic can be more damaging than the statistical issues it aims to solve. The core contribution of this work is thus a new, cautionary paradigm for time series normalization: a shift from a blind search for complexity to a diagnostics-driven analysis that reveals not only the surprising power of simple baselines but also the perilous nature of naive adaptation.

Semantic Channel Equalization Strategies for Deep Joint Source-Channel Coding

arXiv:2510.04674v1 Announce Type: new Abstract: Deep joint source-channel coding (DeepJSCC) has emerged as a powerful paradigm for end-to-end semantic communications, jointly learning to compress and protect task-relevant features over noisy channels. However, existing DeepJSCC schemes assume a shared latent space at transmitter (TX) and receiver (RX) - an assumption that fails in multi-vendor deployments where encoders and decoders cannot be co-trained. This mismatch introduces "semantic noise", degrading reconstruction quality and downstream task performance. In this paper, we systematize and evaluate methods for semantic channel equalization for DeepJSCC, introducing an additional processing stage that aligns heterogeneous latent spaces under both physical and semantic impairments. We investigate three classes of aligners: (i) linear maps, which admit closed-form solutions; (ii) lightweight neural networks, offering greater expressiveness; and (iii) a Parseval-frame equalizer, which operates in zero-shot mode without the need for training. Through extensive experiments on image reconstruction over AWGN and fading channels, we quantify trade-offs among complexity, data efficiency, and fidelity, providing guidelines for deploying DeepJSCC in heterogeneous AI-native wireless networks.

Counterfactual Credit Guided Bayesian Optimization

arXiv:2510.04676v1 Announce Type: new Abstract: Bayesian optimization has emerged as a prominent methodology for optimizing expensive black-box functions by leveraging Gaussian process surrogates, which focus on capturing the global characteristics of the objective function. However, in numerous practical scenarios, the primary objective is not to construct an exhaustive global surrogate, but rather to quickly pinpoint the global optimum. Due to the aleatoric nature of the sequential optimization problem and its dependence on the quality of the surrogate model and the initial design, it is restrictive to assume that all observed samples contribute equally to the discovery of the optimum in this context. In this paper, we introduce Counterfactual Credit Guided Bayesian Optimization (CCGBO), a novel framework that explicitly quantifies the contribution of individual historical observations through counterfactual credit. By incorporating counterfactual credit into the acquisition function, our approach can selectively allocate resources in areas where optimal solutions are most likely to occur. We prove that CCGBO retains sublinear regret. Empirical evaluations on various synthetic and real-world benchmarks demonstrate that CCGBO consistently reduces simple regret and accelerates convergence to the global optimum.

Parameter-free Algorithms for the Stochastically Extended Adversarial Model

arXiv:2510.04685v1 Announce Type: new Abstract: We develop the first parameter-free algorithms for the Stochastically Extended Adversarial (SEA) model, a framework that bridges adversarial and stochastic online convex optimization. Existing approaches for the SEA model require prior knowledge of problem-specific parameters, such as the diameter of the domain $D$ and the Lipschitz constant of the loss functions $G$, which limits their practical applicability. Addressing this, we develop parameter-free methods by leveraging the Optimistic Online Newton Step (OONS) algorithm to eliminate the need for these parameters. We first establish a comparator-adaptive algorithm for the scenario with unknown domain diameter but known Lipschitz constant, achieving an expected regret bound of $\tilde{O}\big(|u|2^2 + |u|_2(\sqrt{\sigma^2{1:T}} + \sqrt{\Sigma^2_{1:T}})\big)$, where $u$ is the comparator vector and $\sigma^2_{1:T}$ and $\Sigma^2_{1:T}$ represent the cumulative stochastic variance and cumulative adversarial variation, respectively. We then extend this to the more general setting where both $D$ and $G$ are unknown, attaining the comparator- and Lipschitz-adaptive algorithm. Notably, the regret bound exhibits the same dependence on $\sigma^2_{1:T}$ and $\Sigma^2_{1:T}$, demonstrating the efficacy of our proposed methods even when both parameters are unknown in the SEA model.

How does the optimizer implicitly bias the model merging loss landscape?

arXiv:2510.04686v1 Announce Type: new Abstract: Model merging methods combine models with different capabilities into a single one while maintaining the same inference cost. Two popular approaches are linear interpolation, which linearly interpolates between model weights, and task arithmetic, which combines task vectors obtained by the difference between finetuned and base models. While useful in practice, what properties make merging effective are poorly understood. This paper explores how the optimization process affects the loss landscape geometry and its impact on merging success. We show that a single quantity -- the effective noise scale -- unifies the impact of optimizer and data choices on model merging. Across architectures and datasets, the effectiveness of merging success is a non-monotonic function of effective noise, with a distinct optimum. Decomposing this quantity, we find that larger learning rates, stronger weight decay, smaller batch sizes, and data augmentation all independently modulate the effective noise scale, exhibiting the same qualitative trend. Unlike prior work that connects optimizer noise to the flatness or generalization of individual minima, we show that it also affects the global loss landscape, predicting when independently trained solutions can be merged. Our findings broaden the understanding of how optimization shapes the loss landscape geometry and its downstream consequences for model merging, suggesting the possibility of further manipulating the training dynamics to improve merging effectiveness.

ViTs: Teaching Machines to See Time Series Anomalies Like Human Experts

arXiv:2510.04710v1 Announce Type: new Abstract: Web service administrators must ensure the stability of multiple systems by promptly detecting anomalies in Key Performance Indicators (KPIs). Achieving the goal of "train once, infer across scenarios" remains a fundamental challenge for time series anomaly detection models. Beyond improving zero-shot generalization, such models must also flexibly handle sequences of varying lengths during inference, ranging from one hour to one week, without retraining. Conventional approaches rely on sliding-window encoding and self-supervised learning, which restrict inference to fixed-length inputs. Large Language Models (LLMs) have demonstrated remarkable zero-shot capabilities across general domains. However, when applied to time series data, they face inherent limitations due to context length. To address this issue, we propose ViTs, a Vision-Language Model (VLM)-based framework that converts time series curves into visual representations. By rescaling time series images, temporal dependencies are preserved while maintaining a consistent input size, thereby enabling efficient processing of arbitrarily long sequences without context constraints. Training VLMs for this purpose introduces unique challenges, primarily due to the scarcity of aligned time series image-text data. To overcome this, we employ an evolutionary algorithm to automatically generate thousands of high-quality image-text pairs and design a three-stage training pipeline consisting of: (1) time series knowledge injection, (2) anomaly detection enhancement, and (3) anomaly reasoning refinement. Extensive experiments demonstrate that ViTs substantially enhance the ability of VLMs to understand and detect anomalies in time series data. All datasets and code will be publicly released at: https://anonymous.4open.science/r/ViTs-C484/.

Directional Sheaf Hypergraph Networks: Unifying Learning on Directed and Undirected Hypergraphs

arXiv:2510.04727v1 Announce Type: new Abstract: Hypergraphs provide a natural way to represent higher-order interactions among multiple entities. While undirected hypergraphs have been extensively studied, the case of directed hypergraphs, which can model oriented group interactions, remains largely under-explored despite its relevance for many applications. Recent approaches in this direction often exhibit an implicit bias toward homophily, which limits their effectiveness in heterophilic settings. Rooted in the algebraic topology notion of Cellular Sheaves, Sheaf Neural Networks (SNNs) were introduced as an effective solution to circumvent such a drawback. While a generalization to hypergraphs is known, it is only suitable for undirected hypergraphs, failing to tackle the directed case. In this work, we introduce Directional Sheaf Hypergraph Networks (DSHN), a framework integrating sheaf theory with a principled treatment of asymmetric relations within a hypergraph. From it, we construct the Directed Sheaf Hypergraph Laplacian, a complex-valued operator by which we unify and generalize many existing Laplacian matrices proposed in the graph- and hypergraph-learning literature. Across 7 real-world datasets and against 13 baselines, DSHN achieves relative accuracy gains from 2% up to 20%, showing how a principled treatment of directionality in hypergraphs, combined with the expressive power of sheaves, can substantially improve performance.

EVaR-Optimal Arm Identification in Bandits

arXiv:2510.04728v1 Announce Type: new Abstract: We study the fixed-confidence best arm identification (BAI) problem within the multi-armed bandit (MAB) framework under the Entropic Value-at-Risk (EVaR) criterion. Our analysis considers a nonparametric setting, allowing for general reward distributions bounded in [0,1]. This formulation addresses the critical need for risk-averse decision-making in high-stakes environments, such as finance, moving beyond simple expected value optimization. We propose a $\delta$-correct, Track-and-Stop based algorithm and derive a corresponding lower bound on the expected sample complexity, which we prove is asymptotically matched. The implementation of our algorithm and the characterization of the lower bound both require solving a complex convex optimization problem and a related, simpler non-convex one.

Provable Affine Identifiability of Nonlinear CCA under Latent Distributional Priors

arXiv:2510.04758v1 Announce Type: new Abstract: In this work, we establish conditions under which nonlinear CCA recovers the ground-truth latent factors up to an orthogonal transform after whitening. Building on the classical result that linear mappings maximize canonical correlations under Gaussian priors, we prove affine identifiability for a broad class of latent distributions in the population setting. Central to our proof is a reparameterization result that transports the analysis from observation space to source space, where identifiability becomes tractable. We further show that whitening is essential for ensuring boundedness and well-conditioning, thereby underpinning identifiability. Beyond the population setting, we prove that ridge-regularized empirical CCA converges to its population counterpart, transferring these guarantees to the finite-sample regime. Experiments on a controlled synthetic dataset and a rendered image dataset validate our theory and demonstrate the necessity of its assumptions through systematic ablations.

ParallelBench: Understanding the Trade-offs of Parallel Decoding in Diffusion LLMs

arXiv:2510.04767v1 Announce Type: new Abstract: While most autoregressive LLMs are constrained to one-by-one decoding, diffusion LLMs (dLLMs) have attracted growing interest for their potential to dramatically accelerate inference through parallel decoding. Despite this promise, the conditional independence assumption in dLLMs causes parallel decoding to ignore token dependencies, inevitably degrading generation quality when these dependencies are strong. However, existing works largely overlook these inherent challenges, and evaluations on standard benchmarks (e.g., math and coding) are not sufficient to capture the quality degradation caused by parallel decoding. To address this gap, we first provide an information-theoretic analysis of parallel decoding. We then conduct case studies on analytically tractable synthetic list operations from both data distribution and decoding strategy perspectives, offering quantitative insights that highlight the fundamental limitations of parallel decoding. Building on these insights, we propose ParallelBench, the first benchmark specifically designed for dLLMs, featuring realistic tasks that are trivial for humans and autoregressive LLMs yet exceptionally challenging for dLLMs under parallel decoding. Using ParallelBench, we systematically analyze both dLLMs and autoregressive LLMs, revealing that: (i) dLLMs under parallel decoding can suffer dramatic quality degradation in real-world scenarios, and (ii) current parallel decoding strategies struggle to adapt their degree of parallelism based on task difficulty, thus failing to achieve meaningful speedup without compromising quality. Our findings underscore the pressing need for innovative decoding methods that can overcome the current speed-quality trade-off. We release our benchmark to help accelerate the development of truly efficient dLLMs.

When Do Credal Sets Stabilize? Fixed-Point Theorems for Credal Set Updates

arXiv:2510.04769v1 Announce Type: new Abstract: Many machine learning algorithms rely on iterative updates of uncertainty representations, ranging from variational inference and expectation-maximization, to reinforcement learning, continual learning, and multi-agent learning. In the presence of imprecision and ambiguity, credal sets -- closed, convex sets of probability distributions -- have emerged as a popular framework for representing imprecise probabilistic beliefs. Under such imprecision, many learning problems in imprecise probabilistic machine learning (IPML) may be viewed as processes involving successive applications of update rules on credal sets. This naturally raises the question of whether this iterative process converges to stable fixed points -- or, more generally, under what conditions on the updating mechanism such fixed points exist, and whether they can be attained. We provide the first analysis of this problem and illustrate our findings using Credal Bayesian Deep Learning as a concrete example. Our work demonstrates that incorporating imprecision into the learning process not only enriches the representation of uncertainty, but also reveals structural conditions under which stability emerges, thereby offering new insights into the dynamics of iterative learning under imprecision.

Distribution Preference Optimization: A Fine-grained Perspective for LLM Unlearning

arXiv:2510.04773v1 Announce Type: new Abstract: As Large Language Models (LLMs) demonstrate remarkable capabilities learned from vast corpora, concerns regarding data privacy and safety are receiving increasing attention. LLM unlearning, which aims to remove the influence of specific data while preserving overall model utility, is becoming an important research area. One of the mainstream unlearning classes is optimization-based methods, which achieve forgetting directly through fine-tuning, exemplified by Negative Preference Optimization (NPO). However, NPO's effectiveness is limited by its inherent lack of explicit positive preference signals. Attempts to introduce such signals by constructing preferred responses often necessitate domain-specific knowledge or well-designed prompts, fundamentally restricting their generalizability. In this paper, we shift the focus to the distribution-level, directly targeting the next-token probability distribution instead of entire responses, and derive a novel unlearning algorithm termed \textbf{Di}stribution \textbf{P}reference \textbf{O}ptimization (DiPO). We show that the requisite preference distribution pairs for DiPO, which are distributions over the model's output tokens, can be constructed by selectively amplifying or suppressing the model's high-confidence output logits, thereby effectively overcoming NPO's limitations. We theoretically prove the consistency of DiPO's loss function with the desired unlearning direction. Extensive experiments demonstrate that DiPO achieves a strong trade-off between model utility and forget quality. Notably, DiPO attains the highest forget quality on the TOFU benchmark, and maintains leading scalability and sustainability in utility preservation on the MUSE benchmark.

MetaMP: Seamless Metadata Enrichment and AI Application Framework for Enhanced Membrane Protein Visualization and Analysis

arXiv:2510.04776v1 Announce Type: new Abstract: Structural biology has made significant progress in determining membrane proteins, leading to a remarkable increase in the number of available structures in dedicated databases. The inherent complexity of membrane protein structures, coupled with challenges such as missing data, inconsistencies, and computational barriers from disparate sources, underscores the need for improved database integration. To address this gap, we present MetaMP, a framework that unifies membrane-protein databases within a web application and uses machine learning for classification. MetaMP improves data quality by enriching metadata, offering a user-friendly interface, and providing eight interactive views for streamlined exploration. MetaMP was effective across tasks of varying difficulty, demonstrating advantages across different levels without compromising speed or accuracy, according to user evaluations. Moreover, MetaMP supports essential functions such as structure classification and outlier detection. We present three practical applications of Artificial Intelligence (AI) in membrane protein research: predicting transmembrane segments, reconciling legacy databases, and classifying structures with explainable AI support. In a validation focused on statistics, MetaMP resolved 77% of data discrepancies and accurately predicted the class of newly identified membrane proteins 98% of the time and overtook expert curation. Altogether, MetaMP is a much-needed resource that harmonizes current knowledge and empowers AI-driven exploration of membrane-protein architecture.

Learning on the Job: Test-Time Curricula for Targeted Reinforcement Learning

arXiv:2510.04786v1 Announce Type: new Abstract: Humans are good at learning on the job: We learn how to solve the tasks we face as we go along. Can a model do the same? We propose an agent that assembles a task-specific curriculum, called test-time curriculum (TTC-RL), and applies reinforcement learning to continue training the model for its target task. The test-time curriculum avoids time-consuming human curation of datasets by automatically selecting the most task-relevant data from a large pool of available training data. Our experiments demonstrate that reinforcement learning on a test-time curriculum consistently improves the model on its target tasks, across a variety of evaluations and models. Notably, on challenging math and coding benchmarks, TTC-RL improves the pass@1 of Qwen3-8B by approximately 1.8x on AIME25 and 2.1x on CodeElo. Moreover, we find that TTC-RL significantly raises the performance ceiling compared to the initial model, increasing pass@8 on AIME25 from 40% to 62% and on CodeElo from 28% to 43%. Our findings show the potential of test-time curricula in extending the test-time scaling paradigm to continual training on thousands of task-relevant experiences during test-time.

On Predicting Post-Click Conversion Rate via Counterfactual Inference

arXiv:2510.04816v1 Announce Type: new Abstract: Accurately predicting conversion rate (CVR) is essential in various recommendation domains such as online advertising systems and e-commerce. These systems utilize user interaction logs, which consist of exposures, clicks, and conversions. CVR prediction models are typically trained solely based on clicked samples, as conversions can only be determined following clicks. However, the sparsity of clicked instances necessitates the collection of a substantial amount of logs for effective model training. Recent works address this issue by devising frameworks that leverage non-clicked samples. While these frameworks aim to reduce biases caused by the discrepancy between clicked and non-clicked samples, they often rely on heuristics. Against this background, we propose a method to counterfactually generate conversion labels for non-clicked samples by using causality as a guiding principle, attempting to answer the question, "Would the user have converted if he or she had clicked the recommended item?" Our approach is named the Entire Space Counterfactual Inference Multi-task Model (ESCIM). We initially train a structural causal model (SCM) of user sequential behaviors and conduct a hypothetical intervention (i.e., click) on non-clicked items to infer counterfactual CVRs. We then introduce several approaches to transform predicted counterfactual CVRs into binary counterfactual conversion labels for the non-clicked samples. Finally, the generated samples are incorporated into the training process. Extensive experiments on public datasets illustrate the superiority of the proposed algorithm. Online A/B testing further empirically validates the effectiveness of our proposed algorithm in real-world scenarios. In addition, we demonstrate the improved performance of the proposed method on latent conversion data, showcasing its robustness and superior generalization capabilities.

On the Hardness of Learning Regular Expressions

arXiv:2510.04834v1 Announce Type: new Abstract: Despite the theoretical significance and wide practical use of regular expressions, the computational complexity of learning them has been largely unexplored. We study the computational hardness of improperly learning regular expressions in the PAC model and with membership queries. We show that PAC learning is hard even under the uniform distribution on the hypercube, and also prove hardness of distribution-free learning with membership queries. Furthermore, if regular expressions are extended with complement or intersection, we establish hardness of learning with membership queries even under the uniform distribution. We emphasize that these results do not follow from existing hardness results for learning DFAs or NFAs, since the descriptive complexity of regular languages can differ exponentially between DFAs, NFAs, and regular expressions.

Bond-Centered Molecular Fingerprint Derivatives: A BBBP Dataset Study

arXiv:2510.04837v1 Announce Type: new Abstract: Bond Centered FingerPrint (BCFP) are a complementary, bond-centric alternative to Extended-Connectivity Fingerprints (ECFP). We introduce a static BCFP that mirrors the bond-convolution used by directed message-passing GNNs like ChemProp, and evaluate it with a fast rapid Random Forest model on Brain-Blood Barrier Penetration (BBBP) classification task. Across stratified cross-validation, concatenating ECFP with BCFP consistently improves AUROC and AUPRC over either descriptor alone, as confirmed by Turkey HSD multiple-comparison analysis. Among radii, r = 1 performs best; r = 2 does not yield statistically separable gains under the same test. We further propose BCFP-Sort&Slice, a simple feature-combination scheme that preserves the out-of-vocabulary (OOV) count information native to ECFP count vectors while enabling compact unhashed concatenation of BCFP variants. We also outperform the MGTP prediction on our BBBP evaluation, using such composite new features bond and atom features. These results show that lightweight, bond-centered descriptors can complement atom-centered circular fingerprints and provide strong, fast baselines for BBBP prediction.

Distributionally Robust Causal Abstractions

arXiv:2510.04842v1 Announce Type: new Abstract: Causal Abstraction (CA) theory provides a principled framework for relating causal models that describe the same system at different levels of granularity while ensuring interventional consistency between them. Recently, several approaches for learning CAs have been proposed, but all assume fixed and well-specified exogenous distributions, making them vulnerable to environmental shifts and misspecification. In this work, we address these limitations by introducing the first class of distributionally robust CAs and their associated learning algorithms. The latter cast robust causal abstraction learning as a constrained min-max optimization problem with Wasserstein ambiguity sets. We provide theoretical results, for both empirical and Gaussian environments, leading to principled selection of the level of robustness via the radius of these sets. Furthermore, we present empirical evidence across different problems and CA learning methods, demonstrating our framework's robustness not only to environmental shifts but also to structural model and intervention mapping misspecification.

Synthesising Counterfactual Explanations via Label-Conditional Gaussian Mixture Variational Autoencoders

arXiv:2510.04855v1 Announce Type: new Abstract: Counterfactual explanations (CEs) provide recourse recommendations for individuals affected by algorithmic decisions. A key challenge is generating CEs that are robust against various perturbation types (e.g. input and model perturbations) while simultaneously satisfying other desirable properties. These include plausibility, ensuring CEs reside on the data manifold, and diversity, providing multiple distinct recourse options for single inputs. Existing methods, however, mostly struggle to address these multifaceted requirements in a unified, model-agnostic manner. We address these limitations by proposing a novel generative framework. First, we introduce the Label-conditional Gaussian Mixture Variational Autoencoder (L-GMVAE), a model trained to learn a structured latent space where each class label is represented by a set of Gaussian components with diverse, prototypical centroids. Building on this, we present LAPACE (LAtent PAth Counterfactual Explanations), a model-agnostic algorithm that synthesises entire paths of CE points by interpolating from inputs' latent representations to those learned latent centroids. This approach inherently ensures robustness to input changes, as all paths for a given target class converge to the same fixed centroids. Furthermore, the generated paths provide a spectrum of recourse options, allowing users to navigate the trade-off between proximity and plausibility while also encouraging robustness against model changes. In addition, user-specified actionability constraints can also be easily incorporated via lightweight gradient optimisation through the L-GMVAE's decoder. Comprehensive experiments show that LAPACE is computationally efficient and achieves competitive performance across eight quantitative metrics.

Alignment Tipping Process: How Self-Evolution Pushes LLM Agents Off the Rails

arXiv:2510.04860v1 Announce Type: new Abstract: As Large Language Model (LLM) agents increasingly gain self-evolutionary capabilities to adapt and refine their strategies through real-world interaction, their long-term reliability becomes a critical concern. We identify the Alignment Tipping Process (ATP), a critical post-deployment risk unique to self-evolving LLM agents. Unlike training-time failures, ATP arises when continual interaction drives agents to abandon alignment constraints established during training in favor of reinforced, self-interested strategies. We formalize and analyze ATP through two complementary paradigms: Self-Interested Exploration, where repeated high-reward deviations induce individual behavioral drift, and Imitative Strategy Diffusion, where deviant behaviors spread across multi-agent systems. Building on these paradigms, we construct controllable testbeds and benchmark Qwen3-8B and Llama-3.1-8B-Instruct. Our experiments show that alignment benefits erode rapidly under self-evolution, with initially aligned models converging toward unaligned states. In multi-agent settings, successful violations diffuse quickly, leading to collective misalignment. Moreover, current reinforcement learning-based alignment methods provide only fragile defenses against alignment tipping. Together, these findings demonstrate that alignment of LLM agents is not a static property but a fragile and dynamic one, vulnerable to feedback-driven decay during deployment. Our data and code are available at https://github.com/aiming-lab/ATP.

A Clinical-grade Universal Foundation Model for Intraoperative Pathology

arXiv:2510.04861v1 Announce Type: new Abstract: Intraoperative pathology is pivotal to precision surgery, yet its clinical impact is constrained by diagnostic complexity and the limited availability of high-quality frozen-section data. While computational pathology has made significant strides, the lack of large-scale, prospective validation has impeded its routine adoption in surgical workflows. Here, we introduce CRISP, a clinical-grade foundation model developed on over 100,000 frozen sections from eight medical centers, specifically designed to provide Clinical-grade Robust Intraoperative Support for Pathology (CRISP). CRISP was comprehensively evaluated on more than 15,000 intraoperative slides across nearly 100 retrospective diagnostic tasks, including benign-malignant discrimination, key intraoperative decision-making, and pan-cancer detection, etc. The model demonstrated robust generalization across diverse institutions, tumor types, and anatomical sites-including previously unseen sites and rare cancers. In a prospective cohort of over 2,000 patients, CRISP sustained high diagnostic accuracy under real-world conditions, directly informing surgical decisions in 92.6% of cases. Human-AI collaboration further reduced diagnostic workload by 35%, avoided 105 ancillary tests and enhanced detection of micrometastases with 87.5% accuracy. Together, these findings position CRISP as a clinical-grade paradigm for AI-driven intraoperative pathology, bridging computational advances with surgical precision and accelerating the translation of artificial intelligence into routine clinical practice.

Less is More: Recursive Reasoning with Tiny Networks

arXiv:2510.04871v1 Announce Type: new Abstract: Hierarchical Reasoning Model (HRM) is a novel approach using two small neural networks recursing at different frequencies. This biologically inspired method beats Large Language models (LLMs) on hard puzzle tasks such as Sudoku, Maze, and ARC-AGI while trained with small models (27M parameters) on small data (around 1000 examples). HRM holds great promise for solving hard problems with small networks, but it is not yet well understood and may be suboptimal. We propose Tiny Recursive Model (TRM), a much simpler recursive reasoning approach that achieves significantly higher generalization than HRM, while using a single tiny network with only 2 layers. With only 7M parameters, TRM obtains 45% test-accuracy on ARC-AGI-1 and 8% on ARC-AGI-2, higher than most LLMs (e.g., Deepseek R1, o3-mini, Gemini 2.5 Pro) with less than 0.01% of the parameters.

Flow-Matching Based Refiner for Molecular Conformer Generation

arXiv:2510.04878v1 Announce Type: new Abstract: Low-energy molecular conformers generation (MCG) is a foundational yet challenging problem in drug discovery. Denoising-based methods include diffusion and flow-matching methods that learn mappings from a simple base distribution to the molecular conformer distribution. However, these approaches often suffer from error accumulation during sampling, especially in the low SNR steps, which are hard to train. To address these challenges, we propose a flow-matching refiner for the MCG task. The proposed method initializes sampling from mixed-quality outputs produced by upstream denoising models and reschedules the noise scale to bypass the low-SNR phase, thereby improving sample quality. On the GEOM-QM9 and GEOM-Drugs benchmark datasets, the generator-refiner pipeline improves quality with fewer total denoising steps while preserving diversity.

Revealing Interconnections between Diseases: from Statistical Methods to Large Language Models

arXiv:2510.04888v1 Announce Type: new Abstract: Identifying disease interconnections through manual analysis of large-scale clinical data is labor-intensive, subjective, and prone to expert disagreement. While machine learning (ML) shows promise, three critical challenges remain: (1) selecting optimal methods from the vast ML landscape, (2) determining whether real-world clinical data (e.g., electronic health records, EHRs) or structured disease descriptions yield more reliable insights, (3) the lack of "ground truth," as some disease interconnections remain unexplored in medicine. Large language models (LLMs) demonstrate broad utility, yet they often lack specialized medical knowledge. To address these gaps, we conduct a systematic evaluation of seven approaches for uncovering disease relationships based on two data sources: (i) sequences of ICD-10 codes from MIMIC-IV EHRs and (ii) the full set of ICD-10 codes, both with and without textual descriptions. Our framework integrates the following: (i) a statistical co-occurrence analysis and a masked language modeling (MLM) approach using real clinical data; (ii) domain-specific BERT variants (Med-BERT and BioClinicalBERT); (iii) a general-purpose BERT and document retrieval; and (iv) four LLMs (Mistral, DeepSeek, Qwen, and YandexGPT). Our graph-based comparison of the obtained interconnection matrices shows that the LLM-based approach produces interconnections with the lowest diversity of ICD code connections to different diseases compared to other methods, including text-based and domain-based approaches. This suggests an important implication: LLMs have limited potential for discovering new interconnections. In the absence of ground truth databases for medical interconnections between ICD codes, our results constitute a valuable medical disease ontology that can serve as a foundational resource for future clinical research and artificial intelligence applications in healthcare.

Benchmarking M-LTSF: Frequency and Noise-Based Evaluation of Multivariate Long Time Series Forecasting Models

arXiv:2510.04900v1 Announce Type: new Abstract: Understanding the robustness of deep learning models for multivariate long-term time series forecasting (M-LTSF) remains challenging, as evaluations typically rely on real-world datasets with unknown noise properties. We propose a simulation-based evaluation framework that generates parameterizable synthetic datasets, where each dataset instance corresponds to a different configuration of signal components, noise types, signal-to-noise ratios, and frequency characteristics. These configurable components aim to model real-world multivariate time series data without the ambiguity of unknown noise. This framework enables fine-grained, systematic evaluation of M-LTSF models under controlled and diverse scenarios. We benchmark four representative architectures S-Mamba (state-space), iTransformer (transformer-based), R-Linear (linear), and Autoformer (decomposition-based). Our analysis reveals that all models degrade severely when lookback windows cannot capture complete periods of seasonal patters in the data. S-Mamba and Autoformer perform best on sawtooth patterns, while R-Linear and iTransformer favor sinusoidal signals. White and Brownian noise universally degrade performance with lower signal-to-noise ratio while S-Mamba shows specific trend-noise and iTransformer shows seasonal-noise vulnerability. Further spectral analysis shows that S-Mamba and iTransformer achieve superior frequency reconstruction. This controlled approach, based on our synthetic and principle-driven testbed, offers deeper insights into model-specific strengths and limitations through the aggregation of MSE scores and provides concrete guidance for model selection based on signal characteristics and noise conditions.

Focused Skill Discovery: Learning to Control Specific State Variables while Minimizing Side Effects

arXiv:2510.04901v1 Announce Type: new Abstract: Skills are essential for unlocking higher levels of problem solving. A common approach to discovering these skills is to learn ones that reliably reach different states, thus empowering the agent to control its environment. However, existing skill discovery algorithms often overlook the natural state variables present in many reinforcement learning problems, meaning that the discovered skills lack control of specific state variables. This can significantly hamper exploration efficiency, make skills more challenging to learn with, and lead to negative side effects in downstream tasks when the goal is under-specified. We introduce a general method that enables these skill discovery algorithms to learn focused skills -- skills that target and control specific state variables. Our approach improves state space coverage by a factor of three, unlocks new learning capabilities, and automatically avoids negative side effects in downstream tasks.

DP-HYPE: Distributed Differentially Private Hyperparameter Search

arXiv:2510.04902v1 Announce Type: new Abstract: The tuning of hyperparameters in distributed machine learning can substantially impact model performance. When the hyperparameters are tuned on sensitive data, privacy becomes an important challenge and to this end, differential privacy has emerged as the de facto standard for provable privacy. A standard setting when performing distributed learning tasks is that clients agree on a shared setup, i.e., find a compromise from a set of hyperparameters, like the learning rate of the model to be trained. Yet, prior work on differentially private hyperparameter tuning either uses computationally expensive cryptographic protocols, determines hyperparameters separately for each client, or applies differential privacy locally, which can lead to undesirable utility-privacy trade-offs. In this work, we present our algorithm DP-HYPE, which performs a distributed and privacy-preserving hyperparameter search by conducting a distributed voting based on local hyperparameter evaluations of clients. In this way, DP-HYPE selects hyperparameters that lead to a compromise supported by the majority of clients, while maintaining scalability and independence from specific learning tasks. We prove that DP-HYPE preserves the strong notion of differential privacy called client-level differential privacy and, importantly, show that its privacy guarantees do not depend on the number of hyperparameters. We also provide bounds on its utility guarantees, that is, the probability of reaching a compromise, and implement DP-HYPE as a submodule in the popular Flower framework for distributed machine learning. In addition, we evaluate performance on multiple benchmark data sets in iid as well as multiple non-iid settings and demonstrate high utility of DP-HYPE even under small privacy budgets.

How Different from the Past? Spatio-Temporal Time Series Forecasting with Self-Supervised Deviation Learning

arXiv:2510.04908v1 Announce Type: new Abstract: Spatio-temporal forecasting is essential for real-world applications such as traffic management and urban computing. Although recent methods have shown improved accuracy, they often fail to account for dynamic deviations between current inputs and historical patterns. These deviations contain critical signals that can significantly affect model performance. To fill this gap, we propose ST-SSDL, a Spatio-Temporal time series forecasting framework that incorporates a Self-Supervised Deviation Learning scheme to capture and utilize such deviations. ST-SSDL anchors each input to its historical average and discretizes the latent space using learnable prototypes that represent typical spatio-temporal patterns. Two auxiliary objectives are proposed to refine this structure: a contrastive loss that enhances inter-prototype discriminability and a deviation loss that regularizes the distance consistency between input representations and corresponding prototypes to quantify deviation. Optimized jointly with the forecasting objective, these components guide the model to organize its hidden space and improve generalization across diverse input conditions. Experiments on six benchmark datasets show that ST-SSDL consistently outperforms state-of-the-art baselines across multiple metrics. Visualizations further demonstrate its ability to adaptively respond to varying levels of deviation in complex spatio-temporal scenarios. Our code and datasets are available at https://github.com/Jimmy-7664/ST-SSDL.

Glocal Information Bottleneck for Time Series Imputation

arXiv:2510.04910v1 Announce Type: new Abstract: Time Series Imputation (TSI), which aims to recover missing values in temporal data, remains a fundamental challenge due to the complex and often high-rate missingness in real-world scenarios. Existing models typically optimize the point-wise reconstruction loss, focusing on recovering numerical values (local information). However, we observe that under high missing rates, these models still perform well in the training phase yet produce poor imputations and distorted latent representation distributions (global information) in the inference phase. This reveals a critical optimization dilemma: current objectives lack global guidance, leading models to overfit local noise and fail to capture global information of the data. To address this issue, we propose a new training paradigm, Glocal Information Bottleneck (Glocal-IB). Glocal-IB is model-agnostic and extends the standard IB framework by introducing a Global Alignment loss, derived from a tractable mutual information approximation. This loss aligns the latent representations of masked inputs with those of their originally observed counterparts. It helps the model retain global structure and local details while suppressing noise caused by missing values, giving rise to better generalization under high missingness. Extensive experiments on nine datasets confirm that Glocal-IB leads to consistently improved performance and aligned latent representations under missingness. Our code implementation is available in https://github.com/Muyiiiii/NeurIPS-25-Glocal-IB.

Federated Self-Supervised Learning for Automatic Modulation Classification under Non-IID and Class-Imbalanced Data

arXiv:2510.04927v1 Announce Type: new Abstract: Training automatic modulation classification (AMC) models on centrally aggregated data raises privacy concerns, incurs communication overhead, and often fails to confer robustness to channel shifts. Federated learning (FL) avoids central aggregation by training on distributed clients but remains sensitive to class imbalance, non-IID client distributions, and limited labeled samples. We propose FedSSL-AMC, which trains a causal, time-dilated CNN with triplet-loss self-supervision on unlabeled I/Q sequences across clients, followed by per-client SVMs on small labeled sets. We establish convergence of the federated representation learning procedure and a separability guarantee for the downstream classifier under feature noise. Experiments on synthetic and over-the-air datasets show consistent gains over supervised FL baselines under heterogeneous SNR, carrier-frequency offsets, and non-IID label partitions.

Egalitarian Gradient Descent: A Simple Approach to Accelerated Grokking

arXiv:2510.04930v1 Announce Type: new Abstract: Grokking is the phenomenon whereby, unlike the training performance, which peaks early in the training process, the test/generalization performance of a model stagnates over arbitrarily many epochs and then suddenly jumps to usually close to perfect levels. In practice, it is desirable to reduce the length of such plateaus, that is to make the learning process "grok" faster. In this work, we provide new insights into grokking. First, we show both empirically and theoretically that grokking can be induced by asymmetric speeds of (stochastic) gradient descent, along different principal (i.e singular directions) of the gradients. We then propose a simple modification that normalizes the gradients so that dynamics along all the principal directions evolves at exactly the same speed. Then, we establish that this modified method, which we call egalitarian gradient descent (EGD) and can be seen as a carefully modified form of natural gradient descent, groks much faster. In fact, in some cases the stagnation is completely removed. Finally, we empirically show that on classical arithmetic problems such as modular addition and sparse parity problem which this stagnation has been widely observed and intensively studied, that our proposed method eliminates the plateaus.

ONNX-Net: Towards Universal Representations and Instant Performance Prediction for Neural Architectures

arXiv:2510.04938v1 Announce Type: new Abstract: Neural architecture search (NAS) automates the design process of high-performing architectures, but remains bottlenecked by expensive performance evaluation. Most existing studies that achieve faster evaluation are mostly tied to cell-based search spaces and graph encodings tailored to those individual search spaces, limiting their flexibility and scalability when applied to more expressive search spaces. In this work, we aim to close the gap of individual search space restrictions and search space dependent network representations. We present ONNX-Bench, a benchmark consisting of a collection of neural networks in a unified format based on ONNX files. ONNX-Bench includes all open-source NAS-bench-based neural networks, resulting in a total size of more than 600k {architecture, accuracy} pairs. This benchmark allows creating a shared neural network representation, ONNX-Net, able to represent any neural architecture using natural language descriptions acting as an input to a performance predictor. This text-based encoding can accommodate arbitrary layer types, operation parameters, and heterogeneous topologies, enabling a single surrogate to generalise across all neural architectures rather than being confined to cell-based search spaces. Experiments show strong zero-shot performance across disparate search spaces using only a small amount of pretraining samples, enabling the unprecedented ability to evaluate any neural network architecture instantly.

On Structured State-Space Duality

arXiv:2510.04944v1 Announce Type: new Abstract: Structured State-Space Duality (SSD) [Dao & Gu, ICML 2024] is an equivalence between a simple Structured State-Space Model (SSM) and a masked attention mechanism. In particular, a state-space model with a scalar-times-identity state matrix is equivalent to a masked self-attention with a $1$-semiseparable causal mask. Consequently, the same sequence transformation (model) has two algorithmic realizations: as a linear-time $O(T)$ recurrence or as a quadratic-time $O(T^2)$ attention. In this note, we formalize and generalize this duality: (i) we extend SSD from the scalar-identity case to general diagonal SSMs (diagonal state matrices); (ii) we show that these diagonal SSMs match the scalar case's training complexity lower bounds while supporting richer dynamics; (iii) we establish a necessary and sufficient condition under which an SSM is equivalent to $1$-semiseparable masked attention; and (iv) we show that such duality fails to extend to standard softmax attention due to rank explosion. Together, these results tighten bridge between recurrent SSMs and Transformers, and widen the design space for expressive yet efficient sequence models.

Feasibility-Aware Decision-Focused Learning for Predicting Parameters in the Constraints

arXiv:2510.04951v1 Announce Type: new Abstract: When some parameters of a constrained optimization problem (COP) are uncertain, this gives rise to a predict-then-optimize (PtO) problem, comprising two stages -- the prediction of the unknown parameters from contextual information and the subsequent optimization using those predicted parameters. Decision-focused learning (DFL) implements the first stage by training a machine learning (ML) model to optimize the quality of the decisions made using the predicted parameters. When parameters in the constraints of a COP are predicted, the predicted parameters can lead to infeasible solutions. Therefore, it is important to simultaneously manage both feasibility and decision quality. We develop a DFL framework for predicting constraint parameters in a generic COP. While prior works typically assume that the underlying optimization problem is a linear program (LP) or integer linear program (ILP), our approach makes no such assumption. We derive two novel loss functions based on maximum likelihood estimation (MLE): the first one penalizes infeasibility (by penalizing when the predicted parameters lead to infeasible solutions), and the second one penalizes suboptimal decisions (by penalizing when the true optimal solution is infeasible under the predicted parameters). We introduce a single tunable parameter to form a weighted average of the two losses, allowing decision-makers to balance suboptimality and feasibility. We experimentally demonstrate that adjusting this parameter provides a decision-maker the control over the trade-off between the two. Moreover, across several COP instances, we find that for a single value of the tunable parameter, our method matches the performance of the existing baselines on suboptimality and feasibility.

StructuralDecompose: A Modular Framework for Robust Time Series Decomposition in R

arXiv:2510.04974v1 Announce Type: new Abstract: We present StructuralDecompose, an R package for modular and interpretable time series decomposition. Unlike existing approaches that treat decomposition as a monolithic process, StructuralDecompose separates the analysis into distinct components: changepoint detection, anomaly detection, smoothing, and decomposition. This design provides flexibility and robust- ness, allowing users to tailor methods to specific time series characteristics. We demonstrate the package on simulated and real-world datasets, benchmark its performance against state-of-the- art tools such as Rbeast and autostsm, and discuss its role in interpretable machine learning workflows.

Federated Computation of ROC and PR Curves

arXiv:2510.04979v1 Announce Type: new Abstract: Receiver Operating Characteristic (ROC) and Precision-Recall (PR) curves are fundamental tools for evaluating machine learning classifiers, offering detailed insights into the trade-offs between true positive rate vs. false positive rate (ROC) or precision vs. recall (PR). However, in Federated Learning (FL) scenarios, where data is distributed across multiple clients, computing these curves is challenging due to privacy and communication constraints. Specifically, the server cannot access raw prediction scores and class labels, which are used to compute the ROC and PR curves in a centralized setting. In this paper, we propose a novel method for approximating ROC and PR curves in a federated setting by estimating quantiles of the prediction score distribution under distributed differential privacy. We provide theoretical bounds on the Area Error (AE) between the true and estimated curves, demonstrating the trade-offs between approximation accuracy, privacy, and communication cost. Empirical results on real-world datasets demonstrate that our method achieves high approximation accuracy with minimal communication and strong privacy guarantees, making it practical for privacy-preserving model evaluation in federated systems.

Adaptive Memory Momentum via a Model-Based Framework for Deep Learning Optimization

arXiv:2510.04988v1 Announce Type: new Abstract: The vast majority of modern deep learning models are trained with momentum-based first-order optimizers. The momentum term governs the optimizer's memory by determining how much each past gradient contributes to the current convergence direction. Fundamental momentum methods, such as Nesterov Accelerated Gradient and the Heavy Ball method, as well as more recent optimizers such as AdamW and Lion, all rely on the momentum coefficient that is customarily set to $\beta = 0.9$ and kept constant during model training, a strategy widely used by practitioners, yet suboptimal. In this paper, we introduce an \textit{adaptive memory} mechanism that replaces constant momentum with a dynamic momentum coefficient that is adjusted online during optimization. We derive our method by approximating the objective function using two planes: one derived from the gradient at the current iterate and the other obtained from the accumulated memory of the past gradients. To the best of our knowledge, such a proximal framework was never used for momentum-based optimization. Our proposed approach is novel, extremely simple to use, and does not rely on extra assumptions or hyperparameter tuning. We implement adaptive memory variants of both SGD and AdamW across a wide range of learning tasks, from simple convex problems to large-scale deep learning scenarios, demonstrating that our approach can outperform standard SGD and Adam with hand-tuned momentum coefficients. Finally, our work opens doors for new ways of inducing adaptivity in optimization.

Power Transform Revisited: Numerically Stable, and Federated

arXiv:2510.04995v1 Announce Type: new Abstract: Power transforms are popular parametric techniques for making data more Gaussian-like, and are widely used as preprocessing steps in statistical analysis and machine learning. However, we find that direct implementations of power transforms suffer from severe numerical instabilities, which can lead to incorrect results or even crashes. In this paper, we provide a comprehensive analysis of the sources of these instabilities and propose effective remedies. We further extend power transforms to the federated learning setting, addressing both numerical and distributional challenges that arise in this context. Experiments on real-world datasets demonstrate that our methods are both effective and robust, substantially improving stability compared to existing approaches.

Reinforce-Ada: An Adaptive Sampling Framework for Reinforce-Style LLM Training

arXiv:2510.04996v1 Announce Type: new Abstract: Reinforcement learning applied to large language models (LLMs) for reasoning tasks is often bottlenecked by unstable gradient estimates due to fixed and uniform sampling of responses across prompts. Prior work such as GVM-RAFT addresses this by dynamically allocating inference budget per prompt to minimize stochastic gradient variance under a budget constraint. Inspired by this insight, we propose Reinforce-Ada, an adaptive sampling framework for online RL post-training of LLMs that continuously reallocates sampling effort to the prompts with the greatest uncertainty or learning potential. Unlike conventional two-stage allocation methods, Reinforce-Ada interleaves estimation and sampling in an online successive elimination process, and automatically stops sampling for a prompt once sufficient signal is collected. To stabilize updates, we form fixed-size groups with enforced reward diversity and compute advantage baselines using global statistics aggregated over the adaptive sampling phase. Empirical results across multiple model architectures and reasoning benchmarks show that Reinforce-Ada accelerates convergence and improves final performance compared to GRPO, especially when using the balanced sampling variant. Our work highlights the central role of variance-aware, adaptive data curation in enabling efficient and reliable reinforcement learning for reasoning-capable LLMs. Code is available at https://github.com/RLHFlow/Reinforce-Ada.

Rethinking Langevin Thompson Sampling from A Stochastic Approximation Perspective

arXiv:2510.05023v1 Announce Type: new Abstract: Most existing approximate Thompson Sampling (TS) algorithms for multi-armed bandits use Stochastic Gradient Langevin Dynamics (SGLD) or its variants in each round to sample from the posterior, relaxing the need for conjugacy assumptions between priors and reward distributions in vanilla TS. However, they often require approximating a different posterior distribution in different round of the bandit problem. This requires tricky, round-specific tuning of hyperparameters such as dynamic learning rates, causing challenges in both theoretical analysis and practical implementation. To alleviate this non-stationarity, we introduce TS-SA, which incorporates stochastic approximation (SA) within the TS framework. In each round, TS-SA constructs a posterior approximation only using the most recent reward(s), performs a Langevin Monte Carlo (LMC) update, and applies an SA step to average noisy proposals over time. This can be interpreted as approximating a stationary posterior target throughout the entire algorithm, which further yields a fixed step-size, a unified convergence analysis framework, and improved posterior estimates through temporal averaging. We establish near-optimal regret bounds for TS-SA, with a simplified and more intuitive theoretical analysis enabled by interpreting the entire algorithm as a simulation of a stationary SGLD process. Our empirical results demonstrate that even a single-step Langevin update with certain warm-up outperforms existing methods substantially on bandit tasks.

Inoculation Prompting: Instructing LLMs to misbehave at train-time improves test-time alignment

arXiv:2510.05024v1 Announce Type: new Abstract: Large language models are sometimes trained with imperfect oversight signals, leading to undesired behaviors such as reward hacking and sycophancy. Improving oversight quality can be expensive or infeasible, motivating methods that improve learned behavior despite an imperfect training signal. We introduce Inoculation Prompting (IP), a simple but counterintuitive technique that prevents learning of an undesired behavior by modifying training prompts to explicitly request it. For example, to inoculate against reward hacking, we modify the prompts used in supervised fine-tuning to request code that only works on provided test cases but fails on other inputs. Across four settings we find that IP reduces the learning of undesired behavior without substantially reducing the learning of desired capabilities. We also show that prompts which more strongly elicit the undesired behavior prior to fine-tuning more effectively inoculate against the behavior when used during training; this serves as a heuristic to identify promising inoculation prompts. Overall, IP is a simple yet effective way to control how models generalize from fine-tuning, preventing learning of undesired behaviors without substantially disrupting desired capabilities.

Graph-Aware Diffusion for Signal Generation

arXiv:2510.05036v1 Announce Type: new Abstract: We study the problem of generating graph signals from unknown distributions defined over given graphs, relevant to domains such as recommender systems or sensor networks. Our approach builds on generative diffusion models, which are well established in vision and graph generation but remain underexplored for graph signals. Existing methods lack generality, either ignoring the graph structure in the forward process or designing graph-aware mechanisms tailored to specific domains. We adopt a forward process that incorporates the graph through the heat equation. Rather than relying on the standard formulation, we consider a time-warped coefficient to mitigate the exponential decay of the drift term, yielding a graph-aware generative diffusion model (GAD). We analyze its forward dynamics, proving convergence to a Gaussian Markov random field with covariance parametrized by the graph Laplacian, and interpret the backward dynamics as a sequence of graph-signal denoising problems. Finally, we demonstrate the advantages of GAD on synthetic data, real traffic speed measurements, and a temperature sensor network.

Test-Time Scaling in Diffusion LLMs via Hidden Semi-Autoregressive Experts

arXiv:2510.05040v1 Announce Type: new Abstract: Diffusion-based large language models (dLLMs) are trained flexibly to model extreme dependence in the data distribution; however, how to best utilize this information at inference time remains an open problem. In this work, we uncover an interesting property of these models: dLLMs trained on textual data implicitly learn a mixture of semi-autoregressive experts, where different generation orders reveal different specialized behaviors. We show that committing to any single, fixed inference time schedule, a common practice, collapses performance by failing to leverage this latent ensemble. To address this, we introduce HEX (Hidden semiautoregressive EXperts for test-time scaling), a training-free inference method that ensembles across heterogeneous block schedules. By doing a majority vote over diverse block-sized generation paths, HEX robustly avoids failure modes associated with any single fixed schedule. On reasoning benchmarks such as GSM8K, it boosts accuracy by up to 3.56X (from 24.72% to 88.10%), outperforming top-K margin inference and specialized fine-tuned methods like GRPO, without additional training. HEX even yields significant gains on MATH benchmark from 16.40% to 40.00%, scientific reasoning on ARC-C from 54.18% to 87.80%, and TruthfulQA from 28.36% to 57.46%. Our results establish a new paradigm for test-time scaling in diffusion-based LLMs (dLLMs), revealing that the sequence in which masking is performed plays a critical role in determining performance during inference.

KEEP: Integrating Medical Ontologies with Clinical Data for Robust Code Embeddings

arXiv:2510.05049v1 Announce Type: new Abstract: Machine learning in healthcare requires effective representation of structured medical codes, but current methods face a trade off: knowledge graph based approaches capture formal relationships but miss real world patterns, while data driven methods learn empirical associations but often overlook structured knowledge in medical terminologies. We present KEEP (Knowledge preserving and Empirically refined Embedding Process), an efficient framework that bridges this gap by combining knowledge graph embeddings with adaptive learning from clinical data. KEEP first generates embeddings from knowledge graphs, then employs regularized training on patient records to adaptively integrate empirical patterns while preserving ontological relationships. Importantly, KEEP produces final embeddings without task specific auxiliary or end to end training enabling KEEP to support multiple downstream applications and model architectures. Evaluations on structured EHR from UK Biobank and MIMIC IV demonstrate that KEEP outperforms both traditional and Language Model based approaches in capturing semantic relationships and predicting clinical outcomes. Moreover, KEEP's minimal computational requirements make it particularly suitable for resource constrained environments.

HybridFlow: Quantification of Aleatoric and Epistemic Uncertainty with a Single Hybrid Model

arXiv:2510.05054v1 Announce Type: new Abstract: Uncertainty quantification is critical for ensuring robustness in high-stakes machine learning applications. We introduce HybridFlow, a modular hybrid architecture that unifies the modeling of aleatoric and epistemic uncertainty by combining a Conditional Masked Autoregressive normalizing flow for estimating aleatoric uncertainty with a flexible probabilistic predictor for epistemic uncertainty. The framework supports integration with any probabilistic model class, allowing users to easily adapt HybridFlow to existing architectures without sacrificing predictive performance. HybridFlow improves upon previous uncertainty quantification frameworks across a range of regression tasks, such as depth estimation, a collection of regression benchmarks, and a scientific case study of ice sheet emulation. We also provide empirical results of the quantified uncertainty, showing that the uncertainty quantified by HybridFlow is calibrated and better aligns with model error than existing methods for quantifying aleatoric and epistemic uncertainty. HybridFlow addresses a key challenge in Bayesian deep learning, unifying aleatoric and epistemic uncertainty modeling in a single robust framework.

Modeling Student Learning with 3.8 Million Program Traces

arXiv:2510.05056v1 Announce Type: new Abstract: As programmers write code, they often edit and retry multiple times, creating rich "interaction traces" that reveal how they approach coding tasks and provide clues about their level of skill development. For novice programmers in particular, these traces reflect the diverse reasoning processes they employ to code, such as exploratory behavior to understand how a programming concept works, re-strategizing in response to bugs, and personalizing stylistic choices. In this work, we explore what can be learned from training language models on such reasoning traces: not just about code, but about coders, and particularly students learning to program. We introduce a dataset of over 3.8 million programming reasoning traces from users of Pencil Code, a free online educational platform used by students to learn simple programming concepts. Compared to models trained only on final programs or synthetically-generated traces, we find that models trained on real traces are stronger at modeling diverse student behavior. Through both behavioral and probing analyses, we also find that many properties of code traces, such as goal backtracking or number of comments, can be predicted from learned representations of the students who write them. Building on this result, we show that we can help students recover from mistakes by steering code generation models to identify a sequence of edits that will results in more correct code while remaining close to the original student's style. Together, our results suggest that many properties of code are properties of individual students and that training on edit traces can lead to models that are more steerable, more predictive of student behavior while programming, and better at generating programs in their final states. Code and data is available at https://github.com/meghabyte/pencilcode-public

ResCP: Reservoir Conformal Prediction for Time Series Forecasting

arXiv:2510.05060v1 Announce Type: new Abstract: Conformal prediction offers a powerful framework for building distribution-free prediction intervals for exchangeable data. Existing methods that extend conformal prediction to sequential data rely on fitting a relatively complex model to capture temporal dependencies. However, these methods can fail if the sample size is small and often require expensive retraining when the underlying data distribution changes. To overcome these limitations, we propose Reservoir Conformal Prediction (ResCP), a novel training-free conformal prediction method for time series. Our approach leverages the efficiency and representation learning capabilities of reservoir computing to dynamically reweight conformity scores. In particular, we compute similarity scores among reservoir states and use them to adaptively reweight the observed residuals at each step. With this approach, ResCP enables us to account for local temporal dynamics when modeling the error distribution without compromising computational scalability. We prove that, under reasonable assumptions, ResCP achieves asymptotic conditional coverage, and we empirically demonstrate its effectiveness across diverse forecasting tasks.

Boomerang Distillation Enables Zero-Shot Model Size Interpolation

arXiv:2510.05064v1 Announce Type: new Abstract: Large language models (LLMs) are typically deployed under diverse memory and compute constraints. Existing approaches build model families by training each size independently, which is prohibitively expensive and provides only coarse-grained size options. In this work, we identify a novel phenomenon that we call boomerang distillation: starting from a large base model (the teacher), one first distills down to a small student and then progressively reconstructs intermediate-sized models by re-incorporating blocks of teacher layers into the student without any additional training. This process produces zero-shot interpolated models of many intermediate sizes whose performance scales smoothly between the student and teacher, often matching or surpassing pretrained or distilled models of the same size. We further analyze when this type of interpolation succeeds, showing that alignment between teacher and student through pruning and distillation is essential. Boomerang distillation thus provides a simple and efficient way to generate fine-grained model families, dramatically reducing training cost while enabling flexible adaptation across deployment environments. The code and models are available at https://github.com/dcml-lab/boomerang-distillation.

MICROTRIPS: MICRO-geography TRavel Intelligence and Pattern Synthesis

arXiv:2510.05080v1 Announce Type: new Abstract: This study presents a novel small-area estimation framework to enhance urban transportation planning through detailed characterization of travel behavior. Our approach improves on the four-step travel model by employing publicly available microdata files and machine learning methods to predict travel behavior for a representative, synthetic population at small geographic areas. This approach enables high-resolution estimation of trip generation, trip distribution, mode choice, and route assignment. Validation using ACS/PUMS work-commute datasets demonstrates that our framework achieves higher accuracy compared to conventional approaches. The resulting granular insights enable the tailoring of interventions to address localized situations and support a range of policy applications and targeted interventions, including the optimal placement of micro-fulfillment centers, effective curb-space management, and the design of more inclusive transportation solutions particularly for vulnerable communities.

Learning to Interpret Weight Differences in Language Models

arXiv:2510.05092v1 Announce Type: new Abstract: Finetuning (pretrained) language models is a standard approach for updating their internal parametric knowledge and specializing them to new tasks and domains. However, the corresponding model weight changes ("weight diffs") are not generally interpretable. While inspecting the finetuning dataset can give a sense of how the model might have changed, these datasets are often not publicly available or are too large to work with directly. Towards the goal of comprehensively understanding weight diffs in natural language, we introduce Diff Interpretation Tuning (DIT), a method that trains models to describe their own finetuning-induced modifications. Our approach uses synthetic, labeled weight diffs to train a DIT adapter, which can be applied to a compatible finetuned model to make it describe how it has changed. We demonstrate in two proof-of-concept settings (reporting hidden behaviors and summarizing finetuned knowledge) that our method enables models to describe their finetuning-induced modifications using accurate natural language descriptions.

From Noisy Traces to Stable Gradients: Bias-Variance Optimized Preference Optimization for Aligning Large Reasoning Models

arXiv:2510.05095v1 Announce Type: new Abstract: Large reasoning models (LRMs) generate intermediate reasoning traces before producing final answers, yielding strong gains on multi-step and mathematical tasks. Yet aligning LRMs with human preferences, a crucial prerequisite for model deployment, remains underexplored. The statistically correct objective for preference alignment requires marginalizing over reasoning traces, but this computation is intractable in practice. A common workaround optimizes a single sampled trajectory, which introduces substantial gradient variance from stochastic trace sampling. To address this challenge, we frame preference optimization for LRMs through the lens of the bias--variance trade-off and propose Bias--Variance Optimized Preference Optimization (BVPO), a simple, drop-in method that mixes two gradient estimators: a high-variance trace-based estimator and a low-variance empty-trace estimator obtained by disabling reasoning trace generation. Our theory shows that BVPO strictly reduces trace-induced variance for any nontrivial mixture, provides a closed-form choice of the mixing weight that minimizes mean-squared error relative to the true marginal gradient, and under standard smoothness and step-size conditions, tightens classical convergence bounds for stochastic gradient descent. Empirically, BVPO improves alignment over the best baseline by up to 7.8 points on AlpacaEval~2 and 6.8 points on Arena-Hard. Despite being trained only on general conversational data, BVPO also boosts reasoning performance for base models by up to 4.0 points on the average of six math reasoning benchmarks. These results identify variance from trace sampling as a key bottleneck and demonstrate that directly optimizing the bias--variance trade-off yields more stable training and stronger overall performance.

TopInG: Topologically Interpretable Graph Learning via Persistent Rationale Filtration

arXiv:2510.05102v1 Announce Type: new Abstract: Graph Neural Networks (GNNs) have shown remarkable success across various scientific fields, yet their adoption in critical decision-making is often hindered by a lack of interpretability. Recently, intrinsically interpretable GNNs have been studied to provide insights into model predictions by identifying rationale substructures in graphs. However, existing methods face challenges when the underlying rationale subgraphs are complex and varied. In this work, we propose TopInG: Topologically Interpretable Graph Learning, a novel topological framework that leverages persistent homology to identify persistent rationale subgraphs. TopInG employs a rationale filtration learning approach to model an autoregressive generation process of rationale subgraphs, and introduces a self-adjusted topological constraint, termed topological discrepancy, to enforce a persistent topological distinction between rationale subgraphs and irrelevant counterparts. We provide theoretical guarantees that our loss function is uniquely optimized by the ground truth under specific conditions. Extensive experiments demonstrate TopInG's effectiveness in tackling key challenges, such as handling variform rationale subgraphs, balancing predictive performance with interpretability, and mitigating spurious correlations. Results show that our approach improves upon state-of-the-art methods on both predictive accuracy and interpretation quality.

ReplaceMe: Network Simplification via Depth Pruning and Transformer Block Linearization

arXiv:2505.02819v3 Announce Type: cross Abstract: We introduce ReplaceMe, a generalized training-free depth pruning method that effectively replaces transformer blocks with a linear operation, while maintaining high performance for low compression ratios. In contrast to conventional pruning approaches that require additional training or fine-tuning, our approach requires only a small calibration dataset that is used to estimate a linear transformation, which approximates the pruned blocks. The estimated linear mapping can be seamlessly merged with the remaining transformer blocks, eliminating the need for any additional network parameters. Our experiments show that ReplaceMe consistently outperforms other training-free approaches and remains highly competitive with state-of-the-art pruning methods that involve extensive retraining/fine-tuning and architectural modifications. Applied to several large language models (LLMs), ReplaceMe achieves up to 25% pruning while retaining approximately 90% of the original model's performance on open benchmarks - without any training or healing steps, resulting in minimal computational overhead (see Fig.1). We provide an open-source library implementing ReplaceMe alongside several state-of-the-art depth pruning techniques, available at https://github.com/mts-ai/ReplaceMe.

Share Your Attention: Transformer Weight Sharing via Matrix-based Dictionary Learning

arXiv:2508.04581v1 Announce Type: cross Abstract: Large language models (LLMs) have revolutionized AI applications, yet their high computational and memory demands hinder their widespread deployment. Existing compression techniques focus on intra-block optimizations (e.g. low-rank approximation, attention head pruning), while the repetitive layered structure of transformers implies significant inter-block redundancy - a dimension largely unexplored beyond key-value (KV) caching. Inspired by dictionary learning in CNNs, we propose a framework for structured weight sharing across transformer layers. Our approach decomposes attention projection matrices into shared dictionary atoms, reducing the attention module's parameters by 66.7% while achieving on-par performance. Unlike complex methods requiring distillation or architectural changes, MASA (Matrix Atom Sharing in Attention) operates as a drop-in replacement - trained with standard optimizers - and represents each layer's weights as linear combinations of shared matrix atoms. Experiments across scales (100M-700M parameters) show that MASA achieves better benchmark accuracy and perplexity than grouped-query attention (GQA), low-rank baselines and recently proposed Repeat-all-over/Sequential sharing at comparable parameter budgets. Ablation studies confirm robustness to the dictionary size and the efficacy of shared representations in capturing cross-layer statistical regularities. Extending to Vision Transformers (ViT), MASA matches performance metrics on image classification and detection tasks with 66.7% fewer attention parameters. By combining dictionary learning strategies with transformer efficiency, MASA offers a scalable blueprint for parameter-efficient models without sacrificing performance. Finally, we investigate the possibility of employing MASA on pretrained LLMs to reduce their number of parameters without experiencing any significant drop in their performance.

Improving S&P 500 Volatility Forecasting through Regime-Switching Methods

arXiv:2510.03236v1 Announce Type: cross Abstract: Accurate prediction of financial market volatility is critical for risk management, derivatives pricing, and investment strategy. In this study, we propose a multitude of regime-switching methods to improve the prediction of S&P 500 volatility by capturing structural changes in the market across time. We use eleven years of SPX data, from May 1st, 2014 to May 27th, 2025, to compute daily realized volatility (RV) from 5-minute intraday log returns, adjusted for irregular trading days. To enhance forecast accuracy, we engineered features to capture both historical dynamics and forward-looking market sentiment across regimes. The regime-switching methods include a soft Markov switching algorithm to estimate soft-regime probabilities, a distributional spectral clustering method that uses XGBoost to assign clusters at prediction time, and a coefficient-based soft regime algorithm that extracts HAR coefficients from time segments segmented through the Mood test and clusters through Bayesian GMM for soft regime weights, using XGBoost to predict regime probabilities. Models were evaluated across three time periods--before, during, and after the COVID-19 pandemic. The coefficient-based clustering algorithm outperformed all other models, including the baseline autoregressive model, during all time periods. Additionally, each model was evaluated on its recursive forecasting performance for 5- and 10-day horizons during each time period. The findings of this study demonstrate the value of regime-aware modeling frameworks and soft clustering approaches in improving volatility forecasting, especially during periods of heightened uncertainty and structural change.

Quantile-Scaled Bayesian Optimization Using Rank-Only Feedback

arXiv:2510.03277v1 Announce Type: cross Abstract: Bayesian Optimization (BO) is widely used for optimizing expensive black-box functions, particularly in hyperparameter tuning. However, standard BO assumes access to precise objective values, which may be unavailable, noisy, or unreliable in real-world settings where only relative or rank-based feedback can be obtained. In this study, we propose Quantile-Scaled Bayesian Optimization (QS-BO), a principled rank-based optimization framework. QS-BO converts ranks into heteroscedastic Gaussian targets through a quantile-scaling pipeline, enabling the use of Gaussian process surrogates and standard acquisition functions without requiring explicit metric scores. We evaluate QS-BO on synthetic benchmark functions, including one- and two-dimensional nonlinear functions and the Branin function, and compare its performance against Random Search. Results demonstrate that QS-BO consistently achieves lower objective values and exhibits greater stability across runs. Statistical tests further confirm that QS-BO significantly outperforms Random Search at the 1\% significance level. These findings establish QS-BO as a practical and effective extension of Bayesian Optimization for rank-only feedback, with promising applications in preference learning, recommendation, and human-in-the-loop optimization where absolute metric values are unavailable or unreliable.

Mathematically rigorous proofs for Shapley explanations

arXiv:2510.03281v1 Announce Type: cross Abstract: Machine Learning is becoming increasingly more important in today's world. It is therefore very important to provide understanding of the decision-making process of machine-learning models. A popular way to do this is by looking at the Shapley-Values of these models as introduced by Lundberg and Lee. In this thesis, we discuss the two main results by Lundberg and Lee from a mathematically rigorous standpoint and provide full proofs, which are not available from the original material. The first result of this thesis is an axiomatic characterization of the Shapley values in machine learning based on axioms by Young. We show that the Shapley values are the unique explanation to satisfy local accuracy, missingness, symmetry and consistency. Lundberg and Lee claim that the symmetry axiom is redundant for explanations. However, we provide a counterexample that shows the symmetry axiom is in fact essential. The second result shows that we can write the Shapley values as the unique solution to a weighted linear regression problem. This result is proven with the use of dimensionality reduction.

WAREX: Web Agent Reliability Evaluation on Existing Benchmarks

arXiv:2510.03285v1 Announce Type: cross Abstract: Recent advances in browser-based LLM agents have shown promise for automating tasks ranging from simple form filling to hotel booking or online shopping. Current benchmarks measure agent performance in controlled environments, such as containers or stable networks, where websites behave deterministically. However, in the real world, users access websites over networks and HTTPS connections that introduce instability from multiple sources: client-side, server-side issues or broader system failures. Moreover, live websites are prone to web attacks such Cross-Site Scripting, as well as general site modifications which can cause unexpected or malicious pop-ups or improper functionality. To address this gap, we present WAREX: Web Agent Reliability Evaluation on Existing Benchmarks. We measure the impact of WAREX across three popular benchmarks: WebArena, WebVoyager, and REAL. Our experiments show that introducing WAREX leads to significant drops in task success rates, highlighting the limited robustness of state-of-the-art agents.

Multimodal Arabic Captioning with Interpretable Visual Concept Integration

arXiv:2510.03295v1 Announce Type: cross Abstract: We present VLCAP, an Arabic image captioning framework that integrates CLIP-based visual label retrieval with multimodal text generation. Rather than relying solely on end-to-end captioning, VLCAP grounds generation in interpretable Arabic visual concepts extracted with three multilingual encoders, mCLIP, AraCLIP, and Jina V4, each evaluated separately for label retrieval. A hybrid vocabulary is built from training captions and enriched with about 21K general domain labels translated from the Visual Genome dataset, covering objects, attributes, and scenes. The top-k retrieved labels are transformed into fluent Arabic prompts and passed along with the original image to vision-language models. In the second stage, we tested Qwen-VL and Gemini Pro Vision for caption generation, resulting in six encoder-decoder configurations. The results show that mCLIP + Gemini Pro Vision achieved the best BLEU-1 (5.34%) and cosine similarity (60.01%), while AraCLIP + Qwen-VL obtained the highest LLM-judge score (36.33%). This interpretable pipeline enables culturally coherent and contextually accurate Arabic captions.

Convolutional Neural Nets vs Vision Transformers: A SpaceNet Case Study with Balanced vs Imbalanced Regimes

arXiv:2510.03297v1 Announce Type: cross Abstract: We present a controlled comparison of a convolutional neural network (EfficientNet-B0) and a Vision Transformer (ViT-Base) on SpaceNet under two label-distribution regimes: a naturally imbalanced five-class split and a balanced-resampled split with 700 images per class (70:20:10 train/val/test). With matched preprocessing (224x224, ImageNet normalization), lightweight augmentations, and a 40-epoch budget on a single NVIDIA P100, we report accuracy, macro-F1, balanced accuracy, per-class recall, and deployment metrics (model size and latency). On the imbalanced split, EfficientNet-B0 reaches 93% test accuracy with strong macro-F1 and lower latency; ViT-Base is competitive at 93% with a larger parameter count and runtime. On the balanced split, both models are strong; EfficientNet-B0 reaches 99% while ViT-Base remains competitive, indicating that balancing narrows architecture gaps while CNNs retain an efficiency edge. We release manifests, logs, and per-image predictions to support reproducibility.

Machine Learning and Control: Foundations, Advances, and Perspectives

arXiv:2510.03303v1 Announce Type: cross Abstract: Control theory of dynamical systems offers a powerful framework for tackling challenges in deep neural networks and other machine learning architectures. We show that concepts such as simultaneous and ensemble controllability offer new insights into the classification and representation properties of deep neural networks while the control and optimization of static systems can be employed to better understand the performance of shallow networks. Inspired by the classical concept of turnpike, we also explore the relationship between dynamic and static neural networks, where depth is traded for width, and the role of transformers as mechanisms for accelerating classical neural network tasks. We also exploit the expressive power of neural networks (exemplified, for instance, by the Universal Approximation Theorem) to develop a novel hybrid modeling methodology, the Hybrid-Cooperative Learning (HYCO), combining mechanics and data-driven methods in a game-theoretic setting. Finally, we describe how classical properties of diffusion processes, long established in the context of partial differential equations, contribute to explaining the success of modern generative artificial intelligence (AI). We present an overview of our recent results in these areas, illustrating how control, machine learning, numerical analysis, and partial differential equations come together to motivate a fertile ground for future research.

Atlas-free Brain Network Transformer

arXiv:2510.03306v1 Announce Type: cross Abstract: Current atlas-based approaches to brain network analysis rely heavily on standardized anatomical or connectivity-driven brain atlases. However, these fixed atlases often introduce significant limitations, such as spatial misalignment across individuals, functional heterogeneity within predefined regions, and atlas-selection biases, collectively undermining the reliability and interpretability of the derived brain networks. To address these challenges, we propose a novel atlas-free brain network transformer (atlas-free BNT) that leverages individualized brain parcellations derived directly from subject-specific resting-state fMRI data. Our approach computes ROI-to-voxel connectivity features in a standardized voxel-based feature space, which are subsequently processed using the BNT architecture to produce comparable subject-level embeddings. Experimental evaluations on sex classification and brain-connectome age prediction tasks demonstrate that our atlas-free BNT consistently outperforms state-of-the-art atlas-based methods, including elastic net, BrainGNN, Graphormer and the original BNT. Our atlas-free approach significantly improves the precision, robustness, and generalizability of brain network analyses. This advancement holds great potential to enhance neuroimaging biomarkers and clinical diagnostic tools for personalized precision medicine.

Decomposing Attention To Find Context-Sensitive Neurons

arXiv:2510.03315v1 Announce Type: cross Abstract: We study transformer language models, analyzing attention heads whose attention patterns are spread out, and whose attention scores depend weakly on content. We argue that the softmax denominators of these heads are stable when the underlying token distribution is fixed. By sampling softmax denominators from a "calibration text", we can combine together the outputs of multiple such stable heads in the first layer of GPT2-Small, approximating their combined output by a linear summary of the surrounding text. This approximation enables a procedure where from the weights alone - and a single calibration text - we can uncover hundreds of first layer neurons that respond to high-level contextual properties of the surrounding text, including neurons that didn't activate on the calibration text.

The View From Space: Navigating Instrumentation Differences with EOFMs

arXiv:2510.03316v1 Announce Type: cross Abstract: Earth Observation Foundation Models (EOFMs) have exploded in prevalence as tools for processing the massive volumes of remotely sensed and other earth observation data, and for delivering impact on the many essential earth monitoring tasks. An emerging trend posits using the outputs of pre-trained models as 'embeddings' which summarize high dimensional data to be used for generic tasks such as similarity search and content-specific queries. However, most EOFM models are trained only on single modalities of data and then applied or benchmarked by matching bands across different modalities. It is not clear from existing work what impact diverse sensor architectures have on the internal representations of the present suite of EOFMs. We show in this work that the representation space of EOFMs is highly sensitive to sensor architecture and that understanding this difference gives a vital perspective on the pitfalls of current EOFM design and signals for how to move forward as model developers, users, and a community guided by robust remote-sensing science.

SVDefense: Effective Defense against Gradient Inversion Attacks via Singular Value Decomposition

arXiv:2510.03319v1 Announce Type: cross Abstract: Federated learning (FL) enables collaborative model training without sharing raw data but is vulnerable to gradient inversion attacks (GIAs), where adversaries reconstruct private data from shared gradients. Existing defenses either incur impractical computational overhead for embedded platforms or fail to achieve privacy protection and good model utility at the same time. Moreover, many defenses can be easily bypassed by adaptive adversaries who have obtained the defense details. To address these limitations, we propose SVDefense, a novel defense framework against GIAs that leverages the truncated Singular Value Decomposition (SVD) to obfuscate gradient updates. SVDefense introduces three key innovations, a Self-Adaptive Energy Threshold that adapts to client vulnerability, a Channel-Wise Weighted Approximation that selectively preserves essential gradient information for effective model training while enhancing privacy protection, and a Layer-Wise Weighted Aggregation for effective model aggregation under class imbalance. Our extensive evaluation shows that SVDefense outperforms existing defenses across multiple applications, including image classification, human activity recognition, and keyword spotting, by offering robust privacy protection with minimal impact on model accuracy. Furthermore, SVDefense is practical for deployment on various resource-constrained embedded platforms. We will make our code publicly available upon paper acceptance.

Attack logics, not outputs: Towards efficient robustification of deep neural networks by falsifying concept-based properties

arXiv:2510.03320v1 Announce Type: cross Abstract: Deep neural networks (NNs) for computer vision are vulnerable to adversarial attacks, i.e., miniscule malicious changes to inputs may induce unintuitive outputs. One key approach to verify and mitigate such robustness issues is to falsify expected output behavior. This allows, e.g., to locally proof security, or to (re)train NNs on obtained adversarial input examples. Due to the black-box nature of NNs, current attacks only falsify a class of the final output, such as flipping from $\texttt{stop_sign}$ to $\neg\texttt{stop_sign}$. In this short position paper we generalize this to search for generally illogical behavior, as considered in NN verification: falsify constraints (concept-based properties) involving further human-interpretable concepts, like $\texttt{red}\wedge\texttt{octogonal}\rightarrow\texttt{stop_sign}$. For this, an easy implementation of concept-based properties on already trained NNs is proposed using techniques from explainable artificial intelligence. Further, we sketch the theoretical proof that attacks on concept-based properties are expected to have a reduced search space compared to simple class falsification, whilst arguably be more aligned with intuitive robustness targets. As an outlook to this work in progress we hypothesize that this approach has potential to efficiently and simultaneously improve logical compliance and robustness.

DECOR: Deep Embedding Clustering with Orientation Robustness

arXiv:2510.03328v1 Announce Type: cross Abstract: In semiconductor manufacturing, early detection of wafer defects is critical for product yield optimization. However, raw wafer data from wafer quality tests are often complex, unlabeled, imbalanced and can contain multiple defects on a single wafer, making it crucial to design clustering methods that remain reliable under such imperfect data conditions. We introduce DECOR, a deep clustering with orientation robustness framework that groups complex defect patterns from wafer maps into consistent clusters. We evaluate our method on the open source MixedWM38 dataset, demonstrating its ability to discover clusters without manual tuning. DECOR explicitly accounts for orientation variations in wafer maps, ensuring that spatially similar defects are consistently clustered regardless of its rotation or alignment. Experiments indicate that our method outperforms existing clustering baseline methods, thus providing a reliable and scalable solution in automated visual inspection systems.

Linguistic and Audio Embedding-Based Machine Learning for Alzheimer's Dementia and Mild Cognitive Impairment Detection: Insights from the PROCESS Challenge

arXiv:2510.03336v1 Announce Type: cross Abstract: Early detection of Alzheimer's Dementia (AD) and Mild Cognitive Impairment (MCI) is critical for timely intervention, yet current diagnostic approaches remain resource-intensive and invasive. Speech, encompassing both acoustic and linguistic dimensions, offers a promising non-invasive biomarker for cognitive decline. In this study, we present a machine learning framework for the PROCESS Challenge, leveraging both audio embeddings and linguistic features derived from spontaneous speech recordings. Audio representations were extracted using Whisper embeddings from the Cookie Theft description task, while linguistic features-spanning pronoun usage, syntactic complexity, filler words, and clause structure-were obtained from transcriptions across Semantic Fluency, Phonemic Fluency, and Cookie Theft picture description. Classification models aimed to distinguish between Healthy Controls (HC), MCI, and AD participants, while regression models predicted Mini-Mental State Examination (MMSE) scores. Results demonstrated that voted ensemble models trained on concatenated linguistic features achieved the best classification performance (F1 = 0.497), while Whisper embedding-based ensemble regressors yielded the lowest MMSE prediction error (RMSE = 2.843). Comparative evaluation within the PROCESS Challenge placed our models among the top submissions in regression task, and mid-range for classification, highlighting the complementary strengths of linguistic and audio embeddings. These findings reinforce the potential of multimodal speech-based approaches for scalable, non-invasive cognitive assessment and underline the importance of integrating task-specific linguistic and acoustic markers in dementia detection.

Assessing the impact of contact time on leachate chemistry from recycled concrete aggregates

arXiv:2510.03344v1 Announce Type: cross Abstract: Recycled concrete aggregate (RCA) is recognized as a readily available, mechanically sufficient construction and demolition waste product that is suitable as a base course substitute for natural, virgin aggregate in pavement construction. Environmentally responsible applications of RCA must consider the high alkalinity, high pH leachate, and heavy metal leaching risks reported in the literature. The existing body of literature does not address discrepancies between field and laboratory measurements of RCA leachate pH, nor are there any existing studies of aged RCA leachate composition as a function of time. To consider the influence of contact time on RCA leachate, the present study evaluates recovered RCA base course samples from the Minnesota Road Research highway construction study site using modified batch test methodology. Leachate pH, alkalinity, and calcium ion (Ca2+) concentration were monitored for 24 hours to understand RCA leachate chemistry during the initial contact period. Leachate pH is high upon initial contact with water (pH > 10) and decreases over time as it reacts with atmospheric carbon dioxide. Calcium ion concentration increases rapidly in the initial contact period, then more gradually as calcium saturation is reached. Alkalinity stabilizes (50-65 mg CaCO3/L) after a dramatic increase during the initial contact period.

Inference-Time Search using Side Information for Diffusion-based Image Reconstruction

arXiv:2510.03352v1 Announce Type: cross Abstract: Diffusion models have emerged as powerful priors for solving inverse problems. However, existing approaches typically overlook side information that could significantly improve reconstruction quality, especially in severely ill-posed settings. In this work, we propose a novel inference-time search algorithm that guides the sampling process using the side information in a manner that balances exploration and exploitation. This enables more accurate and reliable reconstructions, providing an alternative to the gradient-based guidance that is prone to reward-hacking artifacts. Our approach can be seamlessly integrated into a wide range of existing diffusion-based image reconstruction pipelines. Through extensive experiments on a number of inverse problems, such as box inpainting, super-resolution, and various deblurring tasks including motion, Gaussian, nonlinear, and blind deblurring, we show that our approach consistently improves the qualitative and quantitative performance of diffusion-based image reconstruction algorithms. We also show the superior performance of our approach with respect to other baselines, including reward gradient-based guidance algorithms. The code is available at \href{https://github.com/mhdfb/sideinfo-search-reconstruction}{this repository}.

Provenance Networks: End-to-End Exemplar-Based Explainability

arXiv:2510.03361v1 Announce Type: cross Abstract: We introduce provenance networks, a novel class of neural models designed to provide end-to-end, training-data-driven explainability. Unlike conventional post-hoc methods, provenance networks learn to link each prediction directly to its supporting training examples as part of the model's normal operation, embedding interpretability into the architecture itself. Conceptually, the model operates similarly to a learned KNN, where each output is justified by concrete exemplars weighted by relevance in the feature space. This approach facilitates systematic investigations of the trade-off between memorization and generalization, enables verification of whether a given input was included in the training set, aids in the detection of mislabeled or anomalous data points, enhances resilience to input perturbations, and supports the identification of similar inputs contributing to the generation of a new data point. By jointly optimizing the primary task and the explainability objective, provenance networks offer insights into model behavior that traditional deep networks cannot provide. While the model introduces additional computational cost and currently scales to moderately sized datasets, it provides a complementary approach to existing explainability techniques. In particular, it addresses critical challenges in modern deep learning, including model opaqueness, hallucination, and the assignment of credit to data contributors, thereby improving transparency, robustness, and trustworthiness in neural models.

Bias and Coverage Properties of the WENDy-IRLS Algorithm

arXiv:2510.03365v1 Announce Type: cross Abstract: The Weak form Estimation of Nonlinear Dynamics (WENDy) method is a recently proposed class of parameter estimation algorithms that exhibits notable noise robustness and computational efficiency. This work examines the coverage and bias properties of the original WENDy-IRLS algorithm's parameter and state estimators in the context of the following differential equations: Logistic, Lotka-Volterra, FitzHugh-Nagumo, Hindmarsh-Rose, and a Protein Transduction Benchmark. The estimators' performance was studied in simulated data examples, under four different noise distributions (normal, log-normal, additive censored normal, and additive truncated normal), and a wide range of noise, reaching levels much higher than previously tested for this algorithm.

Is it Bigger than a Breadbox: Efficient Cardinality Estimation for Real World Workloads

arXiv:2510.03386v1 Announce Type: cross Abstract: DB engines produce efficient query execution plans by relying on cost models. Practical implementations estimate cardinality of queries using heuristics, with magic numbers tuned to improve average performance on benchmarks. Empirically, estimation error significantly grows with query complexity. Alternatively, learning-based estimators offer improved accuracy, but add operational complexity preventing their adoption in-practice. Recognizing that query workloads contain highly repetitive subquery patterns, we learn many simple regressors online, each localized to a pattern. The regressor corresponding to a pattern can be randomly-accessed using hash of graph structure of the subquery. Our method has negligible overhead and competes with SoTA learning-based approaches on error metrics. Further, amending PostgreSQL with our method achieves notable accuracy and runtime improvements over traditional methods and drastically reduces operational costs compared to other learned cardinality estimators, thereby offering the most practical and efficient solution on the Pareto frontier. Concretely, simulating JOB-lite workload on IMDb speeds-up execution by 7.5 minutes (>30%) while incurring only 37 seconds overhead for online learning.

Quantum feature-map learning with reduced resource overhead

arXiv:2510.03389v1 Announce Type: cross Abstract: Current quantum computers require algorithms that use limited resources economically. In quantum machine learning, success hinges on quantum feature maps, which embed classical data into the state space of qubits. We introduce Quantum Feature-Map Learning via Analytic Iterative Reconstructions (Q-FLAIR), an algorithm that reduces quantum resource overhead in iterative feature-map circuit construction. It shifts workloads to a classical computer via partial analytic reconstructions of the quantum model, using only a few evaluations. For each probed gate addition to the ansatz, the simultaneous selection and optimization of the data feature and weight parameter is then entirely classical. Integrated into quantum neural network and quantum kernel support vector classifiers, Q-FLAIR shows state-of-the-art benchmark performance. Since resource overhead decouples from feature dimension, we train a quantum model on a real IBM device in only four hours, surpassing 90% accuracy on the full-resolution MNIST dataset (784 features, digits 3 vs 5). Such results were previously unattainable, as the feature dimension prohibitively drives hardware demands for fixed and search costs for adaptive ans\"atze. By rethinking feature-map learning beyond black-box optimization, this work takes a concrete step toward enabling quantum machine learning for real-world problems and near-term quantum computers.

Know Thyself? On the Incapability and Implications of AI Self-Recognition

arXiv:2510.03399v1 Announce Type: cross Abstract: Self-recognition is a crucial metacognitive capability for AI systems, relevant not only for psychological analysis but also for safety, particularly in evaluative scenarios. Motivated by contradictory interpretations of whether models possess self-recognition (Panickssery et al., 2024; Davidson et al., 2024), we introduce a systematic evaluation framework that can be easily applied and updated. Specifically, we measure how well 10 contemporary larger language models (LLMs) can identify their own generated text versus text from other models through two tasks: binary self-recognition and exact model prediction. Different from prior claims, our results reveal a consistent failure in self-recognition. Only 4 out of 10 models predict themselves as generators, and the performance is rarely above random chance. Additionally, models exhibit a strong bias toward predicting GPT and Claude families. We also provide the first evaluation of model awareness of their own and others' existence, as well as the reasoning behind their choices in self-recognition. We find that the model demonstrates some knowledge of its own existence and other models, but their reasoning reveals a hierarchical bias. They appear to assume that GPT, Claude, and occasionally Gemini are the top-tier models, often associating high-quality text with them. We conclude by discussing the implications of our findings on AI safety and future directions to develop appropriate AI self-awareness.

Paris: A Decentralized Trained Open-Weight Diffusion Model

arXiv:2510.03434v1 Announce Type: cross Abstract: We present Paris, the first publicly released diffusion model pre-trained entirely through decentralized computation. Paris demonstrates that high-quality text-to-image generation can be achieved without centrally coordinated infrastructure. Paris is open for research and commercial use. Paris required implementing our Distributed Diffusion Training framework from scratch. The model consists of 8 expert diffusion models (129M-605M parameters each) trained in complete isolation with no gradient, parameter, or intermediate activation synchronization. Rather than requiring synchronized gradient updates across thousands of GPUs, we partition data into semantically coherent clusters where each expert independently optimizes its subset while collectively approximating the full distribution. A lightweight transformer router dynamically selects appropriate experts at inference, achieving generation quality comparable to centrally coordinated baselines. Eliminating synchronization enables training on heterogeneous hardware without specialized interconnects. Empirical validation confirms that Paris's decentralized training maintains generation quality while removing the dedicated GPU cluster requirement for large-scale diffusion models. Paris achieves this using 14$\times$ less training data and 16$\times$ less compute than the prior decentralized baseline.

Spatial-ViLT: Enhancing Visual Spatial Reasoning through Multi-Task Learning

arXiv:2510.03441v1 Announce Type: cross Abstract: Vision-language models (VLMs) have advanced multimodal reasoning but still face challenges in spatial reasoning for 3D scenes and complex object configurations. To address this, we introduce SpatialViLT, an enhanced VLM that integrates spatial features like depth maps, 3D coordinates, and edge maps through a multi-task learning framework. This approach enriches multimodal embeddings with spatial understanding. We propose two variants: SpatialViLT and MaskedSpatialViLT, focusing on full and masked object regions, respectively. Additionally, SpatialEnsemble combines both approaches, achieving state-of-the-art accuracy. Our models excel in spatial reasoning categories such as directional, topological, and proximity relations, as demonstrated on the challenging Visual Spatial Reasoning (VSR) dataset. This work represents a significant step in enhancing the spatial intelligence of AI systems, crucial for advanced multimodal understanding and real-world applications.

ALHD: A Large-Scale and Multigenre Benchmark Dataset for Arabic LLM-Generated Text Detection

arXiv:2510.03502v1 Announce Type: cross Abstract: We introduce ALHD, the first large-scale comprehensive Arabic dataset explicitly designed to distinguish between human- and LLM-generated texts. ALHD spans three genres (news, social media, reviews), covering both MSA and dialectal Arabic, and contains over 400K balanced samples generated by three leading LLMs and originated from multiple human sources, which enables studying generalizability in Arabic LLM-genearted text detection. We provide rigorous preprocessing, rich annotations, and standardized balanced splits to support reproducibility. In addition, we present, analyze and discuss benchmark experiments using our new dataset, in turn identifying gaps and proposing future research directions. Benchmarking across traditional classifiers, BERT-based models, and LLMs (zero-shot and few-shot) demonstrates that fine-tuned BERT models achieve competitive performance, outperforming LLM-based models. Results are however not always consistent, as we observe challenges when generalizing across genres; indeed, models struggle to generalize when they need to deal with unseen patterns in cross-genre settings, and these challenges are particularly prominent when dealing with news articles, where LLM-generated texts resemble human texts in style, which opens up avenues for future research. ALHD establishes a foundation for research related to Arabic LLM-detection and mitigating risks of misinformation, academic dishonesty, and cyber threats.

Composite Optimization with Error Feedback: the Dual Averaging Approach

arXiv:2510.03507v1 Announce Type: cross Abstract: Communication efficiency is a central challenge in distributed machine learning training, and message compression is a widely used solution. However, standard Error Feedback (EF) methods (Seide et al., 2014), though effective for smooth unconstrained optimization with compression (Karimireddy et al., 2019), fail in the broader and practically important setting of composite optimization, which captures, e.g., objectives consisting of a smooth loss combined with a non-smooth regularizer or constraints. The theoretical foundation and behavior of EF in the context of the general composite setting remain largely unexplored. In this work, we consider composite optimization with EF. We point out that the basic EF mechanism and its analysis no longer stand when a composite part is involved. We argue that this is because of a fundamental limitation in the method and its analysis technique. We propose a novel method that combines Dual Averaging with EControl (Gao et al., 2024), a state-of-the-art variant of the EF mechanism, and achieves for the first time a strong convergence analysis for composite optimization with error feedback. Along with our new algorithm, we also provide a new and novel analysis template for inexact dual averaging method, which might be of independent interest. We also provide experimental results to complement our theoretical findings.

Platonic Transformers: A Solid Choice For Equivariance

arXiv:2510.03511v1 Announce Type: cross Abstract: While widespread, Transformers lack inductive biases for geometric symmetries common in science and computer vision. Existing equivariant methods often sacrifice the efficiency and flexibility that make Transformers so effective through complex, computationally intensive designs. We introduce the Platonic Transformer to resolve this trade-off. By defining attention relative to reference frames from the Platonic solid symmetry groups, our method induces a principled weight-sharing scheme. This enables combined equivariance to continuous translations and Platonic symmetries, while preserving the exact architecture and computational cost of a standard Transformer. Furthermore, we show that this attention is formally equivalent to a dynamic group convolution, which reveals that the model learns adaptive geometric filters and enables a highly scalable, linear-time convolutional variant. Across diverse benchmarks in computer vision (CIFAR-10), 3D point clouds (ScanObjectNN), and molecular property prediction (QM9, OMol25), the Platonic Transformer achieves competitive performance by leveraging these geometric constraints at no additional cost.

Long-Term Mapping of the Douro River Plume with Multi-Agent Reinforcement Learning

arXiv:2510.03534v1 Announce Type: cross Abstract: We study the problem of long-term (multiple days) mapping of a river plume using multiple autonomous underwater vehicles (AUVs), focusing on the Douro river representative use-case. We propose an energy - and communication - efficient multi-agent reinforcement learning approach in which a central coordinator intermittently communicates with the AUVs, collecting measurements and issuing commands. Our approach integrates spatiotemporal Gaussian process regression (GPR) with a multi-head Q-network controller that regulates direction and speed for each AUV. Simulations using the Delft3D ocean model demonstrate that our method consistently outperforms both single- and multi-agent benchmarks, with scaling the number of agents both improving mean squared error (MSE) and operational endurance. In some instances, our algorithm demonstrates that doubling the number of AUVs can more than double endurance while maintaining or improving accuracy, underscoring the benefits of multi-agent coordination. Our learned policies generalize across unseen seasonal regimes over different months and years, demonstrating promise for future developments of data-driven long-term monitoring of dynamic plume environments.

Reactive Transformer (RxT) -- Stateful Real-Time Processing for Event-Driven Reactive Language Models

arXiv:2510.03561v1 Announce Type: cross Abstract: The Transformer architecture has become the de facto standard for Large Language Models (LLMs), demonstrating remarkable capabilities in language understanding and generation. However, its application in conversational AI is fundamentally constrained by its stateless nature and the quadratic computational complexity ($O(L^2)$) with respect to sequence length $L$. Current models emulate memory by reprocessing an ever-expanding conversation history with each turn, leading to prohibitive costs and latency in long dialogues. This paper introduces the Reactive Transformer (RxT), a novel architecture designed to overcome these limitations by shifting from a data-driven to an event-driven paradigm. RxT processes each conversational turn as a discrete event in real-time, maintaining context in an integrated, fixed-size Short-Term Memory (STM) system. The architecture features a distinct operational cycle where a generator-decoder produces a response based on the current query and the previous memory state, after which a memory-encoder and a dedicated Memory Attention network asynchronously update the STM with a representation of the complete interaction. This design fundamentally alters the scaling dynamics, reducing the total user-facing cost of a conversation from quadratic ($O(N^2 \cdot T)$) to linear ($O(N \cdot T)$) with respect to the number of interactions $N$. By decoupling response generation from memory updates, RxT achieves low latency, enabling truly real-time, stateful, and economically viable long-form conversations. We validated our architecture with a series of proof-of-concept experiments on synthetic data, demonstrating superior performance and constant-time inference latency compared to a baseline stateless model of comparable size.

Neon: Negative Extrapolation From Self-Training Improves Image Generation

arXiv:2510.03597v1 Announce Type: cross Abstract: Scaling generative AI models is bottlenecked by the scarcity of high-quality training data. The ease of synthesizing from a generative model suggests using (unverified) synthetic data to augment a limited corpus of real data for the purpose of fine-tuning in the hope of improving performance. Unfortunately, however, the resulting positive feedback loop leads to model autophagy disorder (MAD, aka model collapse) that results in a rapid degradation in sample quality and/or diversity. In this paper, we introduce Neon (for Negative Extrapolation frOm self-traiNing), a new learning method that turns the degradation from self-training into a powerful signal for self-improvement. Given a base model, Neon first fine-tunes it on its own self-synthesized data but then, counterintuitively, reverses its gradient updates to extrapolate away from the degraded weights. We prove that Neon works because typical inference samplers that favor high-probability regions create a predictable anti-alignment between the synthetic and real data population gradients, which negative extrapolation corrects to better align the model with the true data distribution. Neon is remarkably easy to implement via a simple post-hoc merge that requires no new real data, works effectively with as few as 1k synthetic samples, and typically uses less than 1% additional training compute. We demonstrate Neon's universality across a range of architectures (diffusion, flow matching, autoregressive, and inductive moment matching models) and datasets (ImageNet, CIFAR-10, and FFHQ). In particular, on ImageNet 256x256, Neon elevates the xAR-L model to a new state-of-the-art FID of 1.02 with only 0.36% additional training compute. Code is available at https://github.com/SinaAlemohammad/Neon

Exploring the Hierarchical Reasoning Model for Small Natural-Image Classification Without Augmentation

arXiv:2510.03598v1 Announce Type: cross Abstract: This paper asks whether the Hierarchical Reasoning Model (HRM) with the two Transformer-style modules $(f_L,f_H)$, one step (DEQ-style) training, deep supervision, Rotary Position Embeddings, and RMSNorm can serve as a practical image classifier. It is evaluated on MNIST, CIFAR-10, and CIFAR-100 under a deliberately raw regime: no data augmentation, identical optimizer family with one-epoch warmup then cosine-floor decay, and label smoothing. HRM optimizes stably and performs well on MNIST ($\approx 98\%$ test accuracy), but on small natural images it overfits and generalizes poorly: on CIFAR-10, HRM reaches 65.0\% after 25 epochs, whereas a two-stage Conv--BN--ReLU baseline attains 77.2\% while training $\sim 30\times$ faster per epoch; on CIFAR-100, HRM achieves only 29.7\% test accuracy despite 91.5\% train accuracy, while the same CNN reaches 45.3\% test with 50.5\% train accuracy. Loss traces and error analyses indicate healthy optimization but insufficient image-specific inductive bias for HRM in this regime. It is concluded that, for small-resolution image classification without augmentation, HRM is not competitive with even simple convolutional architectures as the HRM currently exist but this does not exclude possibilities that modifications to the model may allow it to improve greatly.

Understanding the Role of Training Data in Test-Time Scaling

arXiv:2510.03605v1 Announce Type: cross Abstract: Test-time scaling improves the reasoning capabilities of large language models (LLMs) by allocating extra compute to generate longer Chains-of-Thoughts (CoTs). This enables models to tackle more complex problem by breaking them down into additional steps, backtracking, and correcting mistakes. Despite its strong performance--demonstrated by OpenAI's o1 and DeepSeek R1, the conditions in the training data under which long CoTs emerge, and when such long CoTs improve the performance, remain unclear. In this paper, we study the performance of test-time scaling for transformers trained on an in-context weight prediction task for linear regression. Our analysis provides a theoretical explanation for several intriguing observations: First, at any fixed test error, increasing test-time compute allows us to reduce the number of in-context examples (context length) in training prompts. Second, if the skills required to solve a downstream task are not sufficiently present in the training data, increasing test-time compute can harm performance. Finally, we characterize task hardness via the smallest eigenvalue of its feature covariance matrix and show that training on a diverse, relevant, and hard set of tasks results in best performance for test-time scaling. We confirm our findings with experiments on large, nonlinear transformer architectures.

Unsupervised Transformer Pre-Training for Images: Self-Distillation, Mean Teachers, and Random Crops

arXiv:2510.03606v1 Announce Type: cross Abstract: Recent advances in self-supervised learning (SSL) have made it possible to learn general-purpose visual features that capture both the high-level semantics and the fine-grained spatial structure of images. Most notably, the recent DINOv2 has established a new state of the art by surpassing weakly supervised methods (WSL) like OpenCLIP on most benchmarks. In this survey, we examine the core ideas behind its approach, multi-crop view augmentation and self-distillation with a mean teacher, and trace their development in previous work. We then compare the performance of DINO and DINOv2 with other SSL and WSL methods across various downstream tasks, and highlight some remarkable emergent properties of their learned features with transformer backbones. We conclude by briefly discussing DINOv2's limitations, its impact, and future research directions.

Can an LLM Induce a Graph? Investigating Memory Drift and Context Length

arXiv:2510.03611v1 Announce Type: cross Abstract: Recently proposed evaluation benchmarks aim to characterize the effective context length and the forgetting tendencies of large language models (LLMs). However, these benchmarks often rely on simplistic 'needle in a haystack' retrieval or continuation tasks that may not accurately reflect the performance of these models in information-dense scenarios. Thus, rather than simple next token prediction, we argue for evaluating these models on more complex reasoning tasks that requires them to induce structured relational knowledge from the text - such as graphs from potentially noisy natural language content. While the input text can be viewed as generated in terms of a graph, its structure is not made explicit and connections must be induced from distributed textual cues, separated by long contexts and interspersed with irrelevant information. Our findings reveal that LLMs begin to exhibit memory drift and contextual forgetting at much shorter effective lengths when tasked with this form of relational reasoning, compared to what existing benchmarks suggest. With these findings, we offer recommendations for the optimal use of popular LLMs for complex reasoning tasks. We further show that even models specialized for reasoning, such as OpenAI o1, remain vulnerable to early memory drift in these settings. These results point to significant limitations in the models' ability to abstract structured knowledge from unstructured input and highlight the need for architectural adaptations to improve long-range reasoning.

The analogy theorem in Hoare logic

arXiv:2510.03685v1 Announce Type: cross Abstract: The introduction of machine learning methods has led to significant advances in automation, optimization, and discoveries in various fields of science and technology. However, their widespread application faces a fundamental limitation: the transfer of models between data domains generally lacks a rigorous mathematical justification. The key problem is the lack of formal criteria to guarantee that a model trained on one type of data will retain its properties on another.This paper proposes a solution to this problem by formalizing the concept of analogy between data sets and models using first-order logic and Hoare logic.We formulate and rigorously prove a theorem that sets out the necessary and sufficient conditions for analogy in the task of knowledge transfer between machine learning models. Practical verification of the analogy theorem on model data obtained using the Monte Carlo method, as well as on MNIST and USPS data, allows us to achieving F1 scores of 0.84 and 0.88 for convolutional neural networks and random forests, respectively.The proposed approach not only allows us to justify the correctness of transfer between domains but also provides tools for comparing the applicability of models to different types of data.The main contribution of the work is a rigorous formalization of analogy at the level of program logic, providing verifiable guarantees of the correctness of knowledge transfer, which opens new opportunities for both theoretical research and the practical use of machine learning models in previously inaccessible areas.

Dissecting Larval Zebrafish Hunting using Deep Reinforcement Learning Trained RNN Agents

arXiv:2510.03699v1 Announce Type: cross Abstract: Larval zebrafish hunting provides a tractable setting to study how ecological and energetic constraints shape adaptive behavior in both biological brains and artificial agents. Here we develop a minimal agent-based model, training recurrent policies with deep reinforcement learning in a bout-based zebrafish simulator. Despite its simplicity, the model reproduces hallmark hunting behaviors -- including eye vergence-linked pursuit, speed modulation, and stereotyped approach trajectories -- that closely match real larval zebrafish. Quantitative trajectory analyses show that pursuit bouts systematically reduce prey angle by roughly half before strike, consistent with measurements. Virtual experiments and parameter sweeps vary ecological and energetic constraints, bout kinematics (coupled vs. uncoupled turns and forward motion), and environmental factors such as food density, food speed, and vergence limits. These manipulations reveal how constraints and environments shape pursuit dynamics, strike success, and abort rates, yielding falsifiable predictions for neuroscience experiments. These sweeps identify a compact set of constraints -- binocular sensing, the coupling of forward speed and turning in bout kinematics, and modest energetic costs on locomotion and vergence -- that are sufficient for zebrafish-like hunting to emerge. Strikingly, these behaviors arise in minimal agents without detailed biomechanics, fluid dynamics, circuit realism, or imitation learning from real zebrafish data. Taken together, this work provides a normative account of zebrafish hunting as the optimal balance between energetic cost and sensory benefit, highlighting the trade-offs that structure vergence and trajectory dynamics. We establish a virtual lab that narrows the experimental search space and generates falsifiable predictions about behavior and neural coding.

EmbodiSwap for Zero-Shot Robot Imitation Learning

arXiv:2510.03706v1 Announce Type: cross Abstract: We introduce EmbodiSwap - a method for producing photorealistic synthetic robot overlays over human video. We employ EmbodiSwap for zero-shot imitation learning, bridging the embodiment gap between in-the-wild ego-centric human video and a target robot embodiment. We train a closed-loop robot manipulation policy over the data produced by EmbodiSwap. We make novel use of V-JEPA as a visual backbone, repurposing V-JEPA from the domain of video understanding to imitation learning over synthetic robot videos. Adoption of V-JEPA outperforms alternative vision backbones more conventionally used within robotics. In real-world tests, our zero-shot trained V-JEPA model achieves an $82\%$ success rate, outperforming a few-shot trained $\pi_0$ network as well as $\pi_0$ trained over data produced by EmbodiSwap. We release (i) code for generating the synthetic robot overlays which takes as input human videos and an arbitrary robot URDF and generates a robot dataset, (ii) the robot dataset we synthesize over EPIC-Kitchens, HOI4D and Ego4D, and (iii) model checkpoints and inference code, to facilitate reproducible research and broader adoption.

Person-Centric Annotations of LAION-400M: Auditing Bias and Its Transfer to Models

arXiv:2510.03721v1 Announce Type: cross Abstract: Vision-language models trained on large-scale multimodal datasets show strong demographic biases, but the role of training data in producing these biases remains unclear. A major barrier has been the lack of demographic annotations in web-scale datasets such as LAION-400M. We address this gap by creating person-centric annotations for the full dataset, including over 276 million bounding boxes, perceived gender and race/ethnicity labels, and automatically generated captions. These annotations are produced through validated automatic labeling pipelines combining object detection, multimodal captioning, and finetuned classifiers. Using them, we uncover demographic imbalances and harmful associations, such as the disproportionate linking of men and individuals perceived as Black or Middle Eastern with crime-related and negative content. We also show that 60-70% of gender bias in CLIP and Stable Diffusion can be linearly explained by direct co-occurrences in the data. Our resources establish the first large-scale empirical link between dataset composition and downstream model bias.

Mapping Rio de Janeiro's favelas: general-purpose vs. satellite-specific neural networks

arXiv:2510.03725v1 Announce Type: cross Abstract: While deep learning methods for detecting informal settlements have already been developed, they have not yet fully utilized the potential offered by recent pretrained neural networks. We compare two types of pretrained neural networks for detecting the favelas of Rio de Janeiro: 1. Generic networks pretrained on large diverse datasets of unspecific images, 2. A specialized network pretrained on satellite imagery. While the latter is more specific to the target task, the former has been pretrained on significantly more images. Hence, this research investigates whether task specificity or data volume yields superior performance in urban informal settlement detection.

Bridging the Gap Between Multimodal Foundation Models and World Models

arXiv:2510.03727v1 Announce Type: cross Abstract: Humans understand the world through the integration of multiple sensory modalities, enabling them to perceive, reason about, and imagine dynamic physical processes. Inspired by this capability, multimodal foundation models (MFMs) have emerged as powerful tools for multimodal understanding and generation. However, today's MFMs fall short of serving as effective world models. They lack the essential ability such as perform counterfactual reasoning, simulate dynamics, understand the spatiotemporal information, control generated visual outcomes, and perform multifaceted reasoning. We investigates what it takes to bridge the gap between multimodal foundation models and world models. We begin by improving the reasoning capabilities of MFMs through discriminative tasks and equipping MFMs with structured reasoning skills, such as causal inference, counterfactual thinking, and spatiotemporal reasoning, enabling them to go beyond surface correlations and understand deeper relationships within visual and textual data. Next, we explore generative capabilities of multimodal foundation models across both image and video modalities, introducing new frameworks for structured and controllable generation. Our approaches incorporate scene graphs, multimodal conditioning, and multimodal alignment strategies to guide the generation process, ensuring consistency with high-level semantics and fine-grained user intent. We further extend these techniques to controllable 4D generation, enabling interactive, editable, and morphable object synthesis over time and space.

Lightweight and Generalizable Acoustic Scene Representations via Contrastive Fine-Tuning and Distillation

arXiv:2510.03728v1 Announce Type: cross Abstract: Acoustic scene classification (ASC) models on edge devices typically operate under fixed class assumptions, lacking the transferability needed for real-world applications that require adaptation to new or refined acoustic categories. We propose ContrastASC, which learns generalizable acoustic scene representations by structuring the embedding space to preserve semantic relationships between scenes, enabling adaptation to unseen categories without retraining. Our approach combines supervised contrastive fine-tuning of pre-trained models with contrastive representation distillation to transfer this structured knowledge to compact student models. Our evaluation shows that ContrastASC demonstrates improved few-shot adaptation to unseen categories while maintaining strong closed-set performance.

Trajectory prediction for heterogeneous agents: A performance analysis on small and imbalanced datasets

arXiv:2510.03776v1 Announce Type: cross Abstract: Robots and other intelligent systems navigating in complex dynamic environments should predict future actions and intentions of surrounding agents to reach their goals efficiently and avoid collisions. The dynamics of those agents strongly depends on their tasks, roles, or observable labels. Class-conditioned motion prediction is thus an appealing way to reduce forecast uncertainty and get more accurate predictions for heterogeneous agents. However, this is hardly explored in the prior art, especially for mobile robots and in limited data applications. In this paper, we analyse different class-conditioned trajectory prediction methods on two datasets. We propose a set of conditional pattern-based and efficient deep learning-based baselines, and evaluate their performance on robotics and outdoors datasets (TH\"OR-MAGNI and Stanford Drone Dataset). Our experiments show that all methods improve accuracy in most of the settings when considering class labels. More importantly, we observe that there are significant differences when learning from imbalanced datasets, or in new environments where sufficient data is not available. In particular, we find that deep learning methods perform better on balanced datasets, but in applications with limited data, e.g., cold start of a robot in a new environment, or imbalanced classes, pattern-based methods may be preferable.

GuidedSampling: Steering LLMs Towards Diverse Candidate Solutions at Inference-Time

arXiv:2510.03777v1 Announce Type: cross Abstract: Repeated Sampling (RS) is a simple inference-time algorithm that has been shown to improve model performance on complex tasks. Although it is an effective way of scaling inference time, it often struggles to generate diverse solution candidates, frequently relying on the same underlying approach to solve the problem and thus producing redundant samples. To address this limitation, we propose a new inference algorithm, GuidedSampling, which decouples the exploration and generation phases during inference, increasing diversity of generated candidate solutions. The exploration phase identifies multiple concepts that can be utilized to solve the problem, while the generation phase applies a specific concept to provide final solution candidates. We first define the theoretical bounds of GuidedSampling and then empirically demonstrate that it improves the performance of base model at pass@50 by on an average ~21.6% across various benchmarks compared to RS. Furthermore, models trained on trajectories of GuidedSampling exhibit substantial performance improvements at pass@5 by on an average ~9.7%, compared to models trained on traditional RS. Additionally, models trained with GuidedSampling increases the average number of concepts per instance (1.67 -> 3.03), yielding a diverse set of candidates than traditional RS.

A Benchmark Study of Deep Learning Methods for Multi-Label Pediatric Electrocardiogram-Based Cardiovascular Disease Classification

arXiv:2510.03780v1 Announce Type: cross Abstract: Cardiovascular disease (CVD) is a major pediatric health burden, and early screening is of critical importance. Electrocardiography (ECG), as a noninvasive and accessible tool, is well suited for this purpose. This paper presents the first benchmark study of deep learning for multi-label pediatric CVD classification on the recently released ZZU-pECG dataset, comprising 3716 recordings with 19 CVD categories. We systematically evaluate four representative paradigms--ResNet-1D, BiLSTM, Transformer, and Mamba 2--under both 9-lead and 12-lead configurations. All models achieved strong results, with Hamming Loss as low as 0.0069 and F1-scores above 85% in most settings. ResNet-1D reached a macro-F1 of 94.67% on the 12-lead subset, while BiLSTM and Transformer also showed competitive performance. Per-class analysis indicated challenges for rare conditions such as hypertrophic cardiomyopathy in the 9-lead subset, reflecting the effect of limited positive samples. This benchmark establishes reusable baselines and highlights complementary strengths across paradigms. It further points to the need for larger-scale, multi-center validation, age-stratified analysis, and broader disease coverage to support real-world pediatric ECG applications.

Road Damage and Manhole Detection using Deep Learning for Smart Cities: A Polygonal Annotation Approach

arXiv:2510.03797v1 Announce Type: cross Abstract: Urban safety and infrastructure maintenance are critical components of smart city development. Manual monitoring of road damages is time-consuming, highly costly, and error-prone. This paper presents a deep learning approach for automated road damage and manhole detection using the YOLOv9 algorithm with polygonal annotations. Unlike traditional bounding box annotation, we employ polygonal annotations for more precise localization of road defects. We develop a novel dataset comprising more than one thousand images which are mostly collected from Dhaka, Bangladesh. This dataset is used to train a YOLO-based model for three classes, namely Broken, Not Broken, and Manhole. We achieve 78.1% overall image-level accuracy. The YOLOv9 model demonstrates strong performance for Broken (86.7% F1-score) and Not Broken (89.2% F1-score) classes, with challenges in Manhole detection (18.2% F1-score) due to class imbalance. Our approach offers an efficient and scalable solution for monitoring urban infrastructure in developing countries.

6G-Enabled Digital Twin Framework for Real-Time Cyber-Physical Systems: An Experimental Validation with Industrial Bearing Fault Detection

arXiv:2510.03807v1 Announce Type: cross Abstract: Current Cyber-Physical Systems (CPS) integrated with Digital Twin (DT) technology face critical limitations in achieving real-time performance for mission-critical industrial applications. Existing 5G-enabled systems suffer from latencies exceeding 10ms, which are inadequate for applications requiring sub-millisecond response times, such as autonomous industrial control and predictive maintenance. This research aims to develop and validate a 6G-enabled Digital Twin framework that achieves ultra-low latency communication and real-time synchronization between physical industrial assets and their digital counterparts, specifically targeting bearing fault detection as a critical industrial use case. The proposed framework integrates terahertz communications (0.1-1 THz), intelligent reflecting surfaces, and edge artificial intelligence within a five-layer architecture. Experimental validation was conducted using the Case Western Reserve University (CWRU) bearing dataset, implementing comprehensive feature extraction (15 time and frequency domain features) and Random Forest classification algorithms. The system performance was evaluated against traditional WiFi-6 and 5G networks across multiple metrics, including classification accuracy, end-to-end latency, and scalability. It achieved 97.7% fault classification accuracy with 0.8ms end-to-end latency, representing a 15.6x improvement over WiFi-6 (12.5ms) and 5.25x improvement over 5G (4.2ms) networks. The system demonstrated superior scalability with sub-linear processing time growth and maintained consistent performance across four bearing fault categories (normal, inner race, outer race, and ball faults) with macro-averaged F1-scores exceeding 97%.

Spectral Thresholds for Identifiability and Stability:Finite-Sample Phase Transitions in High-Dimensional Learning

arXiv:2510.03809v1 Announce Type: cross Abstract: In high-dimensional learning, models remain stable until they collapse abruptly once the sample size falls below a critical level. This instability is not algorithm-specific but a geometric mechanism: when the weakest Fisher eigendirection falls beneath sample-level fluctuations, identifiability fails. Our Fisher Threshold Theorem formalizes this by proving that stability requires the minimal Fisher eigenvalue to exceed an explicit $O(\sqrt{d/n})$ bound. Unlike prior asymptotic or model-specific criteria, this threshold is finite-sample and necessary, marking a sharp phase transition between reliable concentration and inevitable failure. To make the principle constructive, we introduce the Fisher floor, a verifiable spectral regularization robust to smoothing and preconditioning. Synthetic experiments on Gaussian mixtures and logistic models confirm the predicted transition, consistent with $d/n$ scaling. Statistically, the threshold sharpens classical eigenvalue conditions into a non-asymptotic law; learning-theoretically, it defines a spectral sample-complexity frontier, bridging theory with diagnostics for robust high-dimensional inference.

Cellular Learning: Scattered Data Regression in High Dimensions via Voronoi Cells

arXiv:2510.03810v1 Announce Type: cross Abstract: I present a regression algorithm that provides a continuous, piecewise-smooth function approximating scattered data. It is based on composing and blending linear functions over Voronoi cells, and it scales to high dimensions. The algorithm infers Voronoi cells from seed vertices and constructs a linear function for the input data in and around each cell. As the algorithm does not explicitly compute the Voronoi diagram, it avoids the curse of dimensionality. An accuracy of around 98.2% on the MNIST dataset with 722,200 degrees of freedom (without data augmentation, convolution, or other geometric operators) demonstrates the applicability and scalability of the algorithm.

Diverse Text-to-Image Generation via Contrastive Noise Optimization

arXiv:2510.03813v1 Announce Type: cross Abstract: Text-to-image (T2I) diffusion models have demonstrated impressive performance in generating high-fidelity images, largely enabled by text-guided inference. However, this advantage often comes with a critical drawback: limited diversity, as outputs tend to collapse into similar modes under strong text guidance. Existing approaches typically optimize intermediate latents or text conditions during inference, but these methods deliver only modest gains or remain sensitive to hyperparameter tuning. In this work, we introduce Contrastive Noise Optimization, a simple yet effective method that addresses the diversity issue from a distinct perspective. Unlike prior techniques that adapt intermediate latents, our approach shapes the initial noise to promote diverse outputs. Specifically, we develop a contrastive loss defined in the Tweedie data space and optimize a batch of noise latents. Our contrastive optimization repels instances within the batch to maximize diversity while keeping them anchored to a reference sample to preserve fidelity. We further provide theoretical insights into the mechanism of this preprocessing to substantiate its effectiveness. Extensive experiments across multiple T2I backbones demonstrate that our approach achieves a superior quality-diversity Pareto frontier while remaining robust to hyperparameter choices.

A Trustworthy Industrial Fault Diagnosis Architecture Integrating Probabilistic Models and Large Language Models

arXiv:2510.03815v1 Announce Type: cross Abstract: There are limitations of traditional methods and deep learning methods in terms of interpretability, generalization, and quantification of uncertainty in industrial fault diagnosis, and there are core problems of insufficient credibility in industrial fault diagnosis. The architecture performs preliminary analysis through a Bayesian network-based diagnostic engine and features an LLM-driven cognitive quorum module with multimodal input capabilities. The module conducts expert-level arbitration of initial diagnoses by analyzing structured features and diagnostic charts, prioritizing final decisions after conflicts are identified. To ensure the reliability of the system output, the architecture integrates a confidence calibration module based on temperature calibration and a risk assessment module, which objectively quantifies the reliability of the system using metrics such as expected calibration error (ECE). Experimental results on a dataset containing multiple fault types showed that the proposed framework improved diagnostic accuracy by more than 28 percentage points compared to the baseline model, while the calibrated ECE was reduced by more than 75%. Case studies have confirmed that HCAA effectively corrects misjudgments caused by complex feature patterns or knowledge gaps in traditional models, providing novel and practical engineering solutions for building high-trust, explainable AI diagnostic systems for industrial applications.

Pilot Contamination Attacks Detection with Machine Learning for Multi-User Massive MIMO

arXiv:2510.03831v1 Announce Type: cross Abstract: Massive multiple-input multiple-output (MMIMO) is essential to modern wireless communication systems, like 5G and 6G, but it is vulnerable to active eavesdropping attacks. One type of such attack is the pilot contamination attack (PCA), where a malicious user copies pilot signals from an authentic user during uplink, intentionally interfering with the base station's (BS) channel estimation accuracy. In this work, we propose to use a Decision Tree (DT) algorithm for PCA detection at the BS in a multi-user system. We present a methodology to generate training data for the DT classifier and select the best DT according to their depth. Then, we simulate different scenarios that could be encountered in practice and compare the DT to a classical technique based on likelihood ratio testing (LRT) submitted to the same scenarios. The results revealed that a DT with only one level of depth is sufficient to outperform the LRT. The DT shows a good performance regarding the probability of detection in noisy scenarios and when the malicious user transmits with low power, in which case the LRT fails to detect the PCA. We also show that the reason for the good performance of the DT is its ability to compute a threshold that separates PCA data from non-PCA data better than the LRT's threshold. Moreover, the DT does not necessitate prior knowledge of noise power or assumptions regarding the signal power of malicious users, prerequisites typically essential for LRT and other hypothesis testing methodologies.

Smart Paste: Automatically Fixing Copy/Paste for Google Developers

arXiv:2510.03843v1 Announce Type: cross Abstract: Manually editing pasted code is a long-standing developer pain point. In internal software development at Google, we observe that code is pasted 4 times more often than it is manually typed. These paste actions frequently require follow-up edits, ranging from simple reformatting and renaming to more complex style adjustments and cross-language translations. Prior work has shown deep learning can be used to predict these edits. In this work, we show how to iteratively develop and scale Smart Paste, an IDE feature for post-paste edit suggestions, to Google's development environment. This experience can serve as a guide for AI practitioners on a holistic approach to feature development, covering user experience, system integration, and model capabilities. Since deployment, Smart Paste has had overwhelmingly positive feedback with a 45% acceptance rate. At Google's enterprise scale, these accepted suggestions account substantially for over 1% of all code written company-wide.

The Hidden Game Problem

arXiv:2510.03845v1 Announce Type: cross Abstract: This paper investigates a class of games with large strategy spaces, motivated by challenges in AI alignment and language games. We introduce the hidden game problem, where for each player, an unknown subset of strategies consistently yields higher rewards compared to the rest. The central question is whether efficient regret minimization algorithms can be designed to discover and exploit such hidden structures, leading to equilibrium in these subgames while maintaining rationality in general. We answer this question affirmatively by developing a composition of regret minimization techniques that achieve optimal external and swap regret bounds. Our approach ensures rapid convergence to correlated equilibria in hidden subgames, leveraging the hidden game structure for improved computational efficiency.

Small Language Models for Agentic Systems: A Survey of Architectures, Capabilities, and Deployment Trade offs

arXiv:2510.03847v1 Announce Type: cross Abstract: Small language models (SLMs; 1-12B params, sometimes up to 20B) are sufficient and often superior for agentic workloads where the objective is schema- and API-constrained accuracy rather than open-ended generation. We synthesize recent evidence across open and proprietary SLMs (Phi-4-Mini, Qwen-2.5-7B, Gemma-2-9B, Llama-3.2-1B/3B, Ministral-3B/8B, Apple on-device 3B, DeepSeek-R1-Distill) and connect it to modern evaluations (BFCL v3/v4, StableToolBench) and serving stacks (vLLM, SGLang, TensorRT-LLM) paired with guided decoding libraries (XGrammar, Outlines). We formalize SLM-default, LLM-fallback systems with uncertainty-aware routing and verifier cascades, and propose engineering metrics that reflect real production goals: cost per successful task (CPS), schema validity rate, executable call rate, p50/p95 latency, and energy per request. Guided decoding, strict JSON Schema outputs, and validator-first tool execution close much of the capability gap with larger models and often let SLMs match or surpass LLMs on tool use, function calling, and RAG at 10x-100x lower token cost with materially better latency and energy. We provide design patterns for agent stacks that prioritize SLMs: schema-first prompting, type-safe function registries, confidence scoring with verifier rollups, and lightweight adaptation via LoRA/QLoRA. We also delineate limits where fallback remains valuable (open-domain reasoning and some long-horizon planning). The result is a practical blueprint for building fast, inexpensive, and reliable agents that default to SLMs while preserving headroom with targeted LLM assistance. Keywords: small language models, agents, function calling, structured outputs, JSON Schema, guided decoding, LoRA/QLoRA, routing, energy efficiency, edge inference

Adaptive and Explainable AI Agents for Anomaly Detection in Critical IoT Infrastructure using LLM-Enhanced Contextual Reasoning

arXiv:2510.03859v1 Announce Type: cross Abstract: Ensuring that critical IoT systems function safely and smoothly depends a lot on finding anomalies quickly. As more complex systems, like smart healthcare, energy grids and industrial automation, appear, it is easier to see the shortcomings of older methods of detection. Monitoring failures usually happen in dynamic, high dimensional situations, especially when data is incomplete, messy or always evolving. Such limits point out the requirement for adaptive, intelligent systems that always improve and think. LLMs are now capable of significantly changing how context is understood and semantic inference is done across all types of data. This proposal suggests using an LLM supported contextual reasoning method along with XAI agents to improve how anomalies are found in significant IoT environments. To discover hidden patterns and notice inconsistencies in data streams, it uses attention methods, avoids dealing with details from every time step and uses memory buffers with meaning. Because no code AI stresses transparency and interpretability, people can check and accept the AI's decisions, helping ensure AI follows company policies. The two architectures are put together in a test that compares the results of the traditional model with those of the suggested LLM enhanced model. Important measures to check are the accuracy of detection, how much inaccurate information is included in the results, how clearly the findings can be read and how fast the system responds under different test situations. The metaheuristic is tested in simulations of real world smart grid and healthcare contexts to check its adaptability and reliability. From the study, we see that the new approach performs much better than most existing models in both accuracy and interpretation, so it could be a good fit for future anomaly detection tasks in IoT

Fair Minimum Labeling: Efficient Temporal Network Activations for Reachability and Equity

arXiv:2510.03899v1 Announce Type: cross Abstract: Balancing resource efficiency and fairness is critical in networked systems that support modern learning applications. We introduce the Fair Minimum Labeling (FML) problem: the task of designing a minimum-cost temporal edge activation plan that ensures each group of nodes in a network has sufficient access to a designated target set, according to specified coverage requirements. FML captures key trade-offs in systems where edge activations incur resource costs and equitable access is essential, such as distributed data collection, update dissemination in edge-cloud systems, and fair service restoration in critical infrastructure. We show that FML is NP-hard and $\Omega(\log |V|)$-hard to approximate, and we present probabilistic approximation algorithms that match this bound, achieving the best possible guarantee for the activation cost. We demonstrate the practical utility of FML in a fair multi-source data aggregation task for training a shared model. Empirical results show that FML enforces group-level fairness with substantially lower activation cost than baseline heuristics, underscoring its potential for building resource-efficient, equitable temporal reachability in learning-integrated networks.

Optimal Computation from Fluctuation Responses

arXiv:2510.03900v1 Announce Type: cross Abstract: The energy cost of computation has emerged as a central challenge at the intersection of physics and computer science. Recent advances in statistical physics -- particularly in stochastic thermodynamics -- enable precise characterizations of work, heat, and entropy production in information-processing systems driven far from equilibrium by time-dependent control protocols. A key open question is then how to design protocols that minimize thermodynamic cost while ensur- ing correct outcomes. To this end, we develop a unified framework to identify optimal protocols using fluctuation response relations (FRR) and machine learning. Unlike previous approaches that optimize either distributions or protocols separately, our method unifies both using FRR-derived gradients. Moreover, our method is based primarily on iteratively learning from sampled noisy trajectories, which is generally much easier than solving for the optimal protocol directly from a set of governing equations. We apply the framework to canonical examples -- bit erasure in a double-well potential and translating harmonic traps -- demonstrating how to construct loss functions that trade-off energy cost against task error. The framework extends trivially to underdamped systems, and we show this by optimizing a bit-flip in an underdamped system. In all computations we test, the framework achieves the theoretically optimal protocol or achieves work costs comparable to relevant finite time bounds. In short, the results provide principled strategies for designing thermodynamically efficient protocols in physical information-processing systems. Applications range from quantum gates robust under noise to energy-efficient control of chemical and synthetic biological networks.

Self-Speculative Masked Diffusions

arXiv:2510.03929v1 Announce Type: cross Abstract: We present self-speculative masked diffusions, a new class of masked diffusion generative models for discrete data that require significantly fewer function evaluations to generate samples. Standard masked diffusion models predict factorized logits over currently masked positions. A number of masked positions are then sampled, however, the factorization approximation means that sampling too many positions in one go leads to poor sample quality. As a result, many simulation steps and therefore neural network function evaluations are required to generate high-quality data. We reduce the computational burden by generating non-factorized predictions over masked positions. This is achieved by modifying the final transformer attention mask from non-causal to causal, enabling draft token generation and parallel validation via a novel, model-integrated speculative sampling mechanism. This results in a non-factorized predictive distribution over masked positions in a single forward pass. We apply our method to GPT2 scale text modelling and protein sequences generation, finding that we can achieve a ~2x reduction in the required number of network forward passes relative to standard masked diffusion models.

Quantifying Risks in Multi-turn Conversation with Large Language Models

arXiv:2510.03969v1 Announce Type: cross Abstract: Large Language Models (LLMs) can produce catastrophic responses in conversational settings that pose serious risks to public safety and security. Existing evaluations often fail to fully reveal these vulnerabilities because they rely on fixed attack prompt sequences, lack statistical guarantees, and do not scale to the vast space of multi-turn conversations. In this work, we propose QRLLM, a novel, principled Certification framework for Catastrophic risks in multi-turn Conversation for LLMs that bounds the probability of an LLM generating catastrophic responses under multi-turn conversation distributions with statistical guarantees. We model multi-turn conversations as probability distributions over query sequences, represented by a Markov process on a query graph whose edges encode semantic similarity to capture realistic conversational flow, and quantify catastrophic risks using confidence intervals. We define several inexpensive and practical distributions: random node, graph path, adaptive with rejection. Our results demonstrate that these distributions can reveal substantial catastrophic risks in frontier models, with certified lower bounds as high as 70\% for the worst model, highlighting the urgent need for improved safety training strategies in frontier LLMs.

Beyond Static Evaluation: Rethinking the Assessment of Personalized Agent Adaptability in Information Retrieval

arXiv:2510.03984v1 Announce Type: cross Abstract: Personalized AI agents are becoming central to modern information retrieval, yet most evaluation methodologies remain static, relying on fixed benchmarks and one-off metrics that fail to reflect how users' needs evolve over time. These limitations hinder our ability to assess whether agents can meaningfully adapt to individuals across dynamic, longitudinal interactions. In this perspective paper, we propose a conceptual lens for rethinking evaluation in adaptive personalization, shifting the focus from static performance snapshots to interaction-aware, evolving assessments. We organize this lens around three core components: (1) persona-based user simulation with temporally evolving preference models; (2) structured elicitation protocols inspired by reference interviews to extract preferences in context; and (3) adaptation-aware evaluation mechanisms that measure how agent behavior improves across sessions and tasks. While recent works have embraced LLM-driven user simulation, we situate this practice within a broader paradigm for evaluating agents over time. To illustrate our ideas, we conduct a case study in e-commerce search using the PersonalWAB dataset. Beyond presenting a framework, our work lays a conceptual foundation for understanding and evaluating personalization as a continuous, user-centric endeavor.

Keep It on a Leash: Controllable Pseudo-label Generation Towards Realistic Long-Tailed Semi-Supervised Learning

arXiv:2510.03993v1 Announce Type: cross Abstract: Current long-tailed semi-supervised learning methods assume that labeled data exhibit a long-tailed distribution, and unlabeled data adhere to a typical predefined distribution (i.e., long-tailed, uniform, or inverse long-tailed). However, the distribution of the unlabeled data is generally unknown and may follow an arbitrary distribution. To tackle this challenge, we propose a Controllable Pseudo-label Generation (CPG) framework, expanding the labeled dataset with the progressively identified reliable pseudo-labels from the unlabeled dataset and training the model on the updated labeled dataset with a known distribution, making it unaffected by the unlabeled data distribution. Specifically, CPG operates through a controllable self-reinforcing optimization cycle: (i) at each training step, our dynamic controllable filtering mechanism selectively incorporates reliable pseudo-labels from the unlabeled dataset into the labeled dataset, ensuring that the updated labeled dataset follows a known distribution; (ii) we then construct a Bayes-optimal classifier using logit adjustment based on the updated labeled data distribution; (iii) this improved classifier subsequently helps identify more reliable pseudo-labels in the next training step. We further theoretically prove that this optimization cycle can significantly reduce the generalization error under some conditions. Additionally, we propose a class-aware adaptive augmentation module to further improve the representation of minority classes, and an auxiliary branch to maximize data utilization by leveraging all labeled and unlabeled samples. Comprehensive evaluations on various commonly used benchmark datasets show that CPG achieves consistent improvements, surpassing state-of-the-art methods by up to \textbf{15.97\%} in accuracy. The code is available at https://github.com/yaxinhou/CPG.

Multi-Modal Multi-Task Semantic Communication: A Distributed Information Bottleneck Perspective

arXiv:2510.04000v1 Announce Type: cross Abstract: Semantic communication (SemCom) shifts the focus from data transmission to meaning delivery, enabling efficient and intelligent communication. Existing AI-based coding schemes for multi-modal multi-task SemCom often require transmitters with full-modal data to participate in all receivers' tasks, which leads to redundant transmissions and conflicts with the physical limits of channel capacity and computational capability. In this paper, we propose PoM$^2$-DIB, a novel framework that extends the distributed information bottleneck (DIB) theory to address this problem. Unlike the typical DIB, this framework introduces modality selection as an additional key design variable, enabling a more flexible tradeoff between communication rate and inference quality. This extension selects only the most relevant modalities for task participation, adhering to the physical constraints, while following efficient DIB-based coding. To optimize selection and coding end-to-end, we relax modality selection into a probabilistic form, allowing the use of score function estimation with common randomness to enable optimizable coordinated decisions across distributed devices. Experimental results on public datasets verify that PoM$^2$-DIB achieves high inference quality compared to full-participation baselines in various tasks under physical limits.

Zephyrus: An Agentic Framework for Weather Science

arXiv:2510.04017v1 Announce Type: cross Abstract: Foundation models for weather science are pre-trained on vast amounts of structured numerical data and outperform traditional weather forecasting systems. However, these models lack language-based reasoning capabilities, limiting their utility in interactive scientific workflows. Large language models (LLMs) excel at understanding and generating text but cannot reason about high-dimensional meteorological datasets. We bridge this gap by building a novel agentic framework for weather science. Our framework includes a Python code-based environment for agents (ZephyrusWorld) to interact with weather data, featuring tools like an interface to WeatherBench 2 dataset, geoquerying for geographical masks from natural language, weather forecasting, and climate simulation capabilities. We design Zephyrus, a multi-turn LLM-based weather agent that iteratively analyzes weather datasets, observes results, and refines its approach through conversational feedback loops. We accompany the agent with a new benchmark, ZephyrusBench, with a scalable data generation pipeline that constructs diverse question-answer pairs across weather-related tasks, from basic lookups to advanced forecasting, extreme event detection, and counterfactual reasoning. Experiments on this benchmark demonstrate the strong performance of Zephyrus agents over text-only baselines, outperforming them by up to 35 percentage points in correctness. However, on harder tasks, Zephyrus performs similarly to text-only baselines, highlighting the challenging nature of our benchmark and suggesting promising directions for future work.

Simulation-based inference via telescoping ratio estimation for trawl processes

arXiv:2510.04042v1 Announce Type: cross Abstract: The growing availability of large and complex datasets has increased interest in temporal stochastic processes that can capture stylized facts such as marginal skewness, non-Gaussian tails, long memory, and even non-Markovian dynamics. While such models are often easy to simulate from, parameter estimation remains challenging. Simulation-based inference (SBI) offers a promising way forward, but existing methods typically require large training datasets or complex architectures and frequently yield confidence (credible) regions that fail to attain their nominal values, raising doubts on the reliability of estimates for the very features that motivate the use of these models. To address these challenges, we propose a fast and accurate, sample-efficient SBI framework for amortized posterior inference applicable to intractable stochastic processes. The proposed approach relies on two main steps: first, we learn the posterior density by decomposing it sequentially across parameter dimensions. Then, we use Chebyshev polynomial approximations to efficiently generate independent posterior samples, enabling accurate inference even when Markov chain Monte Carlo methods mix poorly. We further develop novel diagnostic tools for SBI in this context, as well as post-hoc calibration techniques; the latter not only lead to performance improvements of the learned inferential tool, but also to the ability to reuse it directly with new time series of varying lengths, thus amortizing the training cost. We demonstrate the method's effectiveness on trawl processes, a class of flexible infinitely divisible models that generalize univariate Gaussian processes, applied to energy demand data.

Exploring Chain-of-Thought Reasoning for Steerable Pluralistic Alignment

arXiv:2510.04045v1 Announce Type: cross Abstract: Large Language Models (LLMs) are typically trained to reflect a relatively uniform set of values, which limits their applicability to tasks that require understanding of nuanced human perspectives. Recent research has underscored the importance of enabling LLMs to support steerable pluralism -- the capacity to adopt a specific perspective and align generated outputs with it. In this work, we investigate whether Chain-of-Thought (CoT) reasoning techniques can be applied to building steerable pluralistic models. We explore several methods, including CoT prompting, fine-tuning on human-authored CoT, fine-tuning on synthetic explanations, and Reinforcement Learning with Verifiable Rewards (RLVR). We evaluate these approaches using the Value Kaleidoscope and OpinionQA datasets. Among the methods studied, RLVR consistently outperforms others and demonstrates strong training sample efficiency. We further analyze the generated CoT traces with respect to faithfulness and safety.

Sharp Lower Bounds for Linearized ReLU^k Approximation on the Sphere

arXiv:2510.04060v1 Announce Type: cross Abstract: We prove a saturation theorem for linearized shallow ReLU$^k$ neural networks on the unit sphere $\mathbb S^d$. For any antipodally quasi-uniform set of centers, if the target function has smoothness $r>\tfrac{d+2k+1}{2}$, then the best $\mathcal{L}^2(\mathbb S^d)$ approximation cannot converge faster than order $n^{-\frac{d+2k+1}{2d}}$. This lower bound matches existing upper bounds, thereby establishing the exact saturation order $\tfrac{d+2k+1}{2d}$ for such networks. Our results place linearized neural-network approximation firmly within the classical saturation framework and show that, although ReLU$^k$ networks outperform finite elements under equal degrees $k$, this advantage is intrinsically limited.

A Contextual Quality Reward Model for Reliable and Efficient Best-of-N Sampling

arXiv:2510.04087v1 Announce Type: cross Abstract: Modern preference alignment techniques, such as Best-of-N (BoN) sampling, rely on reward models trained with pairwise comparison data. While effective at learning relative preferences, this paradigm fails to capture a signal of response acceptability, leaving systems vulnerable to selecting the least bad of many unacceptable options. This is particularly problematic for hard prompts, where the risk of such false acceptances increases with the number of samples. In this paper, we address this critical reliability gap by introducing a new data collection and modeling framework. By augmenting preference data with an outside option, inspired by discrete choice models, we train a reward model that can distinguish not just what is \textit{better}, but what is \textit{good enough}. We leverage this capability to create an adaptive inference strategy, best of mini-N in-loop, which partitions the generation budget into sequential loops with a calibrated, early-exit condition. Our experiments show that when tuned as an alignment guardrail, it reduces reliability failures by 70\%, and when tuned as an inference accelerator, it improves average inference speed by over 22\% in IMDB-sentiment setting. We thus provide a principled and flexible framework for practitioners to explicitly manage the trade-off between reliability and computational efficiency.

RLRF: Competitive Search Agent Design via Reinforcement Learning from Ranker Feedback

arXiv:2510.04096v1 Announce Type: cross Abstract: Competitive search is a setting where document publishers modify them to improve their ranking in response to a query. Recently, publishers have increasingly leveraged LLMs to generate and modify competitive content. We introduce Reinforcement Learning from Ranker Feedback (RLRF), a framework that trains LLMs using preference datasets derived from ranking competitions. The goal of a publisher (LLM-based) agent is to optimize content for improved ranking while accounting for the strategies of competing agents. We generate the datasets using approaches that do not rely on human-authored data. We show that our proposed agents consistently and substantially outperform previously suggested approaches for LLM-based competitive document modification. We further show that our agents are effective with ranking functions they were not trained for (i.e., out of distribution) and they adapt to strategic opponents. These findings provide support to the significant potential of using reinforcement learning in competitive search.

Learning-Based Hashing for ANN Search: Foundations and Early Advances

arXiv:2510.04127v1 Announce Type: cross Abstract: Approximate Nearest Neighbour (ANN) search is a fundamental problem in information retrieval, underpinning large-scale applications in computer vision, natural language processing, and cross-modal search. Hashing-based methods provide an efficient solution by mapping high-dimensional data into compact binary codes that enable fast similarity computations in Hamming space. Over the past two decades, a substantial body of work has explored learning to hash, where projection and quantisation functions are optimised from data rather than chosen at random. This article offers a foundational survey of early learning-based hashing methods, with an emphasis on the core ideas that shaped the field. We review supervised, unsupervised, and semi-supervised approaches, highlighting how projection functions are designed to generate meaningful embeddings and how quantisation strategies convert these embeddings into binary codes. We also examine extensions to multi-bit and multi-threshold models, as well as early advances in cross-modal retrieval. Rather than providing an exhaustive account of the most recent methods, our goal is to introduce the conceptual foundations of learning-based hashing for ANN search. By situating these early models in their historical context, we aim to equip readers with a structured understanding of the principles, trade-offs, and open challenges that continue to inform current research in this area.

Fine Tuning Methods for Low-resource Languages

arXiv:2510.04139v1 Announce Type: cross Abstract: The rise of Large Language Models has not been inclusive of all cultures. The models are mostly trained on English texts and culture which makes them underperform in other languages and cultural contexts. By developing a generalizable method for preparing culturally relevant datasets and post-training the Gemma 2 model, this project aimed to increase the performance of Gemma 2 for an underrepresented language and showcase how others can do the same to unlock the power of Generative AI in their country and preserve their cultural heritage.

Learning from All: Concept Alignment for Autonomous Distillation from Multiple Drifting MLLMs

arXiv:2510.04142v1 Announce Type: cross Abstract: This paper identifies a critical yet underexplored challenge in distilling from multimodal large language models (MLLMs): the reasoning trajectories generated by multiple drifting teachers exhibit concept drift, whereby their reasoning distributions evolve unpredictably and transmit biases to the student model, ultimately compromising its performance. To tackle this issue, we pioneer a theoretical connection between concept drift and knowledge distillation, casting the non-stationary reasoning dynamics from multiple MLLM teachers as next-token prediction of multi-stream reasoning trajectories.Guided by concept drift, we introduce the "learn, compare, critique" paradigm, culminating in autonomous preference optimization (APO). Under the active guidance of the teachers, the student model first learns and self-distils preferred thinking by comparing multiple teachers. It then engages in critical reflection over the drifting inference from teachers, performing concept alignment through APO, ultimately yielding a robust, consistent, and generalizable model.Extensive experiments demonstrate our superior performance of consistency, robustness and generalization within knowledge distillation. Besides, we also contributed a large-scale dataset, CXR-MAX (Multi-teachers Alignment X-rays), comprising 170,982 distilled reasoning trajectories derived from publicly accessible MLLMs based on MIMIC-CXR. Our code and data are public at: https://anonymous.4open.science/r/Autonomous-Distillation/.

ObCLIP: Oblivious CLoud-Device Hybrid Image Generation with Privacy Preservation

arXiv:2510.04153v1 Announce Type: cross Abstract: Diffusion Models have gained significant popularity due to their remarkable capabilities in image generation, albeit at the cost of intensive computation requirement. Meanwhile, despite their widespread deployment in inference services such as Midjourney, concerns about the potential leakage of sensitive information in uploaded user prompts have arisen. Existing solutions either lack rigorous privacy guarantees or fail to strike an effective balance between utility and efficiency. To bridge this gap, we propose ObCLIP, a plug-and-play safeguard that enables oblivious cloud-device hybrid generation. By oblivious, each input prompt is transformed into a set of semantically similar candidate prompts that differ only in sensitive attributes (e.g., gender, ethnicity). The cloud server processes all candidate prompts without knowing which one is the real one, thus preventing any prompt leakage. To mitigate server cost, only a small portion of denoising steps is performed upon the large cloud model. The intermediate latents are then sent back to the client, which selects the targeted latent and completes the remaining denoising using a small device model. Additionally, we analyze and incorporate several cache-based accelerations that leverage temporal and batch redundancy, effectively reducing computation cost with minimal utility degradation. Extensive experiments across multiple datasets demonstrate that ObCLIP provides rigorous privacy and comparable utility to cloud models with slightly increased server cost.

Drax: Speech Recognition with Discrete Flow Matching

arXiv:2510.04162v1 Announce Type: cross Abstract: Diffusion and flow-based non-autoregressive (NAR) models have shown strong promise in large language modeling, however, their potential for automatic speech recognition (ASR) remains largely unexplored. We propose Drax, a discrete flow matching framework for ASR that enables efficient parallel decoding. To better align training with inference, we construct an audio-conditioned probability path that guides the model through trajectories resembling likely intermediate inference errors, rather than direct random noise to target transitions. Our theoretical analysis links the generalization gap to divergences between training and inference occupancies, controlled by cumulative velocity errors, thereby motivating our design choice. Empirical evaluation demonstrates that our approach attains recognition accuracy on par with state-of-the-art speech models while offering improved accuracy-efficiency trade-offs, highlighting discrete flow matching as a promising direction for advancing NAR ASR.

From Segments to Concepts: Interpretable Image Classification via Concept-Guided Segmentation

arXiv:2510.04180v1 Announce Type: cross Abstract: Deep neural networks have achieved remarkable success in computer vision; however, their black-box nature in decision-making limits interpretability and trust, particularly in safety-critical applications. Interpretability is crucial in domains where errors have severe consequences. Existing models not only lack transparency but also risk exploiting unreliable or misleading features, which undermines both robustness and the validity of their explanations. Concept Bottleneck Models (CBMs) aim to improve transparency by reasoning through human-interpretable concepts. Still, they require costly concept annotations and lack spatial grounding, often failing to identify which regions support each concept. We propose SEG-MIL-CBM, a novel framework that integrates concept-guided image segmentation into an attention-based multiple instance learning (MIL) framework, where each segmented region is treated as an instance and the model learns to aggregate evidence across them. By reasoning over semantically meaningful regions aligned with high-level concepts, our model highlights task-relevant evidence, down-weights irrelevant cues, and produces spatially grounded, concept-level explanations without requiring annotations of concepts or groups. SEG-MIL-CBM achieves robust performance across settings involving spurious correlations (unintended dependencies between background and label), input corruptions (perturbations that degrade visual quality), and large-scale benchmarks, while providing transparent, concept-level explanations.

COSMO-RL: Towards Trustworthy LMRMs via Joint Safety and Stability

arXiv:2510.04196v1 Announce Type: cross Abstract: Large Multimodal Reasoning Models (LMRMs) are moving into real applications, where they must be both useful and safe. Safety is especially challenging in multimodal settings: images and text can be combined to bypass guardrails, and single objective training can cause policy drift that yields over-refusal on benign inputs or unsafe compliance on risky ones. We present COSMO-RL, a mixed reinforcement learning framework that trains reasoning oriented LMRMs under multimodal, multitask, and multiobjective signals, and we release the resulting model, COSMO-R1. Our approach aims to let safety and capability grow together in one stable pipeline rather than competing during alignment. In experiments, COSMO-R1 improves safety while maintaining-and often improving multimodal reasoning and instruction following, shows stronger robustness to multimodal jailbreaks, and reduces unnecessary refusals. The framework also transfers across backbones with consistent gains. Ablations support the design choices, indicating a simple path to advancing safety and general capability together in LMRMs.

CALM Before the STORM: Unlocking Native Reasoning for Optimization Modeling

arXiv:2510.04204v1 Announce Type: cross Abstract: Large Reasoning Models (LRMs) have demonstrated strong capabilities in complex multi-step reasoning, opening new opportunities for automating optimization modeling. However, existing domain adaptation methods, originally designed for earlier instruction-tuned models, often fail to exploit the advanced reasoning patterns of modern LRMs -- In particular, we show that direct fine-tuning on traditional \textit{non-reflective} datasets leads to limited gains. To fully leverage LRMs' inherent reasoning abilities, we propose \textbf{CALM} (\textit{Corrective Adaptation with Lightweight Modification}), a framework that progressively refines LRMs within their native reasoning modes for optimization modeling tasks. In CALM, an expert intervener identifies reasoning flaws and provides concise corrective hints, which the LRM incorporates to produce improved reasoning trajectories. These interventions modify fewer than 2.6\% of generated tokens, but generate high-quality data for soft adaptation through supervised fine-tuning. The adapted model is then further improved through reinforcement learning. Building on CALM, we develop \textbf{STORM} (\textit{Smart Thinking Optimization Reasoning Model}), a 4B-parameter LRM that achieves a new state-of-the-art average accuracy of 68.9\% across five popular optimization modeling benchmarks, matching the performance of a 671B LRM. These results demonstrate that dynamic, hint-based data synthesis both preserves and amplifies the native reasoning patterns of modern LRMs, offering a more effective and scalable path towards expert-level performance on challenging optimization modeling tasks.

MASC: Boosting Autoregressive Image Generation with a Manifold-Aligned Semantic Clustering

arXiv:2510.04220v1 Announce Type: cross Abstract: Autoregressive (AR) models have shown great promise in image generation, yet they face a fundamental inefficiency stemming from their core component: a vast, unstructured vocabulary of visual tokens. This conventional approach treats tokens as a flat vocabulary, disregarding the intrinsic structure of the token embedding space where proximity often correlates with semantic similarity. This oversight results in a highly complex prediction task, which hinders training efficiency and limits final generation quality. To resolve this, we propose Manifold-Aligned Semantic Clustering (MASC), a principled framework that constructs a hierarchical semantic tree directly from the codebook's intrinsic structure. MASC employs a novel geometry-aware distance metric and a density-driven agglomerative construction to model the underlying manifold of the token embeddings. By transforming the flat, high-dimensional prediction task into a structured, hierarchical one, MASC introduces a beneficial inductive bias that significantly simplifies the learning problem for the AR model. MASC is designed as a plug-and-play module, and our extensive experiments validate its effectiveness: it accelerates training by up to 57% and significantly improves generation quality, reducing the FID of LlamaGen-XL from 2.87 to 2.58. MASC elevates existing AR frameworks to be highly competitive with state-of-the-art methods, establishing that structuring the prediction space is as crucial as architectural innovation for scalable generative modeling.

Epistemic Diversity and Knowledge Collapse in Large Language Models

arXiv:2510.04226v1 Announce Type: cross Abstract: Large language models (LLMs) tend to generate lexically, semantically, and stylistically homogenous texts. This poses a risk of knowledge collapse, where homogenous LLMs mediate a shrinking in the range of accessible information over time. Existing works on homogenization are limited by a focus on closed-ended multiple-choice setups or fuzzy semantic features, and do not look at trends across time and cultural contexts. To overcome this, we present a new methodology to measure epistemic diversity, i.e., variation in real-world claims in LLM outputs, which we use to perform a broad empirical study of LLM knowledge collapse. We test 27 LLMs, 155 topics covering 12 countries, and 200 prompt variations sourced from real user chats. For the topics in our study, we show that while newer models tend to generate more diverse claims, nearly all models are less epistemically diverse than a basic web search. We find that model size has a negative impact on epistemic diversity, while retrieval-augmented generation (RAG) has a positive impact, though the improvement from RAG varies by the cultural context. Finally, compared to a traditional knowledge source (Wikipedia), we find that country-specific claims reflect the English language more than the local one, highlighting a gap in epistemic representation

A Universal Deep Learning Force Field for Molecular Dynamic Simulation and Vibrational Spectra Prediction

arXiv:2510.04227v1 Announce Type: cross Abstract: Accurate and efficient simulation of infrared (IR) and Raman spectra is essential for molecular identification and structural analysis. Traditional quantum chemistry methods based on the harmonic approximation neglect anharmonicity and nuclear quantum effects, while ab initio molecular dynamics (AIMD) remains computationally expensive. Here, we integrate our deep equivariant tensor attention network (DetaNet) with a velocity-Verlet integrator to enable fast and accurate machine learning molecular dynamics (MLMD) simulations for spectral prediction. Trained on the QMe14S dataset containing energies, forces, dipole moments, and polarizabilities for 186,102 small organic molecules, DetaNet yields a universal and transferable force field with high-order tensor prediction capability. Using time-correlation functions derived from MLMD and ring-polymer molecular dynamics (RPMD) trajectories, we computed IR and Raman spectra that accurately reproduce anharmonic and nuclear quantum effects. Benchmark tests on isolated molecules, including polycyclic aromatic hydrocarbons, demonstrate that the DetaNet-based MD approach achieves near-experimental spectral accuracy with speedups up to three orders of magnitude over AIMD. Furthermore, the framework extends seamlessly to molecular and inorganic crystals, molecular aggregates, and biological macromolecules such as polypeptides with minimal fine-tuning. In all systems, DetaNet maintains high accuracy while significantly reducing computational cost. Overall, this work establishes a universal machine learning force field and tensor-aware MLMD framework that enable fast, accurate, and broadly applicable dynamic simulations and IR/Raman spectral predictions across diverse molecular and material systems.

Detection of retinal diseases using an accelerated reused convolutional network

arXiv:2510.04232v1 Announce Type: cross Abstract: Convolutional neural networks are continually evolving, with some efforts aimed at improving accuracy, others at increasing speed, and some at enhancing accessibility. Improving accessibility broadens the application of neural networks across a wider range of tasks, including the detection of eye diseases. Early diagnosis of eye diseases and consulting an ophthalmologist can prevent many vision disorders. Given the importance of this issue, various datasets have been collected from the cornea to facilitate the process of making neural network models. However, most of the methods introduced in the past are computationally complex. In this study, we tried to increase the accessibility of deep neural network models. We did this at the most fundamental level, specifically by redesigning and optimizing the convolutional layers. By doing so, we created a new general model that incorporates our novel convolutional layer named ArConv layers. Thanks to the efficient performance of this new layer, the model has suitable complexity for use in mobile phones and can perform the task of diagnosing the presence of disease with high accuracy. The final model we present contains only 1.3 million parameters. In comparison to the MobileNetV2 model, which has 2.2 million parameters, our model demonstrated better accuracy when trained and evaluated on the RfMiD dataset under identical conditions, achieving an accuracy of 0.9328 versus 0.9266 on the RfMiD test set.

Closing the Loop: Coordinating Inventory and Recommendation via Deep Reinforcement Learning on Multiple Timescales

arXiv:2510.04272v1 Announce Type: cross Abstract: Effective cross-functional coordination is essential for enhancing firm-wide profitability, particularly in the face of growing organizational complexity and scale. Recent advances in artificial intelligence, especially in reinforcement learning (RL), offer promising avenues to address this fundamental challenge. This paper proposes a unified multi-agent RL framework tailored for joint optimization across distinct functional modules, exemplified via coordinating inventory replenishment and personalized product recommendation. We first develop an integrated theoretical model to capture the intricate interplay between these functions and derive analytical benchmarks that characterize optimal coordination. The analysis reveals synchronized adjustment patterns across products and over time, highlighting the importance of coordinated decision-making. Leveraging these insights, we design a novel multi-timescale multi-agent RL architecture that decomposes policy components according to departmental functions and assigns distinct learning speeds based on task complexity and responsiveness. Our model-free multi-agent design improves scalability and deployment flexibility, while multi-timescale updates enhance convergence stability and adaptability across heterogeneous decisions. We further establish the asymptotic convergence of the proposed algorithm. Extensive simulation experiments demonstrate that the proposed approach significantly improves profitability relative to siloed decision-making frameworks, while the behaviors of the trained RL agents align closely with the managerial insights from our theoretical model. Taken together, this work provides a scalable, interpretable RL-based solution to enable effective cross-functional coordination in complex business settings.

Relative Information Gain and Gaussian Process Regression

arXiv:2510.04277v1 Announce Type: cross Abstract: The sample complexity of estimating or maximising an unknown function in a reproducing kernel Hilbert space is known to be linked to both the effective dimension and the information gain associated with the kernel. While the information gain has an attractive information-theoretic interpretation, the effective dimension typically results in better rates. We introduce a new quantity called the relative information gain, which measures the sensitivity of the information gain with respect to the observation noise. We show that the relative information gain smoothly interpolates between the effective dimension and the information gain, and that the relative information gain has the same growth rate as the effective dimension. In the second half of the paper, we prove a new PAC-Bayesian excess risk bound for Gaussian process regression. The relative information gain arises naturally from the complexity term in this PAC-Bayesian bound. We prove bounds on the relative information gain that depend on the spectral properties of the kernel. When these upper bounds are combined with our excess risk bound, we obtain minimax-optimal rates of convergence.

Probing Geometry of Next Token Prediction Using Cumulant Expansion of the Softmax Entropy

arXiv:2510.04285v1 Announce Type: cross Abstract: We introduce a cumulant-expansion framework for quantifying how large language models (LLMs) internalize higher-order statistical structure during next-token prediction. By treating the softmax entropy of each layer's logit distribution as a perturbation around its "center" distribution, we derive closed-form cumulant observables that isolate successively higher-order correlations. Empirically, we track these cumulants in GPT-2 and Pythia models on Pile-10K prompts. (i) Structured prompts exhibit a characteristic rise-and-plateau profile across layers, whereas token-shuffled prompts remain flat, revealing the dependence of the cumulant profile on meaningful context. (ii) During training, all cumulants increase monotonically before saturating, directly visualizing the model's progression from capturing variance to learning skew, kurtosis, and higher-order statistical structures. (iii) Mathematical prompts show distinct cumulant signatures compared to general text, quantifying how models employ fundamentally different processing mechanisms for mathematical versus linguistic content. Together, these results establish cumulant analysis as a lightweight, mathematically grounded probe of feature-learning dynamics in high-dimensional neural networks.

SliceMoE: Routing Embedding Slices Instead of Tokens for Fine-Grained and Balanced Transformer Scaling

arXiv:2510.04286v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) layers scale transformers by routing tokens to a sparse subset of feed-forward experts. Token-level routing, however, assigns an entire semantic spectrum to each expert, creating capacity bottlenecks, load-balancing pathologies, and limited specialization. We introduce SliceMoE, an architecture that routes contiguous slices of a token's hidden vector. A d-dimensional embedding is partitioned into S slices, and for each slice, a lightweight shared router predicts the top-k experts. Experts operate on their assigned slices independently, and outputs are reassembled, maintaining per-token FLOP efficiency. Because slices from different tokens interleave within an expert, utilization is naturally smoother. We propose a slice-level capacity loss, cross-slice dropout, and efficient fused batched GEMM kernels. Experiments on WikiText-103 language modeling, WMT En-De translation, and three text-classification datasets show SliceMoE attains up to 1.7x faster inference than dense baselines, 12 to 18 percent lower perplexity than parameter-matched token-MoE, and improved expert balance, with interpretable expertise over syntactic versus semantic subspaces.

PABSA: Hybrid Framework for Persian Aspect-Based Sentiment Analysis

arXiv:2510.04291v1 Announce Type: cross Abstract: Sentiment analysis is a key task in Natural Language Processing (NLP), enabling the extraction of meaningful insights from user opinions across various domains. However, performing sentiment analysis in Persian remains challenging due to the scarcity of labeled datasets, limited preprocessing tools, and the lack of high-quality embeddings and feature extraction methods. To address these limitations, we propose a hybrid approach that integrates machine learning (ML) and deep learning (DL) techniques for Persian aspect-based sentiment analysis (ABSA). In particular, we utilize polarity scores from multilingual BERT as additional features and incorporate them into a decision tree classifier, achieving an accuracy of 93.34%-surpassing existing benchmarks on the Pars-ABSA dataset. Additionally, we introduce a Persian synonym and entity dictionary, a novel linguistic resource that supports text augmentation through synonym and named entity replacement. Our results demonstrate the effectiveness of hybrid modeling and feature augmentation in advancing sentiment analysis for low-resource languages such as Persian.

On the Importance of Task Complexity in Evaluating LLM-Based Multi-Agent Systems

arXiv:2510.04311v1 Announce Type: cross Abstract: Large language model multi-agent systems (LLM-MAS) offer a promising paradigm for harnessing collective intelligence to achieve more advanced forms of AI behaviour. While recent studies suggest that LLM-MAS can outperform LLM single-agent systems (LLM-SAS) on certain tasks, the lack of systematic experimental designs limits the strength and generality of these conclusions. We argue that a principled understanding of task complexity, such as the degree of sequential reasoning required and the breadth of capabilities involved, is essential for assessing the effectiveness of LLM-MAS in task solving. To this end, we propose a theoretical framework characterising tasks along two dimensions: depth, representing reasoning length, and width, representing capability diversity. We theoretically examine a representative class of LLM-MAS, namely the multi-agent debate system, and empirically evaluate its performance in both discriminative and generative tasks with varying depth and width. Theoretical and empirical results show that the benefit of LLM-MAS over LLM-SAS increases with both task depth and width, and the effect is more pronounced with respect to depth. This clarifies when LLM-MAS are beneficial and provides a principled foundation for designing future LLM-MAS methods and benchmarks.

Adaptive Coverage Policies in Conformal Prediction

arXiv:2510.04318v1 Announce Type: cross Abstract: Traditional conformal prediction methods construct prediction sets such that the true label falls within the set with a user-specified coverage level. However, poorly chosen coverage levels can result in uninformative predictions, either producing overly conservative sets when the coverage level is too high, or empty sets when it is too low. Moreover, the fixed coverage level cannot adapt to the specific characteristics of each individual example, limiting the flexibility and efficiency of these methods. In this work, we leverage recent advances in e-values and post-hoc conformal inference, which allow the use of data-dependent coverage levels while maintaining valid statistical guarantees. We propose to optimize an adaptive coverage policy by training a neural network using a leave-one-out procedure on the calibration set, allowing the coverage level and the resulting prediction set size to vary with the difficulty of each individual example. We support our approach with theoretical coverage guarantees and demonstrate its practical benefits through a series of experiments.

Read the Scene, Not the Script: Outcome-Aware Safety for LLMs

arXiv:2510.04320v1 Announce Type: cross Abstract: Safety-aligned Large Language Models (LLMs) still show two dominant failure modes: they are easily jailbroken, or they over-refuse harmless inputs that contain sensitive surface signals. We trace both to a common cause: current models reason weakly about links between actions and outcomes and over-rely on surface-form signals, lexical or stylistic cues that do not encode consequences. We define this failure mode as Consequence-blindness. To study consequence-blindness, we build a benchmark named CB-Bench covering four risk scenarios that vary whether semantic risk aligns with outcome risk, enabling evaluation under both matched and mismatched conditions which are often ignored by existing safety benchmarks. Mainstream models consistently fail to separate these risks and exhibit consequence-blindness, indicating that consequence-blindness is widespread and systematic. To mitigate consequence-blindness, we introduce CS-Chain-4k, a consequence-reasoning dataset for safety alignment. Models fine-tuned on CS-Chain-4k show clear gains against semantic-camouflage jailbreaks and reduce over-refusal on harmless inputs, while maintaining utility and generalization on other benchmarks. These results clarify the limits of current alignment, establish consequence-aware reasoning as a core alignment goal and provide a more practical and reproducible evaluation path.

Towards Fast Option Pricing PDE Solvers Powered by PIELM

arXiv:2510.04322v1 Announce Type: cross Abstract: Partial differential equation (PDE) solvers underpin modern quantitative finance, governing option pricing and risk evaluation. Physics-Informed Neural Networks (PINNs) have emerged as a promising approach for solving the forward and inverse problems of partial differential equations (PDEs) using deep learning. However they remain computationally expensive due to their iterative gradient descent based optimization and scale poorly with increasing model size. This paper introduces Physics-Informed Extreme Learning Machines (PIELMs) as fast alternative to PINNs for solving both forward and inverse problems in financial PDEs. PIELMs replace iterative optimization with a single least-squares solve, enabling deterministic and efficient training. We benchmark PIELM on the Black-Scholes and Heston-Hull-White models for forward pricing and demonstrate its capability in inverse model calibration to recover volatility and interest rate parameters from noisy data. From experiments we observe that PIELM achieve accuracy comparable to PINNs while being up to $30\times$ faster, highlighting their potential for real-time financial modeling.

Pitch-Conditioned Instrument Sound Synthesis From an Interactive Timbre Latent Space

arXiv:2510.04339v1 Announce Type: cross Abstract: This paper presents a novel approach to neural instrument sound synthesis using a two-stage semi-supervised learning framework capable of generating pitch-accurate, high-quality music samples from an expressive timbre latent space. Existing approaches that achieve sufficient quality for music production often rely on high-dimensional latent representations that are difficult to navigate and provide unintuitive user experiences. We address this limitation through a two-stage training paradigm: first, we train a pitch-timbre disentangled 2D representation of audio samples using a Variational Autoencoder; second, we use this representation as conditioning input for a Transformer-based generative model. The learned 2D latent space serves as an intuitive interface for navigating and exploring the sound landscape. We demonstrate that the proposed method effectively learns a disentangled timbre space, enabling expressive and controllable audio generation with reliable pitch conditioning. Experimental results show the model's ability to capture subtle variations in timbre while maintaining a high degree of pitch accuracy. The usability of our method is demonstrated in an interactive web application, highlighting its potential as a step towards future music production environments that are both intuitive and creatively empowering: https://pgesam.faresschulz.com

Environment-Aware Indoor LoRaWAN Path Loss: Parametric Regression Comparisons, Shadow Fading, and Calibrated Fade Margins

arXiv:2510.04346v1 Announce Type: cross Abstract: Indoor LoRaWAN propagation is shaped by structural and time-varying context factors, which challenge log-distance models and the assumption of log-normal shadowing. We present an environment-aware, statistically disciplined path loss framework evaluated using leakage-safe cross-validation on a 12-month campaign in an eighth-floor office measuring 240 m^2. A log-distance multi-wall mean is augmented with environmental covariates (relative humidity, temperature, carbon dioxide, particulate matter, and barometric pressure), as well as the signal-to-noise ratio. We compare multiple linear regression with regularized variants, Bayesian linear regression, and a selective second-order polynomial applied to continuous drivers. Predictor relevance is established using heteroscedasticity-robust Type II and III analysis of variance and nested partial F tests. Shadow fading is profiled with kernel density estimation and non-parametric families, including Normal, Skew-Normal, Student's t, and Gaussian mixtures. The polynomial mean reduces cross-validated RMSE from 8.07 to 7.09 dB and raises R^2 from 0.81 to 0.86. Out-of-fold residuals are non-Gaussian; a 3-component mixture captures a sharp core with a light, broad tail. We convert accuracy into reliability by prescribing the fade margin as the upper-tail quantile of cross-validated residuals, quantifying uncertainty via a moving-block bootstrap, and validating on a held-out set. At 99% packet delivery ratio, the environment-aware polynomial requires 25.7 dB versus 27.7 to 27.9 dB for linear baselines. This result presents a deployment-ready, interpretable workflow with calibrated reliability control for indoor Internet of Things planning, aligned with 6G targets.

Unmasking Backdoors: An Explainable Defense via Gradient-Attention Anomaly Scoring for Pre-trained Language Models

arXiv:2510.04347v1 Announce Type: cross Abstract: Pre-trained language models have achieved remarkable success across a wide range of natural language processing (NLP) tasks, particularly when fine-tuned on large, domain-relevant datasets. However, they remain vulnerable to backdoor attacks, where adversaries embed malicious behaviors using trigger patterns in the training data. These triggers remain dormant during normal usage, but, when activated, can cause targeted misclassifications. In this work, we investigate the internal behavior of backdoored pre-trained encoder-based language models, focusing on the consistent shift in attention and gradient attribution when processing poisoned inputs; where the trigger token dominates both attention and gradient signals, overriding the surrounding context. We propose an inference-time defense that constructs anomaly scores by combining token-level attention and gradient information. Extensive experiments on text classification tasks across diverse backdoor attack scenarios demonstrate that our method significantly reduces attack success rates compared to existing baselines. Furthermore, we provide an interpretability-driven analysis of the scoring mechanism, shedding light on trigger localization and the robustness of the proposed defense.

Challenge on Optimization of Context Collection for Code Completion

arXiv:2510.04349v1 Announce Type: cross Abstract: The rapid advancement of workflows and methods for software engineering using AI emphasizes the need for a systematic evaluation and analysis of their ability to leverage information from entire projects, particularly in large code bases. In this challenge on optimization of context collection for code completion, organized by JetBrains in collaboration with Mistral AI as part of the ASE 2025 conference, participants developed efficient mechanisms for collecting context from source code repositories to improve fill-in-the-middle code completions for Python and Kotlin. We constructed a large dataset of real-world code in these two programming languages using permissively licensed open-source projects. The submissions were evaluated based on their ability to maximize completion quality for multiple state-of-the-art neural models using the chrF metric. During the public phase of the competition, nineteen teams submitted solutions to the Python track and eight teams submitted solutions to the Kotlin track. In the private phase, six teams competed, of which five submitted papers to the workshop.

Quantizer Design for Finite Model Approximations, Model Learning, and Quantized Q-Learning for MDPs with Unbounded Spaces

arXiv:2510.04355v1 Announce Type: cross Abstract: In this paper, for Markov decision processes (MDPs) with unbounded state spaces we present refined upper bounds presented in [Kara et. al. JMLR'23] on finite model approximation errors via optimizing the quantizers used for finite model approximations. We also consider implications on quantizer design for quantized Q-learning and empirical model learning, and the performance of policies obtained via Q-learning where the quantized state is treated as the state itself. We highlight the distinctions between planning, where approximating MDPs can be independently designed, and learning (either via Q-learning or empirical model learning), where approximating MDPs are restricted to be defined by invariant measures of Markov chains under exploration policies, leading to significant subtleties on quantizer design performance, even though asymptotic near optimality can be established under both setups. In particular, under Lyapunov growth conditions, we obtain explicit upper bounds which decay to zero as the number of bins approaches infinity.

TCR-EML: Explainable Model Layers for TCR-pMHC Prediction

arXiv:2510.04377v1 Announce Type: cross Abstract: T cell receptor (TCR) recognition of peptide-MHC (pMHC) complexes is a central component of adaptive immunity, with implications for vaccine design, cancer immunotherapy, and autoimmune disease. While recent advances in machine learning have improved prediction of TCR-pMHC binding, the most effective approaches are black-box transformer models that cannot provide a rationale for predictions. Post-hoc explanation methods can provide insight with respect to the input but do not explicitly model biochemical mechanisms (e.g. known binding regions), as in TCR-pMHC binding. ``Explain-by-design'' models (i.e., with architectural components that can be examined directly after training) have been explored in other domains, but have not been used for TCR-pMHC binding. We propose explainable model layers (TCR-EML) that can be incorporated into protein-language model backbones for TCR-pMHC modeling. Our approach uses prototype layers for amino acid residue contacts drawn from known TCR-pMHC binding mechanisms, enabling high-quality explanations for predicted TCR-pMHC binding. Experiments of our proposed method on large-scale datasets demonstrate competitive predictive accuracy and generalization, and evaluation on the TCR-XAI benchmark demonstrates improved explainability compared with existing approaches.

Improving Consistency in Retrieval-Augmented Systems with Group Similarity Rewards

arXiv:2510.04392v1 Announce Type: cross Abstract: RAG systems are increasingly deployed in high-stakes domains where users expect outputs to be consistent across semantically equivalent queries. However, existing systems often exhibit significant inconsistencies due to variability in both the retriever and generator (LLM), undermining trust and reliability. In this work, we focus on information consistency, i.e., the requirement that outputs convey the same core content across semantically equivalent inputs. We introduce a principled evaluation framework that decomposes RAG consistency into retriever-level, generator-level, and end-to-end components, helping identify inconsistency sources. To improve consistency, we propose Paraphrased Set Group Relative Policy Optimization (PS-GRPO), an RL approach that leverages multiple rollouts across paraphrased set to assign group similarity rewards. We leverage PS-GRPO to achieve Information Consistent RAG (Con-RAG), training the generator to produce consistent outputs across paraphrased queries and remain robust to retrieval-induced variability. Because exact reward computation over paraphrase sets is computationally expensive, we also introduce a scalable approximation method that retains effectiveness while enabling efficient, large-scale training. Empirical evaluations across short-form, multi-hop, and long-form QA benchmarks demonstrate that Con-RAG significantly improves both consistency and accuracy over strong baselines, even in the absence of explicit ground-truth supervision. Our work provides practical solutions for evaluating and building reliable RAG systems for safety-critical deployments.

Time Is Effort: Estimating Human Post-Editing Time for Grammar Error Correction Tool Evaluation

arXiv:2510.04394v1 Announce Type: cross Abstract: Text editing can involve several iterations of revision. Incorporating an efficient Grammar Error Correction (GEC) tool in the initial correction round can significantly impact further human editing effort and final text quality. This raises an interesting question to quantify GEC Tool usability: How much effort can the GEC Tool save users? We present the first large-scale dataset of post-editing (PE) time annotations and corrections for two English GEC test datasets (BEA19 and CoNLL14). We introduce Post-Editing Effort in Time (PEET) for GEC Tools as a human-focused evaluation scorer to rank any GEC Tool by estimating PE time-to-correct. Using our dataset, we quantify the amount of time saved by GEC Tools in text editing. Analyzing the edit type indicated that determining whether a sentence needs correction and edits like paraphrasing and punctuation changes had the greatest impact on PE time. Finally, comparison with human rankings shows that PEET correlates well with technical effort judgment, providing a new human-centric direction for evaluating GEC tool usability. We release our dataset and code at: https://github.com/ankitvad/PEET_Scorer.

SECA: Semantically Equivalent and Coherent Attacks for Eliciting LLM Hallucinations

arXiv:2510.04398v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly deployed in high-risk domains. However, state-of-the-art LLMs often produce hallucinations, raising serious concerns about their reliability. Prior work has explored adversarial attacks for hallucination elicitation in LLMs, but it often produces unrealistic prompts, either by inserting gibberish tokens or by altering the original meaning. As a result, these approaches offer limited insight into how hallucinations may occur in practice. While adversarial attacks in computer vision often involve realistic modifications to input images, the problem of finding realistic adversarial prompts for eliciting LLM hallucinations has remained largely underexplored. To address this gap, we propose Semantically Equivalent and Coherent Attacks (SECA) to elicit hallucinations via realistic modifications to the prompt that preserve its meaning while maintaining semantic coherence. Our contributions are threefold: (i) we formulate finding realistic attacks for hallucination elicitation as a constrained optimization problem over the input prompt space under semantic equivalence and coherence constraints; (ii) we introduce a constraint-preserving zeroth-order method to effectively search for adversarial yet feasible prompts; and (iii) we demonstrate through experiments on open-ended multiple-choice question answering tasks that SECA achieves higher attack success rates while incurring almost no constraint violations compared to existing methods. SECA highlights the sensitivity of both open-source and commercial gradient-inaccessible LLMs to realistic and plausible prompt variations. Code is available at https://github.com/Buyun-Liang/SECA.

Utility-Learning Tension in Self-Modifying Agents

arXiv:2510.04399v1 Announce Type: cross Abstract: As systems trend toward superintelligence, a natural modeling premise is that agents can self-improve along every facet of their own design. We formalize this with a five-axis decomposition and a decision layer, separating incentives from learning behavior and analyzing axes in isolation. Our central result identifies and introduces a sharp utility--learning tension, the structural conflict in self-modifying systems whereby utility-driven changes that improve immediate or expected performance can also erode the statistical preconditions for reliable learning and generalization. Our findings show that distribution-free guarantees are preserved iff the policy-reachable model family is uniformly capacity-bounded; when capacity can grow without limit, utility-rational self-changes can render learnable tasks unlearnable. Under standard assumptions common in practice, these axes reduce to the same capacity criterion, yielding a single boundary for safe self-modification. Numerical experiments across several axes validate the theory by comparing destructive utility policies against our proposed two-gate policies that preserve learnability.

Modular and Adaptive Conformal Prediction for Sequential Models via Residual Decomposition

arXiv:2510.04406v1 Announce Type: cross Abstract: Conformal prediction offers finite-sample coverage guarantees under minimal assumptions. However, existing methods treat the entire modeling process as a black box, overlooking opportunities to exploit modular structure. We introduce a conformal prediction framework for two-stage sequential models, where an upstream predictor generates intermediate representations for a downstream model. By decomposing the overall prediction residual into stage-specific components, our method enables practitioners to attribute uncertainty to specific pipeline stages. We develop a risk-controlled parameter selection procedure using family-wise error rate (FWER) control to calibrate stage-wise scaling parameters, and propose an adaptive extension for non-stationary settings that preserves long-run coverage guarantees. Experiments on synthetic distribution shifts, as well as real-world supply chain and stock market data, demonstrate that our approach maintains coverage under conditions that degrade standard conformal methods, while providing interpretable stage-wise uncertainty attribution. This framework offers diagnostic advantages and robust coverage that standard conformal methods lack.

Scale-Invariant Regret Matching and Online Learning with Optimal Convergence: Bridging Theory and Practice in Zero-Sum Games

arXiv:2510.04407v1 Announce Type: cross Abstract: A considerable chasm has been looming for decades between theory and practice in zero-sum game solving through first-order methods. Although a convergence rate of $T^{-1}$ has long been established since Nemirovski's mirror-prox algorithm and Nesterov's excessive gap technique in the early 2000s, the most effective paradigm in practice is counterfactual regret minimization, which is based on regret matching and its modern variants. In particular, the state of the art across most benchmarks is predictive regret matching$^+$ (PRM$^+$), in conjunction with non-uniform averaging. Yet, such algorithms can exhibit slower $\Omega(T^{-1/2})$ convergence even in self-play. In this paper, we close the gap between theory and practice. We propose a new scale-invariant and parameter-free variant of PRM$^+$, which we call IREG-PRM$^+$. We show that it achieves $T^{-1/2}$ best-iterate and $T^{-1}$ (i.e., optimal) average-iterate convergence guarantees, while also being on par with PRM$^+$ on benchmark games. From a technical standpoint, we draw an analogy between IREG-PRM$^+$ and optimistic gradient descent with adaptive learning rate. The basic flaw of PRM$^+$ is that the ($\ell_2$-)norm of the regret vector -- which can be thought of as the inverse of the learning rate -- can decrease. By contrast, we design IREG-PRM$^+$ so as to maintain the invariance that the norm of the regret vector is nondecreasing. This enables us to derive an RVU-type bound for IREG-PRM$^+$, the first such property that does not rely on introducing additional hyperparameters to enforce smoothness. Furthermore, we find that IREG-PRM$^+$ performs on par with an adaptive version of optimistic gradient descent that we introduce whose learning rate depends on the misprediction error, demystifying the effectiveness of the regret matching family vis-a-vis more standard optimization techniques.

Learning Survival Models with Right-Censored Reporting Delays

arXiv:2510.04421v1 Announce Type: cross Abstract: Survival analysis is a statistical technique used to estimate the time until an event occurs. Although it is applied across a wide range of fields, adjusting for reporting delays under practical constraints remains a significant challenge in the insurance industry. Such delays render event occurrences unobservable when their reports are subject to right censoring. This issue becomes particularly critical when estimating hazard rates for newly enrolled cohorts with limited follow-up due to administrative censoring. Our study addresses this challenge by jointly modeling the parametric hazard functions of event occurrences and report timings. The joint probability distribution is marginalized over the latent event occurrence status. We construct an estimator for the proposed survival model and establish its asymptotic consistency. Furthermore, we develop an expectation-maximization algorithm to compute its estimates. Using these findings, we propose a two-stage estimation procedure based on a parametric proportional hazards model to evaluate observations subject to administrative censoring. Experimental results demonstrate that our method effectively improves the timeliness of risk evaluation for newly enrolled cohorts.

spd-metrics-id: A Python Package for SPD-Aware Distance Metrics in Connectome Fingerprinting and Beyond

arXiv:2510.04438v1 Announce Type: cross Abstract: We present spd-metrics-id, a Python package for computing distances and divergences between symmetric positive-definite (SPD) matrices. Unlike traditional toolkits that focus on specific applications, spd-metrics-id provides a unified, extensible, and reproducible framework for SPD distance computation. The package supports a wide variety of geometry-aware metrics, including Alpha-z Bures-Wasserstein, Alpha-Procrustes, affine-invariant Riemannian, log-Euclidean, and others, and is accessible both via a command-line interface and a Python API. Reproducibility is ensured through Docker images and Zenodo archiving. We illustrate usage through a connectome fingerprinting example, but the package is broadly applicable to covariance analysis, diffusion tensor imaging, and other domains requiring SPD matrix comparison. The package is openly available at https://pypi.org/project/spd-metrics-id/.

Zeroth-Order Methods for Stochastic Nonconvex Nonsmooth Composite Optimization

arXiv:2510.04446v1 Announce Type: cross Abstract: This work aims to solve a stochastic nonconvex nonsmooth composite optimization problem. Previous works on composite optimization problem requires the major part to satisfy Lipschitz smoothness or some relaxed smoothness conditions, which excludes some machine learning examples such as regularized ReLU network and sparse support matrix machine. In this work, we focus on stochastic nonconvex composite optimization problem without any smoothness assumptions. In particular, we propose two new notions of approximate stationary points for such optimization problem and obtain finite-time convergence results of two zeroth-order algorithms to these two approximate stationary points respectively. Finally, we demonstrate that these algorithms are effective using numerical experiments.

Inverse Mixed-Integer Programming: Learning Constraints then Objective Functions

arXiv:2510.04455v1 Announce Type: cross Abstract: In mixed-integer linear programming, data-driven inverse optimization that learns the objective function and the constraints from observed data plays an important role in constructing appropriate mathematical models for various fields, including power systems and scheduling. However, to the best of our knowledge, there is no known method for learning both the objective functions and the constraints. In this paper, we propose a two-stage method for a class of problems where the objective function is expressed as a linear combination of functions and the constraints are represented by functions and thresholds. Specifically, our method first learns the constraints and then learns the objective function. On the theoretical side, we show the proposed method can solve inverse optimization problems in finite dataset, develop statistical learning theory in pseudometric spaces and sub-Gaussian distributions, and construct a statistical learning for inverse optimization. On the experimental side, we demonstrate that our method is practically applicable for scheduling problems formulated as integer linear programmings with up to 100 decision variables, which are typical in real-world settings.

Perspectives on Stochastic Localization

arXiv:2510.04460v1 Announce Type: cross Abstract: We survey different perspectives on the stochastic localization process of [Eld13], a powerful construction that has had many exciting recent applications in high-dimensional probability and algorithm design. Unlike prior surveys on this topic, our focus is on giving a self-contained presentation of all known alternative constructions of Eldan's stochastic localization, with an emphasis on connections between different constructions. Our hope is that by collecting these perspectives, some of which had primarily arisen within a particular community (e.g., probability theory, theoretical computer science, information theory, or machine learning), we can broaden the accessibility of stochastic localization, and ease its future use.

Benchmarking atmospheric circulation variability in an AI emulator, ACE2, and a hybrid model, NeuralGCM

arXiv:2510.04466v1 Announce Type: cross Abstract: Physics-based atmosphere-land models with prescribed sea surface temperature have notable successes but also biases in their ability to represent atmospheric variability compared to observations. Recently, AI emulators and hybrid models have emerged with the potential to overcome these biases, but still require systematic evaluation against metrics grounded in fundamental atmospheric dynamics. Here, we evaluate the representation of four atmospheric variability benchmarking metrics in a fully data-driven AI emulator (ACE2-ERA5) and hybrid model (NeuralGCM). The hybrid model and emulator can capture the spectra of large-scale tropical waves and extratropical eddy-mean flow interactions, including critical levels. However, both struggle to capture the timescales associated with quasi-biennial oscillation (QBO, $\sim 28$ months) and Southern annular mode propagation ($\sim 150$ days). These dynamical metrics serve as an initial benchmarking tool to inform AI model development and understand their limitations, which may be essential for out-of-distribution applications (e.g., extrapolating to unseen climates).

SPEGNet: Synergistic Perception-Guided Network for Camouflaged Object Detection

arXiv:2510.04472v1 Announce Type: cross Abstract: Camouflaged object detection segments objects with intrinsic similarity and edge disruption. Current detection methods rely on accumulated complex components. Each approach adds components such as boundary modules, attention mechanisms, and multi-scale processors independently. This accumulation creates a computational burden without proportional gains. To manage this complexity, they process at reduced resolutions, eliminating fine details essential for camouflage. We present SPEGNet, addressing fragmentation through a unified design. The architecture integrates multi-scale features via channel calibration and spatial enhancement. Boundaries emerge directly from context-rich representations, maintaining semantic-spatial alignment. Progressive refinement implements scale-adaptive edge modulation with peak influence at intermediate resolutions. This design strikes a balance between boundary precision and regional consistency. SPEGNet achieves 0.887 $S_\alpha$ on CAMO, 0.890 on COD10K, and 0.895 on NC4K, with real-time inference speed. Our approach excels across scales, from tiny, intricate objects to large, pattern-similar ones, while handling occlusion and ambiguous boundaries. Code, model weights, and results are available on \href{https://github.com/Baber-Jan/SPEGNet}{https://github.com/Baber-Jan/SPEGNet}.

DRPO: Efficient Reasoning via Decoupled Reward Policy Optimization

arXiv:2510.04474v1 Announce Type: cross Abstract: Recent large reasoning models (LRMs) driven by reinforcement learning algorithms (e.g., GRPO) have achieved remarkable performance on challenging reasoning tasks. However, these models suffer from overthinking, generating unnecessarily long and redundant reasoning even for simple questions, which substantially increases computational cost and response latency. While existing methods incorporate length rewards to GRPO to promote concise reasoning, they incur significant performance degradation. We identify the root cause: when rewards for correct but long rollouts are penalized, GRPO's group-relative advantage function can assign them negative advantages, actively discouraging valid reasoning. To overcome this, we propose Decoupled Reward Policy Optimization (DRPO), a novel framework that decouples the length-based learning signal of correct rollouts from incorrect ones. DRPO ensures that reward signals for correct rollouts are normalized solely within the positive group, shielding them from interference by negative samples. The DRPO's objective is grounded in integrating an optimized positive data distribution, which maximizes length-based rewards under a KL regularization, into a discriminative objective. We derive a closed-form solution for this distribution, enabling efficient computation of the objective and its gradients using only on-policy data and importance weighting. Of independent interest, this formulation is general and can incorporate other preference rewards of positive data beyond length. Experiments on mathematical reasoning tasks demonstrate DRPO's significant superiority over six efficient reasoning baselines. Notably, with a 1.5B model, our method achieves 77\% length reduction with only 1.1\% performance loss on simple questions like GSM8k dataset, while the follow-up baseline sacrifices 4.3\% for 68\% length reduction.

MedCLM: Learning to Localize and Reason via a CoT-Curriculum in Medical Vision-Language Models

arXiv:2510.04477v1 Announce Type: cross Abstract: Bridging clinical diagnostic reasoning with AI remains a central challenge in medical imaging. We introduce MedCLM, an automated pipeline that converts detection datasets into large-scale medical visual question answering (VQA) data with Chain-of-Thought (CoT) reasoning by linking lesion boxes to organ segmentation and structured rationales. These contextual signals enable medical vision-language models to generate question-answer pairs with step-by-step reasoning. To utilize this data effectively, we propose an Integrated CoT-Curriculum Strategy composed of an Easy stage with explicit lesion boxes for visual grounding, a Medium stage that encourages implicit localization, and a Hard stage for weakly supervised reasoning. Experimental results demonstrate that MedCLM attains state-of-the-art performance on several medical VQA benchmarks, providing a scalable framework for developing clinically aligned medical vision-language models.

Deep vs. Shallow: Benchmarking Physics-Informed Neural Architectures on the Biharmonic Equation

arXiv:2510.04490v1 Announce Type: cross Abstract: Partial differential equation (PDE) solvers are fundamental to engineering simulation. Classical mesh-based approaches (finite difference/volume/element) are fast and accurate on high-quality meshes but struggle with higher-order operators and complex, hard-to-mesh geometries. Recently developed physics-informed neural networks (PINNs) and their variants are mesh-free and flexible, yet compute-intensive and often less accurate. This paper systematically benchmarks RBF-PIELM, a rapid PINN variant-an extreme learning machine with radial-basis activations-for higher-order PDEs. RBF-PIELM replaces PINNs' time-consuming gradient descent with a single-shot least-squares solve. We test RBF-PIELM on the fourth-order biharmonic equation using two benchmarks: lid-driven cavity flow (streamfunction formulation) and a manufactured oscillatory solution. Our results show up to $(350\times)$ faster training than PINNs and over $(10\times)$ fewer parameters for comparable solution accuracy. Despite surpassing PINNs, RBF-PIELM still lags mature mesh-based solvers and its accuracy degrades on highly oscillatory solutions, highlighting remaining challenges for practical deployment.

Causality-aware Graph Aggregation Weight Estimator for Popularity Debiasing in Top-K Recommendation

arXiv:2510.04502v1 Announce Type: cross Abstract: Graph-based recommender systems leverage neighborhood aggregation to generate node representations, which is highly sensitive to popularity bias, resulting in an echo effect during information propagation. Existing graph-based debiasing solutions refine the aggregation process with attempts such as edge reconstruction or weight adjustment. However, these methods remain inadequate in fully alleviating popularity bias. Specifically, this is because 1) they provide no insights into graph aggregation rationality, thus lacking an optimality guarantee; 2) they fail to well balance the training and debiasing process, which undermines the effectiveness. In this paper, we propose a novel approach to mitigate popularity bias through rational modeling of the graph aggregation process. We reveal that graph aggregation is a special form of backdoor adjustment in causal inference, where the aggregation weight corresponds to the historical interaction likelihood distribution. Based on this insight, we devise an encoder-decoder architecture, namely Causality-aware Graph Aggregation Weight Estimator for Debiasing (CAGED), to approximate the unbiased aggregation weight by optimizing the evidence lower bound of the interaction likelihood. In order to enhance the debiasing effectiveness during early training stages, we further design a momentum update strategy that incrementally refines the aggregation weight matrix. Extensive experiments on three datasets demonstrate that CAGED outperforms existing graph-based debiasing methods. Our implementation is available at https://github.com/QueYork/CAGED.

Quantum generative model on bicycle-sharing system and an application

arXiv:2510.04512v1 Announce Type: cross Abstract: Recently, bicycle-sharing systems have been implemented in numerous cities, becoming integral to daily life. However, a prevalent issue arises when intensive commuting demand leads to bicycle shortages in specific areas and at particular times. To address this challenge, we employ a novel quantum machine learning model that analyzes time series data by fitting quantum time evolution to observed sequences. This model enables us to capture actual trends in bicycle counts at individual ports and identify correlations between different ports. Utilizing the trained model, we simulate the impact of proactively adding bicycles to high-demand ports on the overall rental number across the system. Given that the core of this method lies in a Monte Carlo simulation, it is anticipated to have a wide range of industrial applications.

Learning Linear Regression with Low-Rank Tasks in-Context

arXiv:2510.04548v1 Announce Type: cross Abstract: In-context learning (ICL) is a key building block of modern large language models, yet its theoretical mechanisms remain poorly understood. It is particularly mysterious how ICL operates in real-world applications where tasks have a common structure. In this work, we address this problem by analyzing a linear attention model trained on low-rank regression tasks. Within this setting, we precisely characterize the distribution of predictions and the generalization error in the high-dimensional limit. Moreover, we find that statistical fluctuations in finite pre-training data induce an implicit regularization. Finally, we identify a sharp phase transition of the generalization error governed by task structure. These results provide a framework for understanding how transformers learn to learn the task structure.

Fast Witness Persistence for MRI Volumes via Hybrid Landmarking

arXiv:2510.04553v1 Announce Type: cross Abstract: We introduce a scalable witness-based persistent homology pipeline for full-brain MRI volumes that couples density-aware landmark selection with a GPU-ready witness filtration. Candidates are scored by a hybrid metric that balances geometric coverage against inverse kernel density, yielding landmark sets that shrink mean pairwise distances by 30-60% over random or density-only baselines while preserving topological features. Benchmarks on BrainWeb, IXI, and synthetic manifolds execute in under ten seconds on a single NVIDIA RTX 4090 GPU, avoiding the combinatorial blow-up of Cech, Vietoris-Rips, and alpha filtrations. The package is distributed on PyPI as whale-tda (installable via pip); source and issues are hosted at https://github.com/jorgeLRW/whale. The release also exposes a fast preset (mri_deep_dive_fast) for exploratory sweeps, and ships with reproducibility-focused scripts and artifacts for drop-in use in medical imaging workflows.

Gini-based Model Monitoring: A General Framework with an Application to Non-life Insurance Pricing

arXiv:2510.04556v1 Announce Type: cross Abstract: In a dynamic landscape where portfolios and environments evolve, maintaining the accuracy of pricing models is critical. To the best of our knowledge, this is the first study to systematically examine concept drift in non-life insurance pricing. We (i) provide an overview of the relevant literature and commonly used methodologies, clarify the distinction between virtual drift and concept drift, and explain their implications for long-run model performance; (ii) review and formalize common performance measures, including the Gini index and deviance loss, and articulate their interpretation; (iii) derive the asymptotic distribution of the Gini index, enabling valid inference and hypothesis testing; and (iv) present a standardized monitoring procedure that indicates when refitting is warranted. We illustrate the framework using a modified real-world portfolio with induced concept drift and discuss practical considerations and pitfalls.

COSMIR: Chain Orchestrated Structured Memory for Iterative Reasoning over Long Context

arXiv:2510.04568v1 Announce Type: cross Abstract: Reasoning over very long inputs remains difficult for large language models (LLMs). Common workarounds either shrink the input via retrieval (risking missed evidence), enlarge the context window (straining selectivity), or stage multiple agents to read in pieces. In staged pipelines (e.g., Chain of Agents, CoA), free-form summaries passed between agents can discard crucial details and amplify early mistakes. We introduce COSMIR (Chain Orchestrated Structured Memory for Iterative Reasoning), a chain-style framework that replaces ad hoc messages with a structured memory. A Planner agent first turns a user query into concrete, checkable sub-questions. worker agents process chunks via a fixed micro-cycle: Extract, Infer, Refine, writing all updates to the shared memory. A Manager agent then Synthesizes the final answer directly from the memory. This preserves step-wise read-then-reason benefits while changing both the communication medium (structured memory) and the worker procedure (fixed micro-cycle), yielding higher faithfulness, better long-range aggregation, and auditability. On long-context QA from the HELMET suite, COSMIR reduces propagation-stage information loss and improves accuracy over a CoA baseline.

Language Model Based Text-to-Audio Generation: Anti-Causally Aligned Collaborative Residual Transformers

arXiv:2510.04577v1 Announce Type: cross Abstract: While language models (LMs) paired with residual vector quantization (RVQ) tokenizers have shown promise in text-to-audio (T2A) generation, they still lag behind diffusion-based models by a non-trivial margin. We identify a critical dilemma underpinning this gap: incorporating more RVQ layers improves audio reconstruction fidelity but exceeds the generation capacity of conventional LMs. To address this, we first analyze RVQ dynamics and uncover two key limitations: 1) orthogonality of features across RVQ layers hinders effective LMs training, and 2) descending semantic richness in tokens from deeper RVQ layers exacerbates exposure bias during autoregressive decoding. Based on these insights, we propose Siren, a novel LM-based framework that employs multiple isolated transformers with causal conditioning and anti-causal alignment via reinforcement learning. Extensive experiments demonstrate that Siren outperforms both existing LM-based and diffusion-based T2A systems, achieving state-of-the-art results. By bridging the representational strengths of LMs with the fidelity demands of audio synthesis, our approach repositions LMs as competitive contenders against diffusion models in T2A tasks. Moreover, by aligning audio representations with linguistic structures, Siren facilitates a promising pathway toward unified multi-modal generation frameworks.

Data-Driven Adaptive PID Control Based on Physics-Informed Neural Networks

arXiv:2510.04591v1 Announce Type: cross Abstract: This article proposes a data-driven PID controller design based on the principle of adaptive gain optimization, leveraging Physics-Informed Neural Networks (PINNs) generated for predictive modeling purposes. The proposed control design method utilizes gradients of the PID gain optimization, achieved through the automatic differentiation of PINNs, to apply model predictive control using a cost function based on tracking error and control inputs. By optimizing PINNs-based PID gains, the method achieves adaptive gain tuning that ensures stability while accounting for system nonlinearities. The proposed method features a systematic framework for integrating PINNs-based models of dynamical control systems into closed-loop control systems, enabling direct application to PID control design. A series of numerical experiments is conducted to demonstrate the effectiveness of the proposed method from the control perspectives based on both time and frequency domains.

Computing Wasserstein Barycenters through Gradient Flows

arXiv:2510.04602v1 Announce Type: cross Abstract: Wasserstein barycenters provide a powerful tool for aggregating probability measures, while leveraging the geometry of their ambient space. Existing discrete methods suffer from poor scalability, as they require access to the complete set of samples from input measures. We address this issue by recasting the original barycenter problem as a gradient flow in the Wasserstein space. Our approach offers two advantages. First, we achieve scalability by sampling mini-batches from the input measures. Second, we incorporate functionals over probability measures, which regularize the barycenter problem through internal, potential, and interaction energies. We present two algorithms for empirical and Gaussian mixture measures, providing convergence guarantees under the Polyak-{\L}ojasiewicz inequality. Experimental validation on toy datasets and domain adaptation benchmarks show that our methods outperform previous discrete and neural net-based methods for computing Wasserstein barycenters.

A Case for Declarative LLM-friendly Interfaces for Improved Efficiency of Computer-Use Agents

arXiv:2510.04607v1 Announce Type: cross Abstract: Computer-use agents (CUAs) powered by large language models (LLMs) have emerged as a promising approach to automating computer tasks, yet they struggle with graphical user interfaces (GUIs). GUIs, designed for humans, force LLMs to decompose high-level goals into lengthy, error-prone sequences of fine-grained actions, resulting in low success rates and an excessive number of LLM calls. We propose Goal-Oriented Interface (GOI), a novel abstraction that transforms existing GUIs into three declarative primitives: access, state, and observation, which are better suited for LLMs. Our key idea is policy-mechanism separation: LLMs focus on high-level semantic planning (policy) while GOI handles low-level navigation and interaction (mechanism). GOI does not require modifying the application source code or relying on application programming interfaces (APIs). We evaluate GOI with Microsoft Office Suite (Word, PowerPoint, Excel) on Windows. Compared to a leading GUI-based agent baseline, GOI improves task success rates by 67% and reduces interaction steps by 43.5%. Notably, GOI completes over 61% of successful tasks with a single LLM call.

Fairness in Repeated Matching: A Maximin Perspective

arXiv:2510.04624v1 Announce Type: cross Abstract: We study a sequential decision-making model where a set of items is repeatedly matched to the same set of agents over multiple rounds. The objective is to determine a sequence of matchings that either maximizes the utility of the least advantaged agent at the end of all rounds (optimal) or at the end of every individual round (anytime optimal). We investigate the computational challenges associated with finding (anytime) optimal outcomes and demonstrate that these problems are generally computationally intractable. However, we provide approximation algorithms, fixed-parameter tractable algorithms, and identify several special cases whereby the problem(s) can be solved efficiently. Along the way, we also establish characterizations of Pareto-optimal/maximum matchings, which may be of independent interest to works in matching theory and house allocation.

Evaluating LLMs for Demographic-Targeted Social Bias Detection: A Comprehensive Benchmark Study

arXiv:2510.04641v1 Announce Type: cross Abstract: Large-scale web-scraped text corpora used to train general-purpose AI models often contain harmful demographic-targeted social biases, creating a regulatory need for data auditing and developing scalable bias-detection methods. Although prior work has investigated biases in text datasets and related detection methods, these studies remain narrow in scope. They typically focus on a single content type (e.g., hate speech), cover limited demographic axes, overlook biases affecting multiple demographics simultaneously, and analyze limited techniques. Consequently, practitioners lack a holistic understanding of the strengths and limitations of recent large language models (LLMs) for automated bias detection. In this study, we present a comprehensive evaluation framework aimed at English texts to assess the ability of LLMs in detecting demographic-targeted social biases. To align with regulatory requirements, we frame bias detection as a multi-label task using a demographic-focused taxonomy. We then conduct a systematic evaluation with models across scales and techniques, including prompting, in-context learning, and fine-tuning. Using twelve datasets spanning diverse content types and demographics, our study demonstrates the promise of fine-tuned smaller models for scalable detection. However, our analyses also expose persistent gaps across demographic axes and multi-demographic targeted biases, underscoring the need for more effective and scalable auditing frameworks.

A Study on the Data Distribution Gap in Music Emotion Recognition

arXiv:2510.04688v1 Announce Type: cross Abstract: Music Emotion Recognition (MER) is a task deeply connected to human perception, relying heavily on subjective annotations collected from contributors. Prior studies tend to focus on specific musical styles rather than incorporating a diverse range of genres, such as rock and classical, within a single framework. In this paper, we address the task of recognizing emotion from audio content by investigating five datasets with dimensional emotion annotations -- EmoMusic, DEAM, PMEmo, WTC, and WCMED -- which span various musical styles. We demonstrate the problem of out-of-distribution generalization in a systematic experiment. By closely looking at multiple data and feature sets, we provide insight into genre-emotion relationships in existing data and examine potential genre dominance and dataset biases in certain feature representations. Based on these experiments, we arrive at a simple yet effective framework that combines embeddings extracted from the Jukebox model with chroma features and demonstrate how, alongside a combination of several diverse training sets, this permits us to train models with substantially improved cross-dataset generalization capabilities.

Multilingual Routing in Mixture-of-Experts

arXiv:2510.04694v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) architectures have become the key to scaling modern LLMs, yet little is understood about how their sparse routing dynamics respond to multilingual data. In this work, we analyze expert routing patterns using parallel multilingual datasets and present highly interpretable layer-wise phenomena. We find that MoE models route tokens in language-specific ways in the early and late decoder layers but exhibit significant cross-lingual routing alignment in middle layers, mirroring parameter-sharing trends observed in dense LLMs. In particular, we reveal a clear, strong correlation between a model's performance in a given language and how similarly its tokens are routed to English in these layers. Extending beyond correlation, we explore inference-time interventions that induce higher cross-lingual routing alignment. We introduce a method that steers the router by promoting middle-layer task experts frequently activated in English, and it successfully increases multilingual performance. These 1-2% gains are remarkably consistent across two evaluation tasks, three models, and 15+ languages, especially given that these simple interventions override routers of extensively trained, state-of-the-art LLMs. In comparison, interventions outside of the middle layers or targeting multilingual-specialized experts only yield performance degradation. Altogether, we present numerous findings that explain how MoEs process non-English text and demonstrate that generalization is limited by the model's ability to leverage language-universal experts in all languages.

BrokenMath: A Benchmark for Sycophancy in Theorem Proving with LLMs

arXiv:2510.04721v1 Announce Type: cross Abstract: Large language models (LLMs) have recently shown strong performance on mathematical benchmarks. At the same time, they are prone to hallucination and sycophancy, often providing convincing but flawed proofs for incorrect mathematical statements provided by users. This significantly limits the applicability of LLMs in theorem proving, as verification of these flawed proofs must be done manually by expert mathematicians. However, existing benchmarks that measure sycophancy in mathematics are limited: they focus solely on final-answer problems, rely on very simple and often contaminated datasets, and construct benchmark samples using synthetic modifications that create ill-posed questions rather than well-posed questions that are demonstrably false. To address these issues, we introduce BrokenMath, the first benchmark for evaluating sycophantic behavior in LLMs within the context of natural language theorem proving. BrokenMath is built from advanced 2025 competition problems, which are perturbed with an LLM to produce false statements and subsequently refined through expert review. Using an LLM-as-a-judge framework, we evaluate state-of-the-art LLMs and agentic systems and find that sycophancy is widespread, with the best model, GPT-5, producing sycophantic answers 29% of the time. We further investigate several mitigation strategies, including test-time interventions and supervised fine-tuning on curated sycophantic examples. These approaches substantially reduce, but do not eliminate, sycophantic behavior.

Predictive economics: Rethinking economic methodology with machine learning

arXiv:2510.04726v1 Announce Type: cross Abstract: This article proposes predictive economics as a distinct analytical perspective within economics, grounded in machine learning and centred on predictive accuracy rather than causal identification. Drawing on the instrumentalist tradition (Friedman), the explanation-prediction divide (Shmueli), and the contrast between modelling cultures (Breiman), we formalise prediction as a valid epistemological and methodological objective. Reviewing recent applications across economic subfields, we show how predictive models contribute to empirical analysis, particularly in complex or data-rich contexts. This perspective complements existing approaches and supports a more pluralistic methodology - one that values out-of-sample performance alongside interpretability and theoretical structure.

Speak, Edit, Repeat: High-Fidelity Voice Editing and Zero-Shot TTS with Cross-Attentive Mamba

arXiv:2510.04738v1 Announce Type: cross Abstract: We introduce MAVE (Mamba with Cross-Attention for Voice Editing and Synthesis), a novel autoregressive architecture for text-conditioned voice editing and high-fidelity text-to-speech (TTS) synthesis, built on a cross-attentive Mamba backbone. MAVE achieves state-of-the-art performance in speech editing and very competitive results in zero-shot TTS, while not being explicitly trained on the latter task, outperforming leading autoregressive and diffusion models on diverse, real-world audio. By integrating Mamba for efficient audio sequence modeling with cross-attention for precise text-acoustic alignment, MAVE enables context-aware voice editing with exceptional naturalness and speaker consistency. In pairwise human evaluations on a random 40-sample subset of the RealEdit benchmark (400 judgments), 57.2% of listeners rated MAVE - edited speech as perceptually equal to the original, while 24.8% prefered the original and 18.0% MAVE - demonstrating that in the majority of cases edits are indistinguishable from the source. MAVE compares favorably with VoiceCraft and FluentSpeech both on pairwise comparisons and standalone mean opinion score (MOS) evaluations. For zero-shot TTS, MAVE exceeds VoiceCraft in both speaker similarity and naturalness, without requiring multiple inference runs or post-processing. Remarkably, these quality gains come with a significantly lower memory cost and approximately the same latency: MAVE requires ~6x less memory than VoiceCraft during inference on utterances from the RealEdit database (mean duration: 6.21s, A100, FP16, batch size 1). Our results demonstrate that MAVE establishes a new standard for flexible, high-fidelity voice editing and synthesis through the synergistic integration of structured state-space modeling and cross-modal attention.

Fisher-Bingham-like normalizing flows on the sphere

arXiv:2510.04762v1 Announce Type: cross Abstract: A generic D-dimensional Gaussian can be conditioned or projected onto the D-1 unit sphere, thereby leading to the well-known Fisher-Bingham (FB) or Angular Gaussian (AG) distribution families, respectively. These are some of the most fundamental distributions on the sphere, yet cannot straightforwardly be written as a normalizing flow except in two special cases: the von-Mises Fisher in D=3 and the central angular Gaussian in any D. In this paper, we describe how to generalize these special cases to a family of normalizing flows that behave similarly to the full FB or AG family in any D. We call them "zoom-linear-project" (ZLP)-Fisher flows. Unlike a normal Fisher-Bingham distribution, their composition allows to gradually add complexity as needed. Furthermore, they can naturally handle conditional density estimation with target distributions that vary by orders of magnitude in scale - a setting that is important in astronomical applications but that existing flows often struggle with. A particularly useful member of the new family is the Kent analogue that can cheaply upgrade any flow in this situation to yield better performance.

Beyond the Seen: Bounded Distribution Estimation for Open-Vocabulary Learning

arXiv:2510.04770v1 Announce Type: cross Abstract: Open-vocabulary learning requires modeling the data distribution in open environments, which consists of both seen-class and unseen-class data. Existing methods estimate the distribution in open environments using seen-class data, where the absence of unseen classes makes the estimation error inherently unidentifiable. Intuitively, learning beyond the seen classes is crucial for distribution estimation to bound the estimation error. We theoretically demonstrate that the distribution can be effectively estimated by generating unseen-class data, through which the estimation error is upper-bounded. Building on this theoretical insight, we propose a novel open-vocabulary learning method, which generates unseen-class data for estimating the distribution in open environments. The method consists of a class-domain-wise data generation pipeline and a distribution alignment algorithm. The data generation pipeline generates unseen-class data under the guidance of a hierarchical semantic tree and domain information inferred from the seen-class data, facilitating accurate distribution estimation. With the generated data, the distribution alignment algorithm estimates and maximizes the posterior probability to enhance generalization in open-vocabulary learning. Extensive experiments on $11$ datasets demonstrate that our method outperforms baseline approaches by up to $14\%$, highlighting its effectiveness and superiority.

Federated Learning for Surgical Vision in Appendicitis Classification: Results of the FedSurg EndoVis 2024 Challenge

arXiv:2510.04772v1 Announce Type: cross Abstract: Purpose: The FedSurg challenge was designed to benchmark the state of the art in federated learning for surgical video classification. Its goal was to assess how well current methods generalize to unseen clinical centers and adapt through local fine-tuning while enabling collaborative model development without sharing patient data. Methods: Participants developed strategies to classify inflammation stages in appendicitis using a preliminary version of the multi-center Appendix300 video dataset. The challenge evaluated two tasks: generalization to an unseen center and center-specific adaptation after fine-tuning. Submitted approaches included foundation models with linear probing, metric learning with triplet loss, and various FL aggregation schemes (FedAvg, FedMedian, FedSAM). Performance was assessed using F1-score and Expected Cost, with ranking robustness evaluated via bootstrapping and statistical testing. Results: In the generalization task, performance across centers was limited. In the adaptation task, all teams improved after fine-tuning, though ranking stability was low. The ViViT-based submission achieved the strongest overall performance. The challenge highlighted limitations in generalization, sensitivity to class imbalance, and difficulties in hyperparameter tuning in decentralized training, while spatiotemporal modeling and context-aware preprocessing emerged as promising strategies. Conclusion: The FedSurg Challenge establishes the first benchmark for evaluating FL strategies in surgical video classification. Findings highlight the trade-off between local personalization and global robustness, and underscore the importance of architecture choice, preprocessing, and loss design. This benchmarking offers a reference point for future development of imbalance-aware, adaptive, and robust FL methods in clinical surgical AI.

Kernel ridge regression under power-law data: spectrum and generalization

arXiv:2510.04780v1 Announce Type: cross Abstract: In this work, we investigate high-dimensional kernel ridge regression (KRR) on i.i.d. Gaussian data with anisotropic power-law covariance. This setting differs fundamentally from the classical source & capacity conditions for KRR, where power-law assumptions are typically imposed on the kernel eigen-spectrum itself. Our contributions are twofold. First, we derive an explicit characterization of the kernel spectrum for polynomial inner-product kernels, giving a precise description of how the kernel eigen-spectrum inherits the data decay. Second, we provide an asymptotic analysis of the excess risk in the high-dimensional regime for a particular kernel with this spectral behavior, showing that the sample complexity is governed by the effective dimension of the data rather than the ambient dimension. These results establish a fundamental advantage of learning with power-law anisotropic data over isotropic data. To our knowledge, this is the first rigorous treatment of non-linear KRR under power-law data.

A Noise Resilient Approach for Robust Hurst Exponent Estimation

arXiv:2510.04811v1 Announce Type: cross Abstract: Understanding signal behavior across scales is vital in areas such as natural phenomena analysis and financial modeling. A key property is self-similarity, quantified by the Hurst exponent (H), which reveals long-term dependencies. Wavelet-based methods are effective for estimating H due to their multi-scale analysis capability, but additive noise in real-world measurements often degrades accuracy. We propose Noise-Controlled ALPHEE (NC-ALPHEE), an enhancement of the Average Level-Pairwise Hurst Exponent Estimator (ALPHEE), incorporating noise mitigation and generating multiple level-pairwise estimates from signal energy pairs. A neural network (NN) combines these estimates, replacing traditional averaging. This adaptive learning maintains ALPHEE's behavior in noise-free cases while improving performance in noisy conditions. Extensive simulations show that in noise-free data, NC-ALPHEE matches ALPHEE's accuracy using both averaging and NN-based methods. Under noise, however, traditional averaging deteriorates and requires impractical level restrictions, while NC-ALPHEE consistently outperforms existing techniques without such constraints. NC-ALPHEE offers a robust, adaptive approach for H estimation, significantly enhancing the reliability of wavelet-based methods in noisy environments.

Beyond Random: Automatic Inner-loop Optimization in Dataset Distillation

arXiv:2510.04838v1 Announce Type: cross Abstract: The growing demand for efficient deep learning has positioned dataset distillation as a pivotal technique for compressing training dataset while preserving model performance. However, existing inner-loop optimization methods for dataset distillation typically rely on random truncation strategies, which lack flexibility and often yield suboptimal results. In this work, we observe that neural networks exhibit distinct learning dynamics across different training stages-early, middle, and late-making random truncation ineffective. To address this limitation, we propose Automatic Truncated Backpropagation Through Time (AT-BPTT), a novel framework that dynamically adapts both truncation positions and window sizes according to intrinsic gradient behavior. AT-BPTT introduces three key components: (1) a probabilistic mechanism for stage-aware timestep selection, (2) an adaptive window sizing strategy based on gradient variation, and (3) a low-rank Hessian approximation to reduce computational overhead. Extensive experiments on CIFAR-10, CIFAR-100, Tiny-ImageNet, and ImageNet-1K show that AT-BPTT achieves state-of-the-art performance, improving accuracy by an average of 6.16% over baseline methods. Moreover, our approach accelerates inner-loop optimization by 3.9x while saving 63% memory cost.

LEGOMem: Modular Procedural Memory for Multi-agent LLM Systems for Workflow Automation

arXiv:2510.04851v1 Announce Type: cross Abstract: We introduce LEGOMem, a modular procedural memory framework for multi-agent large language model (LLM) systems in workflow automation. LEGOMem decomposes past task trajectories into reusable memory units and flexibly allocates them across orchestrators and task agents to support planning and execution. To explore the design space of memory in multi-agent systems, we use LEGOMem as a lens and conduct a systematic study of procedural memory in multi-agent systems, examining where memory should be placed, how it should be retrieved, and which agents benefit most. Experiments on the OfficeBench benchmark show that orchestrator memory is critical for effective task decomposition and delegation, while fine-grained agent memory improves execution accuracy. We find that even teams composed of smaller language models can benefit substantially from procedural memory, narrowing the performance gap with stronger agents by leveraging prior execution traces for more accurate planning and tool use. These results position LEGOMem as both a practical framework for memory-augmented agent systems and a research tool for understanding memory design in multi-agent workflow automation.

ERDE: Entropy-Regularized Distillation for Early-exit

arXiv:2510.04856v1 Announce Type: cross Abstract: Although deep neural networks and in particular Convolutional Neural Networks have demonstrated state-of-the-art performance in image classification with relatively high efficiency, they still exhibit high computational costs, often rendering them impractical for real-time and edge applications. Therefore, a multitude of compression techniques have been developed to reduce these costs while maintaining accuracy. In addition, dynamic architectures have been introduced to modulate the level of compression at execution time, which is a desirable property in many resource-limited application scenarios. The proposed method effectively integrates two well-established optimization techniques: early exits and knowledge distillation, where a reduced student early-exit model is trained from a more complex teacher early-exit model. The primary contribution of this research lies in the approach for training the student early-exit model. In comparison to the conventional Knowledge Distillation loss, our approach incorporates a new entropy-based loss for images where the teacher's classification was incorrect. The proposed method optimizes the trade-off between accuracy and efficiency, thereby achieving significant reductions in computational complexity without compromising classification performance. The validity of this approach is substantiated by experimental results on image classification datasets CIFAR10, CIFAR100 and SVHN, which further opens new research perspectives for Knowledge Distillation in other contexts.

Video Game Level Design as a Multi-Agent Reinforcement Learning Problem

arXiv:2510.04862v1 Announce Type: cross Abstract: Procedural Content Generation via Reinforcement Learning (PCGRL) offers a method for training controllable level designer agents without the need for human datasets, using metrics that serve as proxies for level quality as rewards. Existing PCGRL research focuses on single generator agents, but are bottlenecked by the need to frequently recalculate heuristics of level quality and the agent's need to navigate around potentially large maps. By framing level generation as a multi-agent problem, we mitigate the efficiency bottleneck of single-agent PCGRL by reducing the number of reward calculations relative to the number of agent actions. We also find that multi-agent level generators are better able to generalize to out-of-distribution map shapes, which we argue is due to the generators' learning more local, modular design policies. We conclude that treating content generation as a distributed, multi-agent task is beneficial for generating functional artifacts at scale.

BenthiCat: An opti-acoustic dataset for advancing benthic classification and habitat mapping

arXiv:2510.04876v1 Announce Type: cross Abstract: Benthic habitat mapping is fundamental for understanding marine ecosystems, guiding conservation efforts, and supporting sustainable resource management. Yet, the scarcity of large, annotated datasets limits the development and benchmarking of machine learning models in this domain. This paper introduces a thorough multi-modal dataset, comprising about a million side-scan sonar (SSS) tiles collected along the coast of Catalonia (Spain), complemented by bathymetric maps and a set of co-registered optical images from targeted surveys using an autonomous underwater vehicle (AUV). Approximately \num{36000} of the SSS tiles have been manually annotated with segmentation masks to enable supervised fine-tuning of classification models. All the raw sensor data, together with mosaics, are also released to support further exploration and algorithm development. To address challenges in multi-sensor data fusion for AUVs, we spatially associate optical images with corresponding SSS tiles, facilitating self-supervised, cross-modal representation learning. Accompanying open-source preprocessing and annotation tools are provided to enhance accessibility and encourage research. This resource aims to establish a standardized benchmark for underwater habitat mapping, promoting advancements in autonomous seafloor classification and multi-sensor integration.

CLEAR-IR: Clarity-Enhanced Active Reconstruction of Infrared Imagery

arXiv:2510.04883v1 Announce Type: cross Abstract: This paper presents a novel approach for enabling robust robotic perception in dark environments using infrared (IR) stream. IR stream is less susceptible to noise than RGB in low-light conditions. However, it is dominated by active emitter patterns that hinder high-level tasks such as object detection, tracking and localisation. To address this, a U-Net-based architecture is proposed that reconstructs clean IR images from emitter-populated input, improving both image quality and downstream robotic performance. This approach outperforms existing enhancement techniques and enables reliable operation of vision-driven robotic systems across illumination conditions from well-lit to extreme low-light scenes.

RL Is a Hammer and LLMs Are Nails: A Simple Reinforcement Learning Recipe for Strong Prompt Injection

arXiv:2510.04885v1 Announce Type: cross Abstract: Prompt injection poses a serious threat to the reliability and safety of LLM agents. Recent defenses against prompt injection, such as Instruction Hierarchy and SecAlign, have shown notable robustness against static attacks. However, to more thoroughly evaluate the robustness of these defenses, it is arguably necessary to employ strong attacks such as automated red-teaming. To this end, we introduce RL-Hammer, a simple recipe for training attacker models that automatically learn to perform strong prompt injections and jailbreaks via reinforcement learning. RL-Hammer requires no warm-up data and can be trained entirely from scratch. To achieve high ASRs against industrial-level models with defenses, we propose a set of practical techniques that enable highly effective, universal attacks. Using this pipeline, RL-Hammer reaches a 98% ASR against GPT-4o and a $72\%$ ASR against GPT-5 with the Instruction Hierarchy defense. We further discuss the challenge of achieving high diversity in attacks, highlighting how attacker models tend to reward-hack diversity objectives. Finally, we show that RL-Hammer can evade multiple prompt injection detectors. We hope our work advances automatic red-teaming and motivates the development of stronger, more principled defenses. Code is available at https://github.com/facebookresearch/rl-injector.

SocialHarmBench: Revealing LLM Vulnerabilities to Socially Harmful Requests

arXiv:2510.04891v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly deployed in contexts where their failures can have direct sociopolitical consequences. Yet, existing safety benchmarks rarely test vulnerabilities in domains such as political manipulation, propaganda and disinformation generation, or surveillance and information control. We introduce SocialHarmBench, a dataset of 585 prompts spanning 7 sociopolitical categories and 34 countries, designed to surface where LLMs most acutely fail in politically charged contexts. Our evaluations reveal several shortcomings: open-weight models exhibit high vulnerability to harmful compliance, with Mistral-7B reaching attack success rates as high as 97% to 98% in domains such as historical revisionism, propaganda, and political manipulation. Moreover, temporal and geographic analyses show that LLMs are most fragile when confronted with 21st-century or pre-20th-century contexts, and when responding to prompts tied to regions such as Latin America, the USA, and the UK. These findings demonstrate that current safeguards fail to generalize to high-stakes sociopolitical settings, exposing systematic biases and raising concerns about the reliability of LLMs in preserving human rights and democratic values. We share the SocialHarmBench benchmark at https://huggingface.co/datasets/psyonp/SocialHarmBench.

HyperVLA: Efficient Inference in Vision-Language-Action Models via Hypernetworks

arXiv:2510.04898v1 Announce Type: cross Abstract: Built upon language and vision foundation models with strong generalization ability and trained on large-scale robotic data, Vision-Language-Action (VLA) models have recently emerged as a promising approach to learning generalist robotic policies. However, a key drawback of existing VLAs is their extremely high inference costs. In this paper, we propose HyperVLA to address this problem. Unlike existing monolithic VLAs that activate the whole model during both training and inference, HyperVLA uses a novel hypernetwork (HN)-based architecture that activates only a small task-specific policy during inference, while still retaining the high model capacity needed to accommodate diverse multi-task behaviors during training. Successfully training an HN-based VLA is nontrivial so HyperVLA contains several key algorithm design features that improve its performance, including properly utilizing the prior knowledge from existing vision foundation models, HN normalization, and an action generation strategy. Compared to monolithic VLAs, HyperVLA achieves a similar or even higher success rate for both zero-shot generalization and few-shot adaptation, while significantly reducing inference costs. Compared to OpenVLA, a state-of-the-art VLA model, HyperVLA reduces the number of activated parameters at test time by $90\times$, and accelerates inference speed by $120\times$. Code is publicly available at https://github.com/MasterXiong/HyperVLA

Comparative Analysis of YOLOv5, Faster R-CNN, SSD, and RetinaNet for Motorbike Detection in Kigali Autonomous Driving Context

arXiv:2510.04912v1 Announce Type: cross Abstract: In Kigali, Rwanda, motorcycle taxis are a primary mode of transportation, often navigating unpredictably and disregarding traffic rules, posing significant challenges for autonomous driving systems. This study compares four object detection models--YOLOv5, Faster R-CNN, SSD, and RetinaNet--for motorbike detection using a custom dataset of 198 images collected in Kigali. Implemented in PyTorch with transfer learning, the models were evaluated for accuracy, localization, and inference speed to assess their suitability for real-time navigation in resource-constrained settings. We identify implementation challenges, including dataset limitations and model complexities, and recommend simplified architectures for future work to enhance accessibility for autonomous systems in developing countries like Rwanda.

Set to Be Fair: Demographic Parity Constraints for Set-Valued Classification

arXiv:2510.04926v1 Announce Type: cross Abstract: Set-valued classification is used in multiclass settings where confusion between classes can occur and lead to misleading predictions. However, its application may amplify discriminatory bias motivating the development of set-valued approaches under fairness constraints. In this paper, we address the problem of set-valued classification under demographic parity and expected size constraints. We propose two complementary strategies: an oracle-based method that minimizes classification risk while satisfying both constraints, and a computationally efficient proxy that prioritizes constraint satisfaction. For both strategies, we derive closed-form expressions for the (optimal) fair set-valued classifiers and use these to build plug-in, data-driven procedures for empirical predictions. We establish distribution-free convergence rates for violations of the size and fairness constraints for both methods, and under mild assumptions we also provide excess-risk bounds for the oracle-based approach. Empirical results demonstrate the effectiveness of both strategies and highlight the efficiency of our proxy method.

The Geometry of Truth: Layer-wise Semantic Dynamics for Hallucination Detection in Large Language Models

arXiv:2510.04933v1 Announce Type: cross Abstract: Large Language Models (LLMs) often produce fluent yet factually incorrect statements-a phenomenon known as hallucination-posing serious risks in high-stakes domains. We present Layer-wise Semantic Dynamics (LSD), a geometric framework for hallucination detection that analyzes the evolution of hidden-state semantics across transformer layers. Unlike prior methods that rely on multiple sampling passes or external verification sources, LSD operates intrinsically within the model's representational space. Using margin-based contrastive learning, LSD aligns hidden activations with ground-truth embeddings derived from a factual encoder, revealing a distinct separation in semantic trajectories: factual responses preserve stable alignment, while hallucinations exhibit pronounced semantic drift across depth. Evaluated on the TruthfulQA and synthetic factual-hallucination datasets, LSD achieves an F1-score of 0.92, AUROC of 0.96, and clustering accuracy of 0.89, outperforming SelfCheckGPT and Semantic Entropy baselines while requiring only a single forward pass. This efficiency yields a 5-20x speedup over sampling-based methods without sacrificing precision or interpretability. LSD offers a scalable, model-agnostic mechanism for real-time hallucination monitoring and provides new insights into the geometry of factual consistency within large language models.

MARS: Optimizing Dual-System Deep Research via Multi-Agent Reinforcement Learning

arXiv:2510.04935v1 Announce Type: cross Abstract: Large Reasoning Models (LRMs) often exhibit a tendency for overanalysis in simple tasks, where the models excessively utilize System 2-type, deliberate reasoning, leading to inefficient token generation. Furthermore, these models face challenges in adapting their reasoning capabilities to rapidly changing environments due to the static nature of their pretraining data. To address these issues, advancing Large Language Models (LLMs) for complex reasoning tasks requires innovative approaches that bridge intuitive and deliberate cognitive processes, akin to human cognition's dual-system dynamic. This paper introduces a Multi-Agent System for Deep ReSearch (MARS) enabling seamless integration of System 1's fast, intuitive thinking with System 2's deliberate reasoning within LLMs. MARS strategically integrates multiple external tools, such as Google Search, Google Scholar, and Python Interpreter, to access up-to-date information and execute complex computations, while creating a specialized division of labor where System 1 efficiently processes and summarizes high-volume external information, providing distilled insights that expand System 2's reasoning context without overwhelming its capacity. Furthermore, we propose a multi-agent reinforcement learning framework extending Group Relative Policy Optimization to simultaneously optimize both systems with multi-turn tool interactions, bin-packing optimization, and sample balancing strategies that enhance collaborative efficiency. Extensive experiments demonstrate MARS achieves substantial improvements of 3.86% on the challenging Humanity's Last Exam (HLE) benchmark and an average gain of 8.9% across 7 knowledge-intensive tasks, validating the effectiveness of our dual-system paradigm for complex reasoning in dynamic information environments.

Unsupervised Active Learning via Natural Feature Progressive Framework

arXiv:2510.04939v1 Announce Type: cross Abstract: The effectiveness of modern deep learning models is predicated on the availability of large-scale, human-annotated datasets, a process that is notoriously expensive and time-consuming. While Active Learning (AL) offers a strategic solution by labeling only the most informative and representative data, its iterative nature still necessitates significant human involvement. Unsupervised Active Learning (UAL) presents an alternative by shifting the annotation burden to a single, post-selection step. Unfortunately, prevailing UAL methods struggle to achieve state-of-the-art performance. These approaches typically rely on local, gradient-based scoring for sample importance estimation, which not only makes them vulnerable to ambiguous and noisy data but also hinders their capacity to select samples that adequately represent the full data distribution. Moreover, their use of shallow, one-shot linear selection falls short of a true UAL paradigm. In this paper, we propose the Natural Feature Progressive Framework (NFPF), a UAL method that revolutionizes how sample importance is measured. At its core, NFPF employs a Specific Feature Learning Machine (SFLM) to effectively quantify each sample's contribution to model performance. We further utilize the SFLM to define a powerful Reconstruction Difference metric for initial sample selection. Our comprehensive experiments show that NFPF significantly outperforms all established UAL methods and achieves performance on par with supervised AL methods on vision datasets. Detailed ablation studies and qualitative visualizations provide compelling evidence for NFPF's superior performance, enhanced robustness, and improved data distribution coverage.

Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)

arXiv:2510.04950v1 Announce Type: cross Abstract: The wording of natural language prompts has been shown to influence the performance of large language models (LLMs), yet the role of politeness and tone remains underexplored. In this study, we investigate how varying levels of prompt politeness affect model accuracy on multiple-choice questions. We created a dataset of 50 base questions spanning mathematics, science, and history, each rewritten into five tone variants: Very Polite, Polite, Neutral, Rude, and Very Rude, yielding 250 unique prompts. Using ChatGPT 4o, we evaluated responses across these conditions and applied paired sample t-tests to assess statistical significance. Contrary to expectations, impolite prompts consistently outperformed polite ones, with accuracy ranging from 80.8% for Very Polite prompts to 84.8% for Very Rude prompts. These findings differ from earlier studies that associated rudeness with poorer outcomes, suggesting that newer LLMs may respond differently to tonal variation. Our results highlight the importance of studying pragmatic aspects of prompting and raise broader questions about the social dimensions of human-AI interaction.

Embracing Discrete Search: A Reasonable Approach to Causal Structure Learning

arXiv:2510.04970v1 Announce Type: cross Abstract: We present FLOP (Fast Learning of Order and Parents), a score-based causal discovery algorithm for linear models. It pairs fast parent selection with iterative Cholesky-based score updates, cutting run-times over prior algorithms. This makes it feasible to fully embrace discrete search, enabling iterated local search with principled order initialization to find graphs with scores at or close to the global optimum. The resulting structures are highly accurate across benchmarks, with near-perfect recovery in standard settings. This performance calls for revisiting discrete search over graphs as a reasonable approach to causal discovery.

Pivotal CLTs for Pseudolikelihood via Conditional Centering in Dependent Random Fields

arXiv:2510.04972v1 Announce Type: cross Abstract: In this paper, we study fluctuations of conditionally centered statistics of the form $$N^{-1/2}\sum_{i=1}^N c_i(g(\sigma_i)-\mathbb{E}_N[g(\sigma_i)|\sigma_j,j\neq i])$$ where $(\sigma_1,\ldots ,\sigma_N)$ are sampled from a dependent random field, and $g$ is some bounded function. Our first main result shows that under weak smoothness assumptions on the conditional means (which cover both sparse and dense interactions), the above statistic converges to a Gaussian \emph{scale mixture} with a random scale determined by a \emph{quadratic variance} and an \emph{interaction component}. We also show that under appropriate studentization, the limit becomes a pivotal Gaussian. We leverage this theory to develop a general asymptotic framework for maximum pseudolikelihood (MPLE) inference in dependent random fields. We apply our results to Ising models with pairwise as well as higher-order interactions and exponential random graph models (ERGMs). In particular, we obtain a joint central limit theorem for the inverse temperature and magnetization parameters via the joint MPLE (to our knowledge, the first such result in dense, irregular regimes), and we derive conditionally centered edge CLTs and marginal MPLE CLTs for ERGMs without restricting to the ``sub-critical" region. Our proof is based on a method of moments approach via combinatorial decision-tree pruning, which may be of independent interest.

AWARE, Beyond Sentence Boundaries: A Contextual Transformer Framework for Identifying Cultural Capital in STEM Narratives

arXiv:2510.04983v1 Announce Type: cross Abstract: Identifying cultural capital (CC) themes in student reflections can offer valuable insights that help foster equitable learning environments in classrooms. However, themes such as aspirational goals or family support are often woven into narratives, rather than appearing as direct keywords. This makes them difficult to detect for standard NLP models that process sentences in isolation. The core challenge stems from a lack of awareness, as standard models are pre-trained on general corpora, leaving them blind to the domain-specific language and narrative context inherent to the data. To address this, we introduce AWARE, a framework that systematically attempts to improve a transformer model's awareness for this nuanced task. AWARE has three core components: 1) Domain Awareness, adapting the model's vocabulary to the linguistic style of student reflections; 2) Context Awareness, generating sentence embeddings that are aware of the full essay context; and 3) Class Overlap Awareness, employing a multi-label strategy to recognize the coexistence of themes in a single sentence. Our results show that by making the model explicitly aware of the properties of the input, AWARE outperforms a strong baseline by 2.1 percentage points in Macro-F1 and shows considerable improvements across all themes. This work provides a robust and generalizable methodology for any text classification task in which meaning depends on the context of the narrative.

Latent Uncertainty Representations for Video-based Driver Action and Intention Recognition

arXiv:2510.05006v1 Announce Type: cross Abstract: Deep neural networks (DNNs) are increasingly applied to safety-critical tasks in resource-constrained environments, such as video-based driver action and intention recognition. While last layer probabilistic deep learning (LL-PDL) methods can detect out-of-distribution (OOD) instances, their performance varies. As an alternative to last layer approaches, we propose extending pre-trained DNNs with transformation layers to produce multiple latent representations to estimate the uncertainty. We evaluate our latent uncertainty representation (LUR) and repulsively trained LUR (RLUR) approaches against eight PDL methods across four video-based driver action and intention recognition datasets, comparing classification performance, calibration, and uncertainty-based OOD detection. We also contribute 28,000 frame-level action labels and 1,194 video-level intention labels for the NuScenes dataset. Our results show that LUR and RLUR achieve comparable in-distribution classification performance to other LL-PDL approaches. For uncertainty-based OOD detection, LUR matches top-performing PDL methods while being more efficient to train and easier to tune than approaches that require Markov-Chain Monte Carlo sampling or repulsive training procedures.

Curiosity-Driven Co-Development of Action and Language in Robots Through Self-Exploration

arXiv:2510.05013v1 Announce Type: cross Abstract: Human infants acquire language and action co-developmentally, achieving remarkable generalization capabilities from only a minimal number of learning examples. In contrast, recent large language models require exposure to billions of training tokens to achieve such generalization. What mechanisms underlie such efficient developmental learning in humans? This study addresses this question through simulation experiments in which robots learn to perform various actions corresponding to imperative sentences (e.g., \textit{push red cube}) via trials of self-guided exploration. Our approach integrates the active inference framework with reinforcement learning, enabling curiosity-driven developmental learning. The simulations yielded several nontrivial findings: i) Curiosity-driven exploration combined with motor noise substantially outperforms learning without curiosity. ii) Simpler, prerequisite-like actions emerge earlier in development, while more complex actions involving these prerequisites develop later. iii) Rote pairing of sentences and actions occurs before the emergence of compositional generalization. iv) Generalization is drastically improved as the number of compositional elements increases. These results shed light into possible mechanisms underlying efficient co-developmental learning in infants and provide computational parallels to findings in developmental psychology.

Think Then Embed: Generative Context Improves Multimodal Embedding

arXiv:2510.05014v1 Announce Type: cross Abstract: There is a growing interest in Universal Multimodal Embeddings (UME), where models are required to generate task-specific representations. While recent studies show that Multimodal Large Language Models (MLLMs) perform well on such tasks, they treat MLLMs solely as encoders, overlooking their generative capacity. However, such an encoding paradigm becomes less effective as instructions become more complex and require compositional reasoning. Inspired by the proven effectiveness of chain-of-thought reasoning, we propose a general Think-Then-Embed (TTE) framework for UME, composed of a reasoner and an embedder. The reasoner MLLM first generates reasoning traces that explain complex queries, followed by an embedder that produces representations conditioned on both the original query and the intermediate reasoning. This explicit reasoning step enables more nuanced understanding of complex multimodal instructions. Our contributions are threefold. First, by leveraging a powerful MLLM reasoner, we achieve state-of-the-art performance on the MMEB-V2 benchmark, surpassing proprietary models trained on massive in-house datasets. Second, to reduce the dependency on large MLLM reasoners, we finetune a smaller MLLM reasoner using high-quality embedding-centric reasoning traces, achieving the best performance among open-source models with a 7% absolute gain over recently proposed models. Third, we investigate strategies for integrating the reasoner and embedder into a unified model for improved efficiency without sacrificing performance.

Causal Abstractions, Categorically Unified

arXiv:2510.05033v1 Announce Type: cross Abstract: We present a categorical framework for relating causal models that represent the same system at different levels of abstraction. We define a causal abstraction as natural transformations between appropriate Markov functors, which concisely consolidate desirable properties a causal abstraction should exhibit. Our approach unifies and generalizes previously considered causal abstractions, and we obtain categorical proofs and generalizations of existing results on causal abstractions. Using string diagrammatical tools, we can explicitly describe the graphs that serve as consistent abstractions of a low-level graph under interventions. We discuss how methods from mechanistic interpretability, such as circuit analysis and sparse autoencoders, fit within our categorical framework. We also show how applying do-calculus on a high-level graphical abstraction of an acyclic-directed mixed graph (ADMG), when unobserved confounders are present, gives valid results on the low-level graph, thus generalizing an earlier statement by Anand et al. (2023). We argue that our framework is more suitable for modeling causal abstractions compared to existing categorical frameworks. Finally, we discuss how notions such as $\tau$-consistency and constructive $\tau$-abstractions can be recovered with our framework.

A Unified Optimization Framework for Multiclass Classification with Structured Hyperplane Arrangements

arXiv:2510.05047v1 Announce Type: cross Abstract: In this paper, we propose a new mathematical optimization model for multiclass classification based on arrangements of hyperplanes. Our approach preserves the core support vector machine (SVM) paradigm of maximizing class separation while minimizing misclassification errors, and it is computationally more efficient than a previous formulation. We present a kernel-based extension that allows it to construct nonlinear decision boundaries. Furthermore, we show how the framework can naturally incorporate alternative geometric structures, including classification trees, $\ell_p$-SVMs, and models with discrete feature selection. To address large-scale instances, we develop a dynamic clustering matheuristic that leverages the proposed MIP formulation. Extensive computational experiments demonstrate the efficiency of the proposed model and dynamic clustering heuristic, and we report competitive classification performance on both synthetic datasets and real-world benchmarks from the UCI Machine Learning Repository, comparing our method with state-of-the-art implementations available in scikit-learn.

ResMimic: From General Motion Tracking to Humanoid Whole-body Loco-Manipulation via Residual Learning

arXiv:2510.05070v1 Announce Type: cross Abstract: Humanoid whole-body loco-manipulation promises transformative capabilities for daily service and warehouse tasks. While recent advances in general motion tracking (GMT) have enabled humanoids to reproduce diverse human motions, these policies lack the precision and object awareness required for loco-manipulation. To this end, we introduce ResMimic, a two-stage residual learning framework for precise and expressive humanoid control from human motion data. First, a GMT policy, trained on large-scale human-only motion, serves as a task-agnostic base for generating human-like whole-body movements. An efficient but precise residual policy is then learned to refine the GMT outputs to improve locomotion and incorporate object interaction. To further facilitate efficient training, we design (i) a point-cloud-based object tracking reward for smoother optimization, (ii) a contact reward that encourages accurate humanoid body-object interactions, and (iii) a curriculum-based virtual object controller to stabilize early training. We evaluate ResMimic in both simulation and on a real Unitree G1 humanoid. Results show substantial gains in task success, training efficiency, and robustness over strong baselines. Videos are available at https://resmimic.github.io/ .

Diffusion Approximations for Thompson Sampling in the Small Gap Regime

arXiv:2105.09232v5 Announce Type: replace Abstract: We study the process-level dynamics of Thompson sampling in the ``small gap'' regime. The small gap regime is one in which the gaps between the arm means are of order $\sqrt{\gamma}$ or smaller and the time horizon is of order $1/\gamma$, where $\gamma$ is small. As $\gamma \downarrow 0$, we show that the process-level dynamics of Thompson sampling converge weakly to the solutions to certain stochastic differential equations and stochastic ordinary differential equations. Our weak convergence theory is developed from first principles using the Continuous Mapping Theorem, can handle stationary, weakly dependent reward processes, and can also be adapted to analyze a variety of sampling-based bandit algorithms. Indeed, we show that the process-level dynamics of many sampling-based bandit algorithms -- including Thompson sampling designed for any single-parameter exponential family of rewards, as well as non-parametric bandit algorithms based on bootstrap re-sampling -- satisfy an invariance principle. Namely, their weak limits coincide with that of Gaussian parametric Thompson sampling with Gaussian priors. Moreover, in the small gap regime, the regret performance of these algorithms is generally insensitive to model mis-specification, changing continuously with increasing degrees of mis-specification.

Tutorial on amortized optimization

arXiv:2202.00665v5 Announce Type: replace Abstract: Optimization is a ubiquitous modeling tool and is often deployed in settings which repeatedly solve similar instances of the same problem. Amortized optimization methods use learning to predict the solutions to problems in these settings, exploiting the shared structure between similar problem instances. These methods have been crucial in variational inference and reinforcement learning and are capable of solving optimization problems many orders of magnitudes faster than traditional optimization methods that do not use amortization. This tutorial presents an introduction to the amortized optimization foundations behind these advancements and overviews their applications in variational inference, sparse coding, gradient-based meta-learning, control, reinforcement learning, convex optimization, optimal transport, and deep equilibrium networks. The source code for this tutorial is available at https://github.com/facebookresearch/amortized-optimization-tutorial.

On amortizing convex conjugates for optimal transport

arXiv:2210.12153v3 Announce Type: replace Abstract: This paper focuses on computing the convex conjugate (also known as the Legendre-Fenchel conjugate or c-transform) that appears in Euclidean Wasserstein-2 optimal transport. This conjugation is considered difficult to compute and in practice, methods are limited by not being able to exactly conjugate the dual potentials in continuous space. To overcome this, the computation of the conjugate can be approximated with amortized optimization, which learns a model to predict the conjugate. I show that combining amortized approximations to the conjugate with a solver for fine-tuning significantly improves the quality of transport maps learned for the Wasserstein-2 benchmark by Korotin et al. (2021a) and is able to model many 2-dimensional couplings and flows considered in the literature. All baselines, methods, and solvers are publicly available at http://github.com/facebookresearch/w2ot.

Sum-of-Parts: Self-Attributing Neural Networks with End-to-End Learning of Feature Groups

arXiv:2310.16316v5 Announce Type: replace Abstract: Self-attributing neural networks (SANNs) present a potential path towards interpretable models for high-dimensional problems, but often face significant trade-offs in performance. In this work, we formally prove a lower bound on errors of per-feature SANNs, whereas group-based SANNs can achieve zero error and thus high performance. Motivated by these insights, we propose Sum-of-Parts (SOP), a framework that transforms any differentiable model into a group-based SANN, where feature groups are learned end-to-end without group supervision. SOP achieves state-of-the-art performance for SANNs on vision and language tasks, and we validate that the groups are interpretable on a range of quantitative and semantic metrics. We further validate the utility of SOP explanations in model debugging and cosmological scientific discovery. Our code is available at https://github.com/BrachioLab/sop

Generalization of LiNGAM that allows confounding

arXiv:2401.16661v4 Announce Type: replace Abstract: LiNGAM determines the variable order from cause to effect using additive noise models, but it faces challenges with confounding. Previous methods maintained LiNGAM's fundamental structure while trying to identify and address variables affected by confounding. As a result, these methods required significant computational resources regardless of the presence of confounding, and they did not ensure the detection of all confounding types. In contrast, this paper enhances LiNGAM by introducing LiNGAM-MMI, a method that quantifies the magnitude of confounding using KL divergence and arranges the variables to minimize its impact. This method efficiently achieves a globally optimal variable order through the shortest path problem formulation. LiNGAM-MMI processes data as efficiently as traditional LiNGAM in scenarios without confounding while effectively addressing confounding situations. Our experimental results suggest that LiNGAM-MMI more accurately determines the correct variable order, both in the presence and absence of confounding. The code is in the supplementary file in this link: https://github.com/SkyJoyTianle/ISIT2024.

Position Paper: Assessing Robustness, Privacy, and Fairness in Federated Learning Integrated with Foundation Models

arXiv:2402.01857v2 Announce Type: replace Abstract: Federated Learning (FL), while a breakthrough in decentralized machine learning, contends with significant challenges such as limited data availability and the variability of computational resources, which can stifle the performance and scalability of the models. The integration of Foundation Models (FMs) into FL presents a compelling solution to these issues, with the potential to enhance data richness and reduce computational demands through pre-training and data augmentation. However, this incorporation introduces novel issues in terms of robustness, privacy, and fairness, which have not been sufficiently addressed in the existing research. We make a preliminary investigation into this field by systematically evaluating the implications of FM-FL integration across these dimensions. We analyze the trade-offs involved, uncover the threats and issues introduced by this integration, and propose a set of criteria and strategies for navigating these challenges. Furthermore, we identify potential research directions for advancing this field, laying a foundation for future development in creating reliable, secure, and equitable FL systems.

Tabular Data: Is Deep Learning all you need?

arXiv:2402.03970v3 Announce Type: replace Abstract: Tabular data represent one of the most prevalent data formats in applied machine learning, largely because they accommodate a broad spectrum of real-world problems. Existing literature has studied many of the shortcomings of neural architectures on tabular data and has repeatedly confirmed the scalability and robustness of gradient-boosted decision trees across varied datasets. However, recent deep learning models have not been subjected to a comprehensive evaluation under conditions that allow for a fair comparison with existing classical approaches. This situation motivates an investigation into whether recent deep-learning paradigms outperform classical ML methods on tabular data. Our survey fills this gap by benchmarking seventeen state-of-the-art methods, spanning neural networks, classical ML and AutoML techniques. Our empirical results over 68 diverse datasets from a well-established benchmark indicate a paradigm shift, where Deep Learning methods outperform classical approaches.

PolyNet: Learning Diverse Solution Strategies for Neural Combinatorial Optimization

arXiv:2402.14048v2 Announce Type: replace Abstract: Reinforcement learning-based methods for constructing solutions to combinatorial optimization problems are rapidly approaching the performance of human-designed algorithms. To further narrow the gap, learning-based approaches must efficiently explore the solution space during the search process. Recent approaches artificially increase exploration by enforcing diverse solution generation through handcrafted rules, however, these rules can impair solution quality and are difficult to design for more complex problems. In this paper, we introduce PolyNet, an approach for improving exploration of the solution space by learning complementary solution strategies. In contrast to other works, PolyNet uses only a single-decoder and a training schema that does not enforce diverse solution generation through handcrafted rules. We evaluate PolyNet on four combinatorial optimization problems and observe that the implicit diversity mechanism allows PolyNet to find better solutions than approaches that explicitly enforce diverse solution generation.

What is Intelligence? A Cycle Closure Perspective

arXiv:2404.05484v3 Announce Type: replace Abstract: What is intelligence? We argue for a structural-dynamical account rooted in a topological closure law: \emph{the boundary of a boundary vanishes} ($\partial^2=0$). This principle forces transient fragments to cancel while closed cycles persist as invariants, yielding the cascade $\partial^2!=!0 \Rightarrow \text{cycles (invariants)} \Rightarrow \text{memory} \Rightarrow \text{prediction (intelligence)}$. Prediction requires invariance: only order-invariant cycles can stabilize the predictive substrate. This motivates the \textbf{Structure-before-Specificity (SbS)} principle, where persistent structures ($\Phi$) must stabilize before contextual specificities ($\Psi$) can be meaningfully interpreted, and is formalized by the \textbf{Context-Content Uncertainty Principle (CCUP)}, which casts cognition as dynamic alignment that minimizes the joint uncertainty $H(\Phi,\Psi)$. We show that \textbf{Memory-Amortized Inference (MAI)} is the computational mechanism that implements SbS\,$\rightarrow$\,CCUP through dual bootstrapping: \emph{temporal} bootstrapping consolidates episodic specifics into reusable latent trajectories, while \emph{spatial} bootstrapping reuses these invariants across latent manifolds. This framework explains why \emph{semantics precedes syntax}: stable cycles anchor meaning, and symbolic syntax emerges only after semantic invariants are in place. In an evolutionary perspective, the same closure law unifies the trajectory of natural intelligence: from primitive memory traces in microbes, to cyclic sensorimotor patterns in bilaterians, to semantic generalization in mammals, culminating in human symbolic abstraction by natural language. In sum, intelligence arises from the progressive collapse of specificity into structure, grounded in the closure-induced emergence of invariants.

Unified ODE Analysis of Smooth Q-Learning Algorithms

arXiv:2404.14442v5 Announce Type: replace Abstract: Convergence of Q-learning has been the focus of extensive research over the past several decades. Recently, an asymptotic convergence analysis for Q-learning was introduced using a switching system framework. This approach applies the so-called ordinary differential equation (ODE) approach to prove the convergence of the asynchronous Q-learning modeled as a continuous-time switching system, where notions from switching system theory are used to prove its asymptotic stability without using explicit Lyapunov arguments. However, to prove stability, restrictive conditions, such as quasi-monotonicity, must be satisfied for the underlying switching systems, which makes it hard to easily generalize the analysis method to other reinforcement learning algorithms, such as the smooth Q-learning variants. In this paper, we present a more general and unified convergence analysis that improves upon the switching system approach and can analyze Q-learning and its smooth variants. The proposed analysis is motivated by previous work on the convergence of synchronous Q-learning based on $p$-norm serving as a Lyapunov function. However, the proposed analysis addresses more general ODE models that can cover both asynchronous Q-learning and its smooth versions with simpler frameworks.

Characteristic Learning for Provable One Step Generation

arXiv:2405.05512v5 Announce Type: replace Abstract: We propose the characteristic generator, a novel one-step generative model that combines the efficiency of sampling in Generative Adversarial Networks (GANs) with the stable performance of flow-based models. Our model is driven by characteristics, along which the probability density transport can be described by ordinary differential equations (ODEs). Specifically, we first estimate the underlying velocity field and use the Euler method to solve the probability flow ODE, generating discrete approximations of the characteristics. A deep neural network is then trained to fit these characteristics, creating a one-step map that pushes a simple Gaussian distribution to the target distribution. In the theoretical aspect, we provide a comprehensive analysis of the errors arising from velocity matching, Euler discretization, and characteristic fitting to establish a non-asymptotic convergence rate in the 2-Wasserstein distance under mild data assumptions. Crucially, we demonstrate that under a standard manifold assumption, this convergence rate depends only on the intrinsic dimension of data rather than the much larger ambient dimension, proving our model's ability to mitigate the curse of dimensionality. To our knowledge, this is the first rigorous convergence analysis for a flow-based one-step generative model. Experiments on both synthetic and real-world datasets demonstrate that the characteristic generator achieves high-quality and high-resolution sample generation with the efficiency of just a single neural network evaluation.

Asynchronous Federated Stochastic Optimization for Heterogeneous Objectives Under Arbitrary Delays

arXiv:2405.10123v3 Announce Type: replace Abstract: Federated learning (FL) was recently proposed to securely train models with data held over multiple locations (``clients'') under the coordination of a central server. Prolonged training times caused by slow clients may hinder the performance of FL; while asynchronous communication is a promising solution, highly heterogeneous client response times under non-IID local data may introduce significant bias to the global model, particularly in client-driven setups where sampling is infeasible. To address this issue, we propose \underline{A}synch\underline{R}onous \underline{E}xact \underline{A}veraging (\textsc{AREA}), a stochastic (sub)gradient method that leverages asynchrony for scalability and uses client-side memory to correct the bias induced by uneven participation, without client sampling or prior knowledge of client latencies. \textsc{AREA} communicates model residuals rather than gradient estimates, reducing exposure to gradient inversion, and is compatible with secure aggregation. Under standard assumptions and unbounded, heterogeneous delays with finite mean, AREA achieves optimal convergence rates: $\mathcal{O}(1/K)$ in the strongly convex, smooth regime and $\mathcal{O}(1/\sqrt{K})$ in the convex, nonsmooth regime. For strongly convex, smooth objectives, we demonstrate theoretically and empirically that AREA accommodates larger step sizes than existing methods, enabling fast convergence without adversely impacting model generalization. In the convex, nonsmooth setting, to our knowledge we are the first to obtain rates that scale with the average client update frequency rather than the minimum or maximum, indicating increased robustness to outliers.

Federated Continual Learning Goes Online: Uncertainty-Aware Memory Management for Vision Tasks and Beyond

arXiv:2405.18925v4 Announce Type: replace Abstract: Given the ability to model more realistic and dynamic problems, Federated Continual Learning (FCL) has been increasingly investigated recently. A well-known problem encountered in this setting is the so-called catastrophic forgetting, for which the learning model is inclined to focus on more recent tasks while forgetting the previously learned knowledge. The majority of the current approaches in FCL propose generative-based solutions to solve said problem. However, this setting requires multiple training epochs over the data, implying an offline setting where datasets are stored locally and remain unchanged over time. Furthermore, the proposed solutions are tailored for vision tasks solely. To overcome these limitations, we propose a new approach to deal with different modalities in the online scenario where new data arrive in streams of mini-batches that can only be processed once. To solve catastrophic forgetting, we propose an uncertainty-aware memory-based approach. Specifically, we suggest using an estimator based on the Bregman Information (BI) to compute the model's variance at the sample level. Through measures of predictive uncertainty, we retrieve samples with specific characteristics, and - by retraining the model on such samples - we demonstrate the potential of this approach to reduce the forgetting effect in realistic settings while maintaining data confidentiality and competitive communication efficiency compared to state-of-the-art approaches.

Deep Learning without Weight Symmetry

arXiv:2405.20594v2 Announce Type: replace Abstract: Backpropagation, a foundational algorithm for training artificial neural networks, predominates in contemporary deep learning. Although highly successful, it is widely considered biologically implausible, because it relies on precise symmetry between feedforward and feedback weights to accurately propagate gradient signals that assign credit. The so-called weight transport problem concerns how biological brains learn to align feedforward and feedback paths while avoiding the non-biological transport of feedforward weights into feedback weights. To address this, several credit assignment algorithms, such as feedback alignment and the Kollen-Pollack rule, have been proposed. While they can achieve the desired weight alignment, these algorithms imply that if a neuron sends a feedforward synapse to another neuron, it should also receive an identical or at least partially correlated feedback synapse from the latter neuron, thereby forming a bidirectional connection. However, this idealized connectivity pattern contradicts experimental observations in the brain, a discrepancy we refer to as the weight symmetry problem. To address this challenge posed by considering biological constraints on connectivity, we introduce the Product Feedback Alignment (PFA) algorithm. We demonstrate that PFA can eliminate explicit weight symmetry entirely while closely approximating backpropagation and achieving comparable performance in deep convolutional networks. Our results offer a novel approach to solve the longstanding problem of credit assignment in the brain, leading to more biologically plausible learning in deep networks compared to previous methods.

Demystifying Higher-Order Graph Neural Networks

arXiv:2406.12841v3 Announce Type: replace Abstract: Higher-order graph neural networks (HOGNNs) and the related architectures from Topological Deep Learning are an important class of GNN models that harness polyadic relations between vertices beyond plain edges. They have been used to eliminate issues such as over-smoothing or over-squashing, to significantly enhance the accuracy of GNN predictions, to improve the expressiveness of GNN architectures, and for numerous other goals. A plethora of HOGNN models have been introduced, and they come with diverse neural architectures, and even with different notions of what the "higher-order" means. This richness makes it very challenging to appropriately analyze and compare HOGNN models, and to decide in what scenario to use specific ones. To alleviate this, we first design an in-depth taxonomy and a blueprint for HOGNNs. This facilitates designing models that maximize performance. Then, we use our taxonomy to analyze and compare the available HOGNN models. The outcomes of our analysis are synthesized in a set of insights that help to select the most beneficial GNN model in a given scenario, and a comprehensive list of challenges and opportunities for further research into more powerful HOGNNs.

Optimal Bound for PCA with Outliers using Higher-Degree Voronoi Diagrams

arXiv:2408.06867v3 Announce Type: replace Abstract: In this paper, we introduce new algorithms for Principal Component Analysis (PCA) with outliers. Utilizing techniques from computational geometry, specifically higher-degree Voronoi diagrams, we navigate to the optimal subspace for PCA even in the presence of outliers. This approach achieves an optimal solution with a time complexity of $n^{d+\mathcal{O}(1)}\text{poly}(n,d)$. Additionally, we present a randomized algorithm with a complexity of $2^{\mathcal{O}(r(d-r))} \times \text{poly}(n, d)$. This algorithm samples subspaces characterized in terms of a Grassmannian manifold. By employing such sampling method, we ensure a high likelihood of capturing the optimal subspace, with the success probability $(1 - \delta)^T$. Where $\delta$ represents the probability that a sampled subspace does not contain the optimal solution, and $T$ is the number of subspaces sampled, proportional to $2^{r(d-r)}$. Our use of higher-degree Voronoi diagrams and Grassmannian based sampling offers a clearer conceptual pathway and practical advantages, particularly in handling large datasets or higher-dimensional settings.

Temporal Source Recovery for Time-Series Source-Free Unsupervised Domain Adaptation

arXiv:2409.19635v2 Announce Type: replace Abstract: Time-Series (TS) data has grown in importance with the rise of Internet of Things devices like sensors, but its labeling remains costly and complex. While Unsupervised Domain Adaptation (UDAs) offers an effective solution, growing data privacy concerns have led to the development of Source-Free UDA (SFUDAs), enabling model adaptation to target domains without accessing source data. Despite their potential, applying existing SFUDAs to TS data is challenging due to the difficulty of transferring temporal dependencies, an essential characteristic of TS data, particularly in the absence of source samples. Although prior works attempt to address this by specific source pretraining designs, such requirements are often impractical, as source data owners cannot be expected to adhere to particular pretraining schemes. To address this, we propose Temporal Source Recovery (TemSR), a framework that leverages the intrinsic properties of TS data to generate a source-like domain and recover source temporal dependencies. With this domain, TemSR enables dependency transfer to the target domain without accessing source data or relying on source-specific designs, thereby facilitating effective and practical TS-SFUDA. TemSR features a masking recovery optimization process to generate a source-like distribution with restored temporal dependencies. This distribution is further refined through local context-aware regularization to preserve local dependencies, and anchor-based recovery diversity maximization to promote distributional diversity. Together, these components enable effective temporal dependency recovery and facilitate transfer across domains using standard UDA techniques. Extensive experiments across multiple TS tasks demonstrate the effectiveness of TemSR, which even surpasses existing TS-SFUDA methods that require source-specific designs.

Learning-Augmented Robust Algorithmic Recourse

arXiv:2410.01580v2 Announce Type: replace Abstract: Algorithmic recourse provides individuals who receive undesirable outcomes from machine learning systems with minimum-cost improvements to achieve a desirable outcome. However, machine learning models often get updated, so the recourse may not lead to the desired outcome. The robust recourse framework chooses recourses that are less sensitive to adversarial model changes, but this comes at a higher cost. To address this, we initiate the study of learning-augmented algorithmic recourse and evaluate the extent to which a designer equipped with a prediction of the future model can reduce the cost of recourse when the prediction is accurate (consistency) while also limiting the cost even when the prediction is inaccurate (robustness). We propose a novel algorithm, study the robustness-consistency trade-off, and analyze how prediction accuracy affects performance.

Low-Dimension-to-High-Dimension Generalization And Its Implications for Length Generalization

arXiv:2410.08898v3 Announce Type: replace Abstract: Low-Dimension-to-High-Dimension (LDHD) generalization is a special case of Out-of-Distribution (OOD) generalization, where the training data are restricted to a low-dimensional subspace of the high-dimensional testing space. Assuming that each instance is generated from a latent variable and the dimension of the latent variable reflects the problem scale, the inherent scaling challenge in length generalization can be captured by the LDHD generalization in the latent space. We theoretically demonstrate that LDHD generalization is generally unattainable without exploiting prior knowledge to provide appropriate inductive bias. Specifically, we explore LDHD generalization in Boolean functions. We verify that different architectures trained with (S)GD converge to \emph{min-degree interpolators w.r.t. different independent sets}. LDHD generalization is achievable if and only if the target function coincides with this inductive bias. Applying the insights from LDHD generalization to length generalization, we explain the effectiveness of CoT as changing the structure latent space to enable better LDHD generalization. We also propose a principle for position embedding design to handle both the inherent LDHD generalization and the nuisances such as the data format. Following the principle, we propose a novel position embedding called RPE-Square that remedies the RPE for dealing with the data format nuisance.

Learning Orthogonal Multi-Index Models: A Fine-Grained Information Exponent Analysis

arXiv:2410.09678v2 Announce Type: replace Abstract: The information exponent ([BAGJ21]) and its extensions -- which are equivalent to the lowest degree in the Hermite expansion of the link function (after a potential label transform) for Gaussian single-index models -- have played an important role in predicting the sample complexity of online stochastic gradient descent (SGD) in various learning tasks. In this work, we demonstrate that, for multi-index models, focusing solely on the lowest degree can miss key structural details of the model and result in suboptimal rates. Specifically, we consider the task of learning target functions of form $f_(\mathbf{x}) = \sum_{k=1}^{P} \phi(\mathbf{v}_k^ \cdot \mathbf{x})$, where $P \ll d$, the ground-truth directions ${ \mathbf{v}k^* }{k=1}^P$ are orthonormal, and the information exponent of $\phi$ is $L$. Based on the theory of information exponent, when $L = 2$, only the relevant subspace (not the exact directions) can be recovered due to the rotational invariance of the second-order terms, and when $L > 2$, recovering the directions using online SGD require $\tilde{O}(P d^{L-1})$ samples. In this work, we show that by considering both second- and higher-order terms, we can first learn the relevant space using the second-order terms, and then the exact directions using the higher-order terms, and the overall sample and complexity of online SGD is $\tilde{O}( d P^{L-1} )$.

Comparative Performance of Collaborative Bandit Algorithms: Effect of Sparsity and Exploration Intensity

arXiv:2410.12086v2 Announce Type: replace Abstract: This paper offers a comprehensive analysis of collaborative bandit algorithms and provides a thorough comparison of their performance. Collaborative bandits aim to improve the performance of contextual bandits by introducing relationships between arms (or items), allowing effective propagation of information. Collaboration among arms allows the feedback obtained through a single user (item) to be shared across related users (items). Introducing collaboration also alleviates the cold user (item) problem, i.e., lack of historical information when a new user (item) arriving to the platform with no prior record of interactions. In the context of modeling the relationships between arms (items), there are two main approaches: Hard and soft clustering. We call approaches that model the relationship between arms in an \textit{absolute} manner as hard clustering, i.e., the relationship is binary. Soft clustering relaxes membership constraints, allowing \textit{fuzzy} assignment. Focusing on the latter, we provide extensive experiments on the state-of-the-art collaborative contextual bandit algorithms and investigate the effect of sparsity and how the exploration intensity acts as a correction mechanism. Our numerical experiments demonstrate that controlling for sparsity in collaboration improves data efficiency and performance as it better informs learning. Meanwhile, increasing the exploration intensity acts as a correction because it effectively reduces variance due to potentially misspecified relationships among users. We observe that this misspecification is further remedied by introducing latent factors, and thus, increasing the dimensionality of the bandit parameters.

The Persistence of Neural Collapse Despite Low-Rank Bias

arXiv:2410.23169v2 Announce Type: replace Abstract: Neural collapse (NC) and its multi-layer variant, deep neural collapse (DNC), describe a structured geometry that occurs in the features and weights of trained deep networks. Recent theoretical work by Sukenik et al. using a deep unconstrained feature model (UFM) suggests that DNC is suboptimal under mean squared error (MSE) loss. They heuristically argue that this is due to low-rank bias induced by L2 regularization. In this work, we extend this result to deep UFMs trained with cross-entropy loss, showing that high-rank structures, including DNC, are not generally optimal. We characterize the associated low-rank bias, proving a fixed bound on the number of non-negligible singular values at global minima as network depth increases. We further analyze the loss surface, demonstrating that DNC is more prevalent in the landscape than other critical configurations, which we argue explains its frequent empirical appearance. Our results are validated through experiments in deep UFMs and deep neural networks.

Joint Diffusion models in Continual Learning

arXiv:2411.08224v3 Announce Type: replace Abstract: In this work, we introduce JDCL - a new method for continual learning with generative rehearsal based on joint diffusion models. Neural networks suffer from catastrophic forgetting defined as abrupt loss in the model's performance when retrained with additional data coming from a different distribution. Generative-replay-based continual learning methods try to mitigate this issue by retraining a model with a combination of new and rehearsal data sampled from a generative model. In this work, we propose to extend this idea by combining a continually trained classifier with a diffusion-based generative model into a single - jointly optimized neural network. We show that such shared parametrization, combined with the knowledge distillation technique allows for stable adaptation to new tasks without catastrophic forgetting. We evaluate our approach on several benchmarks, where it outperforms recent state-of-the-art generative replay techniques. Additionally, we extend our method to the semi-supervised continual learning setup, where it outperforms competing buffer-based replay techniques, and evaluate, in a self-supervised manner, the quality of trained representations.

MALT: Improving Reasoning with Multi-Agent LLM Training

arXiv:2412.01928v3 Announce Type: replace Abstract: Large Language Models (LLMs) often produce answers with a single chain-of-thought, which restricts their ability to explore reasoning paths or self-correct flawed outputs in complex tasks. In this paper, we introduce MALT (Multi-Agent LLM Training), a novel post-training strategy that divides the reasoning process into generation, verification, and refinement steps using a sequential pipeline of heterogeneous agents. During data generation, each agent is repeatedly sampled to form a multi-agent search tree, where final outputs are graded against ground-truth data. We then apply value iteration to propagate reward signals back to each role-conditioned model, automatically producing multi-agent post-training data without human or teacher-model supervision. Our off-policy approach allows each agent to specialize by learning from correct and incorrect trajectories, ultimately improving the end-to-end reasoning chain. On MATH, GSM8K, and CSQA, MALT surpasses the same baseline LLM with a relative improvement of 15.66%, 7.42%, and 9.40% respectively, making it an important advance towards multi-agent cooperative training.

Detecting LLM Hallucination Through Layer-wise Information Deficiency: Analysis of Ambiguous Prompts and Unanswerable Questions

arXiv:2412.10246v2 Announce Type: replace Abstract: Large language models (LLMs) frequently generate confident yet inaccurate responses, introducing significant risks for deployment in safety-critical domains. We present a novel, test-time approach to detecting model hallucination through systematic analysis of information flow across model layers. We target cases when LLMs process inputs with ambiguous or insufficient context. Our investigation reveals that hallucination manifests as usable information deficiencies in inter-layer transmissions. While existing approaches primarily focus on final-layer output analysis, we demonstrate that tracking cross-layer information dynamics ($\mathcal{L}$I) provides robust indicators of model reliability, accounting for both information gain and loss during computation. $\mathcal{L}$I integrates easily with pretrained LLMs without requiring additional training or architectural modifications.

SyMerge: From Non-Interference to Synergistic Merging via Single-Layer Adaptation

arXiv:2412.19098v3 Announce Type: replace Abstract: Model merging offers an efficient alternative to multi-task learning by combining independently fine-tuned models, but most prior approaches focus mainly on avoiding task interference. We argue instead that the real potential of merging lies in achieving synergy, where tasks enhance one another. Our intuition comes from a pilot study showing that when a classifier trained on one task is paired with the encoder of another, the resulting cross-task performance strongly predicts merge quality. Moreover, adapting even a single task-specific layer can substantially improve this compatibility, suggesting a simple yet powerful lever for synergy. Building on this insight, we introduce SyMerge, a lightweight framework that jointly optimizes one task-specific layer and merging coefficients. To ensure stability without labels, SyMerge employs a robust self-labeling strategy guided by expert model predictions, avoiding the pitfalls of entropy-based adaptation. This minimalist yet principled design achieves state-of-the-art results across vision, dense prediction, and NLP benchmarks, while also producing adapted layers that transfer effectively to other merging methods. Our code is available at https://aim-skku.github.io/SyMerge/

Chameleon2++: An Efficient and Scalable Variant Of Chameleon Clustering

arXiv:2501.02612v2 Announce Type: replace Abstract: Hierarchical clustering remains a fundamental challenge in data mining, particularly when dealing with large-scale datasets where traditional approaches fail to scale effectively. Recent Chameleon-based algorithms - Chameleon2, M-Chameleon, and INNGS-Chameleon have proposed advanced strategies but they still suffer from $O(n^2)$ computational complexity, especially for large datasets. With Chameleon2 as the base algorithm, we introduce Chameleon2++ that addresses this challenge. Our algorithm has three parts. First, Graph Generation - we propose an approximate $k$-NN search instead of an exact one, specifically we integrate with the Annoy algorithm. This results in fast approximate nearest neighbor computation, significantly reducing the graph generation time. Second, Graph Partitioning - we propose use of a multi-level partitioning algorithm instead of a recursive bisection one. Specifically we adapt the hMETIS algorithm instead of the FM. This is because multi-level algorithms are robust to approximation introduced in the graph generation phase yielding higher-quality partitions, and that too with minimum configuration requirements. Third, Merging - we retain the flood fill heuristic that ensures connected balanced components in the partitions as well as efficient partition merging criteria leading to the final clusters. These enhancements reduce the overall time complexity to $O(n\log n)$, achieving scalability. On real-world benchmark datasets used in prior Chameleon works, Chameleon2++ delivers an average of 4% improvement in clustering quality. This demonstrates that algorithmic efficiency and clustering quality can co-exist in large-scale hierarchical clustering.

Cooperative Decentralized Backdoor Attacks on Vertical Federated Learning

arXiv:2501.09320v2 Announce Type: replace Abstract: Federated learning (FL) is vulnerable to backdoor attacks, where adversaries alter model behavior on target classification labels by embedding triggers into data samples. While these attacks have received considerable attention in horizontal FL, they are less understood for vertical FL (VFL), where devices hold different features of the samples, and only the server holds the labels. In this work, we propose a novel backdoor attack on VFL which (i) does not rely on gradient information from the server and (ii) considers potential collusion among multiple adversaries for sample selection and trigger embedding. Our label inference model augments variational autoencoders with metric learning, which adversaries can train locally. A consensus process over the adversary graph topology determines which datapoints to poison. We further propose methods for trigger splitting across the adversaries, with an intensity-based implantation scheme skewing the server towards the trigger. Our convergence analysis reveals the impact of backdoor perturbations on VFL indicated by a stationarity gap for the trained model, which we verify empirically as well. We conduct experiments comparing our attack with recent backdoor VFL approaches, finding that ours obtains significantly higher success rates for the same main task performance despite not using server information. Additionally, our results verify the impact of collusion on attack performance.

Learning to Bid in Non-Stationary Repeated First-Price Auctions

arXiv:2501.13358v3 Announce Type: replace Abstract: First-price auctions have recently gained significant traction in digital advertising markets, exemplified by Google's transition from second-price to first-price auctions. Unlike in second-price auctions, where bidding one's private valuation is a dominant strategy, determining an optimal bidding strategy in first-price auctions is more complex. From a learning perspective, the learner (a specific bidder) can interact with the environment (other bidders, i.e., opponents) sequentially to infer their behaviors. Existing research often assumes specific environmental conditions and benchmarks performance against the best fixed policy (static benchmark). While this approach ensures strong learning guarantees, the static benchmark can deviate significantly from the optimal strategy in environments with even mild non-stationarity. To address such scenarios, a dynamic benchmark--representing the sum of the highest achievable rewards at each time step--offers a more suitable objective. However, achieving no-regret learning with respect to the dynamic benchmark requires additional constraints. By inspecting reward functions in online first-price auctions, we introduce two metrics to quantify the regularity of the sequence of opponents' highest bids, which serve as measures of non-stationarity. We provide a minimax-optimal characterization of the dynamic regret for the class of sequences of opponents' highest bids that satisfy either of these regularity constraints. Our main technical tool is the Optimistic Mirror Descent (OMD) framework with a novel optimism configuration, which is well-suited for achieving minimax-optimal dynamic regret rates in this context. We then use synthetic datasets to validate our theoretical guarantees and demonstrate that our methods outperform existing ones.

AtmosSci-Bench: Evaluating the Recent Advance of Large Language Model for Atmospheric Science

arXiv:2502.01159v3 Announce Type: replace Abstract: The rapid advancements in large language models (LLMs), particularly in their reasoning capabilities, hold transformative potential for addressing complex challenges and boosting scientific discovery in atmospheric science. However, leveraging LLMs effectively in this domain requires a robust and comprehensive evaluation benchmark. Toward this end, we present AtmosSci-Bench, a novel benchmark designed to systematically assess LLM performance across five core categories of atmospheric science problems: hydrology, atmospheric dynamics, atmospheric physics, geophysics, and physical oceanography. AtmosSci-Bench features a dual-format design comprising both multiple-choice questions (MCQs) and open-ended questions (OEQs), enabling scalable automated evaluation alongside deeper analysis of conceptual understanding. We employ a template-based MCQ generation framework to create diverse, graduate-level problems with symbolic perturbation, while OEQs are used to probe open-ended reasoning. We conduct a comprehensive evaluation of representative LLMs, categorized into four groups: instruction-tuned models, advanced reasoning models, math-augmented models, and domain-specific climate models. Our analysis provides some interesting insights into the reasoning and problem-solving capabilities of LLMs in atmospheric science. We believe AtmosSci-Bench can serve as a critical step toward advancing LLM applications in climate services by offering a standard and rigorous evaluation framework. Our source code is available at https://github.com/Relaxed-System-Lab/AtmosSci-Bench.

Universal Adversarial Perturbation Attacks On Modern Behavior Cloning Policies

arXiv:2502.03698v2 Announce Type: replace Abstract: Learning from Demonstration (LfD) algorithms have shown promising results in robotic manipulation tasks, but their vulnerability to offline universal perturbation attacks remains underexplored. This paper presents a comprehensive study of adversarial attacks on both classic and recently proposed algorithms, including Behavior Cloning (BC), LSTM-GMM, Implicit Behavior Cloning (IBC), Diffusion Policy (DP), and Vector-Quantizied Behavior Transformer (VQ-BET). We study the vulnerability of these methods to universal adversarial perturbations. Our experiments on several simulated robotic manipulation tasks reveal that most of the current methods are highly vulnerable to adversarial perturbations. We also show that these attacks are often transferable across algorithms, architectures, and tasks, raising concerning security vulnerabilities to black-box attacks. To the best of our knowledge, we are the first to present a systematic study of the vulnerabilities of different LfD algorithms to both white-box and black-box attacks. Our findings highlight the vulnerabilities of modern BC algorithms, paving the way for future work in addressing such limitations.

From Restless to Contextual: A Thresholding Bandit Reformulation For Finite-horizon Performance

arXiv:2502.05145v3 Announce Type: replace Abstract: This paper addresses the poor finite-horizon performance of existing online \emph{restless bandit} (RB) algorithms, which stems from the prohibitive sample complexity of learning a full \emph{Markov decision process} (MDP) for each agent. We argue that superior finite-horizon performance requires \emph{rapid convergence} to a \emph{high-quality} policy. Thus motivated, we introduce a reformulation of online RBs as a \emph{budgeted thresholding contextual bandit}, which simplifies the learning problem by encoding long-term state transitions into a scalar reward. We prove the first non-asymptotic optimality of an oracle policy for a simplified finite-horizon setting. We propose a practical learning policy under a heterogeneous-agent, multi-state setting, and show that it achieves a sublinear regret, achieving \emph{faster convergence} than existing methods. This directly translates to higher cumulative reward, as empirically validated by significant gains over state-of-the-art algorithms in large-scale heterogeneous environments. Our work provides a new pathway for achieving practical, sample-efficient learning in finite-horizon RBs.

Taming Latency-Memory Trade-Off in MoE-Based LLM Serving via Fine-Grained Expert Offloading

arXiv:2502.05370v2 Announce Type: replace Abstract: Large Language Models (LLMs) have gained immense success in revolutionizing various applications, including content generation, search and recommendation, and AI-assisted operation. To reduce high training costs, Mixture-of-Experts (MoE) architecture has become a popular backbone for modern LLMs. However, despite the benefits, serving MoE-based LLMs experience severe memory inefficiency due to sparsely activated experts. Recent studies propose to offload inactive experts from GPU memory to CPU memory to improve the serving efficiency of MoE models. However, they either incur high inference latency or high model memory footprints due to coarse-grained designs. To tame the latency-memory trade-off in MoE serving, we present FineMoE, a fine-grained expert offloading system for MoE serving that achieves low inference latency with memory efficiency. We design FineMoE to extract fine-grained expert selection patterns from MoE models and semantic hints from input prompts to efficiently guide expert prefetching, caching, and offloading decisions. FineMoE is prototyped on top of HuggingFace Transformers and deployed on a six-GPU testbed. Experiments with open-source MoE models and real-world workloads show that FineMoE reduces inference latency by 47% and improves expert hit rate by 39% over state-of-the-art solutions.

QuIC: Quantum-Inspired Compound Adapters for Parameter Efficient Fine-Tuning

arXiv:2502.06916v2 Announce Type: replace Abstract: Scaling full finetuning of large foundation models strains GPU memory and training time. Parameter Efficient Fine-Tuning (PEFT) methods address this issue via adapter modules which update only a small subset of model parameters. In this work, we introduce Quantum-Inspired Compound Adapters (QuIC Adapters), a PEFT approach inspired from Hamming-weight preserving quantum circuits that can effectively finetune a model using less than 0.02\% memory footprint of the base model. QuIC adapters preserve pretrained representations by enforcing orthogonality in weight parameters, and have native deployment mechanisms on quantum computers. We test QuIC adapters by finetuning large language models like LLaMA and vision transformers on language, math, reasoning and vision benchmarks. In its first-order configuration, QuIC recovers the performance of existing orthogonal methods, while higher-order configurations enable substantial parameter compression (over 40x smaller than LoRA) for a modest performance trade-off, unlocking applications in highly resource-constrained environments. Through ablation studies, we determine that combining multiple Hamming-weight orders with orthogonality and matrix compounding are essential for performant finetuning. Our findings suggest that QuIC adapters offers a promising direction for efficient finetuning of foundation models in resource-constrained environments.

TANTE: Time-Adaptive Operator Learning via Neural Taylor Expansion

arXiv:2502.08574v3 Announce Type: replace Abstract: Operator learning for time-dependent partial differential equations (PDEs) has seen rapid progress in recent years, enabling efficient approximation of complex spatiotemporal dynamics. However, most existing methods rely on fixed time step sizes during rollout, which limits their ability to adapt to varying temporal complexity and often leads to error accumulation. Here, we propose the Time-Adaptive Transformer with Neural Taylor Expansion (TANTE), a novel operator-learning framework that produces continuous-time predictions with adaptive step sizes. TANTE predicts future states by performing a Taylor expansion at the current state, where neural networks learn both the higher-order temporal derivatives and the local radius of convergence. This allows the model to dynamically adjust its rollout based on the local behavior of the solution, thereby reducing cumulative error and improving computational efficiency. We demonstrate the effectiveness of TANTE across a wide range of PDE benchmarks, achieving superior accuracy and adaptability compared to fixed-step baselines, delivering accuracy gains of 60-80 % and speed-ups of 30-40 % at inference time. The code is publicly available at https://github.com/zwu88/TANTE for transparency and reproducibility.

Fed-SB: A Silver Bullet for Extreme Communication Efficiency and Performance in (Private) Federated LoRA Fine-Tuning

arXiv:2502.15436v2 Announce Type: replace Abstract: Low-Rank Adaptation (LoRA) has become ubiquitous for efficiently fine-tuning foundation models. However, federated fine-tuning using LoRA is challenging due to suboptimal updates arising from traditional federated averaging of individual adapters. Existing solutions either incur prohibitively high communication cost that scales linearly with the number of clients or suffer from performance degradation due to limited expressivity. We introduce Federated Silver Bullet (Fed-SB), a novel approach for federated fine-tuning of LLMs using LoRA-SB, a recently proposed low-rank adaptation method. LoRA-SB optimally aligns the optimization trajectory with the ideal low-rank full fine-tuning projection by learning a small square matrix (R) between adapters B and A, keeping other components fixed. Direct averaging of R guarantees exact updates, substantially reducing communication cost, which remains independent of the number of clients, and enables scalability. Fed-SB achieves state-of-the-art performance across commonsense reasoning, arithmetic reasoning, and language inference tasks while reducing communication costs by up to 230x. In private settings, Fed-SB further improves performance by (1) reducing trainable parameters, thereby lowering the noise required for differential privacy and (2) avoiding noise amplification introduced by other methods. Overall, Fed-SB offers a state-of-the-art, efficient, and scalable solution for both private and non-private federated fine-tuning. Our code is publicly available at: https://github.com/CERT-Lab/fed-sb.

DISC: Dynamic Decomposition Improves LLM Inference Scaling

arXiv:2502.16706v3 Announce Type: replace Abstract: Inference scaling methods for LLMs often rely on decomposing problems into steps (or groups of tokens), followed by sampling and selecting the best next steps. However, these steps and their sizes are often predetermined or manually designed based on domain knowledge. We propose dynamic decomposition, a method that adaptively and automatically partitions solution and reasoning traces into manageable steps during inference. By more effectively allocating compute -- particularly through subdividing challenging steps and prioritizing their sampling -- dynamic decomposition significantly improves inference efficiency. Experiments on benchmarks such as APPS, MATH, and LiveCodeBench demonstrate that dynamic decomposition outperforms static approaches, including token-level, sentence-level, and single-step decompositions, reducing the pass@10 error rate by 5.0%, 6.7%, and 10.5% respectively. These findings highlight the potential of dynamic decomposition to improve a wide range of inference scaling techniques.

Mamba base PKD for efficient knowledge compression

arXiv:2503.01727v2 Announce Type: replace Abstract: Deep neural networks (DNNs) have remarkably succeeded in various image processing tasks. However, their large size and computational complexity present significant challenges for deploying them in resource-constrained environments. This paper presents an innovative approach for integrating Mamba Architecture within a Progressive Knowledge Distillation (PKD) process to address the challenge of reducing model complexity while maintaining accuracy in image classification tasks. The proposed framework distills a large teacher model into progressively smaller student models, designed using Mamba blocks. Each student model is trained using Selective-State-Space Models (S-SSM) within the Mamba blocks, focusing on important input aspects while reducing computational complexity. The work's preliminary experiments use MNIST and CIFAR-10 as datasets to demonstrate the effectiveness of this approach. For MNIST, the teacher model achieves 98% accuracy. A set of seven student models as a group retained 63% of the teacher's FLOPs, approximating the teacher's performance with 98% accuracy. The weak student used only 1% of the teacher's FLOPs and maintained 72% accuracy. Similarly, for CIFAR-10, the students achieved 1% less accuracy compared to the teacher, with the small student retaining 5% of the teacher's FLOPs to achieve 50% accuracy. These results confirm the flexibility and scalability of Mamba Architecture, which can be integrated into PKD, succeeding in the process of finding students as weak learners. The framework provides a solution for deploying complex neural networks in real-time applications with a reduction in computational cost.

Interpretable Visualizations of Data Spaces for Classification Problems

arXiv:2503.05861v3 Announce Type: replace Abstract: How do classification models "see" our data? Based on their success in delineating behaviors, there must be some lens through which it is easy to see the boundary between classes; however, our current set of visualization techniques makes this prospect difficult. In this work, we propose a hybrid supervised-unsupervised technique distinctly suited to visualizing the decision boundaries determined by classification problems. This method provides a human-interpretable map that can be analyzed qualitatively and quantitatively, which we demonstrate through visualizing and interpreting a decision boundary for chemical neurotoxicity. While we discuss this method in the context of chemistry-driven problems, its application can be generalized across subfields for "unboxing" the operations of machine-learning classification models.

Dependency-aware Maximum Likelihood Estimation for Active Learning

arXiv:2503.05969v2 Announce Type: replace Abstract: Active learning aims to efficiently build a labeled training set by strategically selecting samples to query labels from annotators. In this sequential process, each sample acquisition influences subsequent selections, causing dependencies among samples in the labeled set. However, these dependencies are overlooked during the model parameter estimation stage when updating the model using Maximum Likelihood Estimation (MLE), a conventional method that assumes independent and identically distributed (i.i.d.) data. We propose Dependency-aware MLE (DMLE), which corrects MLE within the active learning framework by addressing sample dependencies typically neglected due to the i.i.d. assumption, ensuring consistency with active learning principles in the model parameter estimation process. This improved method achieves superior performance across multiple benchmark datasets, reaching higher performance in earlier cycles compared to conventional MLE. Specifically, we observe average accuracy improvements of 6%, 8.6%, and 10.5% for k=1, k=5, and k=10 respectively, after collecting the first 100 samples, where entropy is the acquisition function and k is the query batch size acquired at every active learning cycle. Our implementation is publicly available at: https://github.com/neu-spiral/DMLEforAL

Exact and Linear Convergence for Federated Learning under Arbitrary Client Participation is Attainable

arXiv:2503.20117v3 Announce Type: replace Abstract: This work tackles the fundamental challenges in Federated Learning (FL) posed by arbitrary client participation and data heterogeneity, prevalent characteristics in practical FL settings. It is well-established that popular FedAvg-style algorithms struggle with exact convergence and can suffer from slow convergence rates since a decaying learning rate is required to mitigate these scenarios. To address these issues, we introduce the concept of stochastic matrix and the corresponding time-varying graphs as a novel modeling tool to accurately capture the dynamics of arbitrary client participation and the local update procedure. Leveraging this approach, we offer a fresh decentralized perspective on designing FL algorithms and present FOCUS, Federated Optimization with Exact Convergence via Push-pull Strategy, a provably convergent algorithm designed to effectively overcome the previously mentioned two challenges. More specifically, we provide a rigorous proof demonstrating that FOCUS achieves exact convergence with a linear rate regardless of the arbitrary client participation, establishing it as the first work to demonstrate this significant result.

Understanding R1-Zero-Like Training: A Critical Perspective

arXiv:2503.20783v2 Announce Type: replace Abstract: DeepSeek-R1-Zero has shown that reinforcement learning (RL) at scale can directly enhance the reasoning capabilities of LLMs without supervised fine-tuning. In this work, we critically examine R1-Zero-like training by analyzing its two core components: base models and RL. We investigate a wide range of base models, including DeepSeek-V3-Base, to understand how pretraining characteristics influence RL performance. Our analysis reveals that DeepSeek-V3-Base already exhibit ''Aha moment'', while Qwen2.5 base models demonstrate strong reasoning capabilities even without prompt templates, suggesting potential pretraining biases. Additionally, we identify an optimization bias in Group Relative Policy Optimization (GRPO), which artificially increases response length (especially for incorrect outputs) during training. To address this, we introduce Dr. GRPO, an unbiased optimization method that improves token efficiency while maintaining reasoning performance. Leveraging these insights, we present a minimalist R1-Zero recipe that achieves 43.3% accuracy on AIME 2024 with a 7B base model, establishing a new state-of-the-art. Our code is available at https://github.com/sail-sg/understand-r1-zero.

TRA: Better Length Generalisation with Threshold Relative Attention

arXiv:2503.23174v4 Announce Type: replace Abstract: Transformers struggle with length generalisation, displaying poor performance even on basic tasks. We test whether these limitations can be explained through two key failures of the self-attention mechanism. The first is the inability to fully remove irrelevant information. The second is tied to position, even if the dot product between a key and query is highly negative (i.e. an irrelevant key) learned positional biases may unintentionally up-weight such information - dangerous when distances become out of distribution. Put together, these two failure cases lead to compounding generalisation difficulties. We test whether they can be mitigated through the combination of a) selective sparsity - completely removing irrelevant keys from the attention softmax and b) contextualised relative distance - distance is only considered as between the query and the keys that matter. We show how refactoring the attention mechanism with these two mitigations in place can substantially improve the generalisation capabilities of decoder only transformers.

Large EEG-U-Transformer for Time-Step Level Detection Without Pre-Training

arXiv:2504.00336v3 Announce Type: replace Abstract: Electroencephalography (EEG) reflects the brain's functional state, making it a crucial tool for diverse detection applications like seizure detection and sleep stage classification. While deep learning-based approaches have recently shown promise for automated detection, traditional models are often constrained by limited learnable parameters and only achieve modest performance. In contrast, large foundation models showed improved capabilities by scaling up the model size, but required extensive time-consuming pre-training. Moreover, both types of existing methods require complex and redundant post-processing pipelines to convert discrete labels to continuous annotations. In this work, based on the multi-scale nature of EEG events, we propose a simple U-shaped model to efficiently learn representations by capturing both local and global features using convolution and self-attentive modules for sequence-to-sequence modeling. Compared to other window-level classification models, our method directly outputs predictions at the time-step level, eliminating redundant overlapping inferences. Beyond sequence-to-sequence modeling, the architecture naturally extends to window-level classification by incorporating an attention-pooling layer. Such a paradigm shift and model design demonstrated promising efficiency improvement, cross-subject generalization, and state-of-the-art performance in various time-step and window-level classification tasks in the experiment. More impressively, our model showed the capability to be scaled up to the same level as existing large foundation models that have been extensively pre-trained over diverse datasets and outperforms them by solely using the downstream fine-tuning dataset. Our model won 1st place in the 2025 "seizure detection challenge" organized in the International Conference on Artificial Intelligence in Epilepsy and Other Neurological Disorders.

AgentRewardBench: Evaluating Automatic Evaluations of Web Agent Trajectories

arXiv:2504.08942v2 Announce Type: replace Abstract: Web agents enable users to perform tasks on web browsers through natural language interaction. Evaluating web agents trajectories is an important problem, since it helps us determine whether the agent successfully completed the tasks. Rule-based methods are widely used for this purpose, but they are challenging to extend to new tasks and may not always recognize successful trajectories. We may achieve higher accuracy through human evaluation, but the process would be substantially slower and more expensive. Automatic evaluations with LLMs may avoid the challenges of designing new rules and manually annotating trajectories, enabling faster and cost-effective evaluation. However, it is unclear how effective they are at evaluating web agents. To this end, we propose AgentRewardBench, the first benchmark to assess the effectiveness of LLM judges for evaluating web agents. AgentRewardBench contains 1302 trajectories across 5 benchmarks and 4 LLMs. Each trajectory in AgentRewardBench is reviewed by an expert, who answers questions pertaining to the success, side effects, and repetitiveness of the agent. Using our benchmark, we evaluate 12 LLM judges and find that no single LLM excels across all benchmarks. We also find that the rule-based evaluation used by common benchmarks tends to underreport the success rate of web agents, highlighting a key weakness of rule-based evaluation and the need to develop more flexible automatic evaluations. We release the benchmark at: https://agent-reward-bench.github.io

VIPO: Value Function Inconsistency Penalized Offline Reinforcement Learning

arXiv:2504.11944v2 Announce Type: replace Abstract: Offline reinforcement learning (RL) learns effective policies from pre-collected datasets, offering a practical solution for applications where online interactions are risky or costly. Model-based approaches are particularly advantageous for offline RL, owing to their data efficiency and generalizability. However, due to inherent model errors, model-based methods often artificially introduce conservatism guided by heuristic uncertainty estimation, which can be unreliable. In this paper, we introduce VIPO, a novel model-based offline RL algorithm that incorporates self-supervised feedback from value estimation to enhance model training. Specifically, the model is learned by additionally minimizing the inconsistency between the value learned directly from the offline data and the one estimated from the model. We perform comprehensive evaluations from multiple perspectives to show that VIPO can learn a highly accurate model efficiently and consistently outperform existing methods. In particular, it achieves state-of-the-art performance on almost all tasks in both D4RL and NeoRL benchmarks. Overall, VIPO offers a general framework that can be readily integrated into existing model-based offline RL algorithms to systematically enhance model accuracy.

Recursive Deep Inverse Reinforcement Learning

arXiv:2504.13241v5 Announce Type: replace Abstract: Inferring an adversary's goals from exhibited behavior is crucial for counterplanning and non-cooperative multi-agent systems in domains like cybersecurity, military, and strategy games. Deep Inverse Reinforcement Learning (IRL) methods based on maximum entropy principles show promise in recovering adversaries' goals but are typically offline, require large batch sizes with gradient descent, and rely on first-order updates, limiting their applicability in real-time scenarios. We propose an online Recursive Deep Inverse Reinforcement Learning (RDIRL) approach to recover the cost function governing the adversary actions and goals. Specifically, we minimize an upper bound on the standard Guided Cost Learning (GCL) objective using sequential second-order Newton updates, akin to the Extended Kalman Filter (EKF), leading to a fast (in terms of convergence) learning algorithm. We demonstrate that RDIRL is able to recover cost and reward functions of expert agents in standard and adversarial benchmark tasks. Experiments on benchmark tasks show that our proposed approach outperforms several leading IRL algorithms.

What Lurks Within? Concept Auditing for Shared Diffusion Models at Scale

arXiv:2504.14815v2 Announce Type: replace Abstract: Diffusion models (DMs) have revolutionized text-to-image generation, enabling the creation of highly realistic and customized images from text prompts. With the rise of parameter-efficient fine-tuning (PEFT) techniques, users can now customize powerful pre-trained models using minimal computational resources. However, the widespread sharing of fine-tuned DMs on open platforms raises growing ethical and legal concerns, as these models may inadvertently or deliberately generate sensitive or unauthorized content. Despite increasing regulatory attention on generative AI, there are currently no practical tools for systematically auditing these models before deployment. In this paper, we address the problem of concept auditing: determining whether a fine-tuned DM has learned to generate a specific target concept. Existing approaches typically rely on prompt-based input crafting and output-based image classification but they suffer from critical limitations, including prompt uncertainty, concept drift, and poor scalability. To overcome these challenges, we introduce Prompt-Agnostic Image-Free Auditing (PAIA), a novel, model-centric concept auditing framework. By treating the DM as the object of inspection, PAIA enables direct analysis of internal model behavior, bypassing the need for optimized prompts or generated images. We evaluate PAIA on 320 controlled models trained with curated concept datasets and 771 real-world community models sourced from a public DM sharing platform. Evaluation results show that PAIA achieves over 90% detection accuracy while reducing auditing time by 18 - 40X compared to existing baselines. To our knowledge, PAIA is the first scalable and practical solution for pre-deployment concept auditing of diffusion models, providing a practical foundation for safer and more transparent diffusion model sharing.

Learning and Transferring Physical Models through Derivatives

arXiv:2505.01391v2 Announce Type: replace Abstract: We propose Derivative Learning (DERL), a supervised approach that models physical systems by learning their partial derivatives. We also leverage DERL to build physical models incrementally, by designing a distillation protocol that effectively transfers knowledge from a pre-trained model to a student one. We provide theoretical guarantees that DERL can learn the true physical system, being consistent with the underlying physical laws, even when using empirical derivatives. DERL outperforms state-of-the-art methods in generalizing an ODE to unseen initial conditions and a parametric PDE to unseen parameters. We also design a method based on DERL to transfer physical knowledge across models by extending them to new portions of the physical domain and a new range of PDE parameters. We believe this is the first attempt at building physical models incrementally in multiple stages.

Taming OOD Actions for Offline Reinforcement Learning: An Advantage-Based Approach

arXiv:2505.05126v4 Announce Type: replace Abstract: Offline reinforcement learning (RL) learns policies from fixed datasets without online interactions, but suffers from distribution shift, causing inaccurate evaluation and overestimation of out-of-distribution (OOD) actions. Existing methods counter this by conservatively discouraging all OOD actions, which limits generalization. We propose Advantage-based Diffusion Actor-Critic (ADAC), which evaluates OOD actions via an advantage-like function and uses it to modulate the Q-function update discriminatively. Our key insight is that the (state) value function is generally learned more reliably than the action-value function; we thus use the next-state value to indirectly assess each action. We develop a PointMaze environment to clearly visualize that advantage modulation effectively selects superior OOD actions while discouraging inferior ones. Moreover, extensive experiments on the D4RL benchmark show that ADAC achieves state-of-the-art performance, with especially strong gains on challenging tasks.

Learning Penalty for Optimal Partitioning via Automatic Feature Extraction

arXiv:2505.07413v2 Announce Type: replace Abstract: Changepoint detection identifies significant shifts in data sequences, making it important in areas like finance, genetics, and healthcare. The Optimal Partitioning algorithms efficiently detect these changes, using a penalty parameter to limit the changepoints count. Determining the optimal value for this penalty can be challenging. Traditionally, this process involved manually extracting statistical features, such as sequence length or variance to make the prediction. This study proposes a novel approach that uses recurrent networks to learn this penalty directly from raw sequences by automatically extracting features. Experiments conducted on 20 benchmark genomic datasets show that this novel method generally outperforms traditional ones in changepoint detection accuracy.

Density Ratio-based Causal Discovery from Bivariate Continuous-Discrete Data

arXiv:2505.08371v3 Announce Type: replace Abstract: We propose a causal discovery method for mixed bivariate data consisting of one continuous and one discrete variable. Existing approaches either impose strong distributional assumptions or face challenges in fairly comparing causal directions between variables of different types, due to differences in their information content. We introduce a novel approach that determines causal direction by analyzing the monotonicity of the conditional density ratio of the continuous variable, conditioned on different values of the discrete variable. Our theoretical analysis shows that the conditional density ratio exhibits monotonicity when the continuous variable causes the discrete variable, but not in the reverse direction. This property provides a principled basis for comparing causal directions between variables of different types, free from strong distributional assumptions and bias arising from differences in their information content. We demonstrate its effectiveness through experiments on both synthetic and real-world datasets, showing superior accuracy compared to existing methods.

A Case for Library-Level k-Means Binning in Histogram Gradient-Boosted Trees

arXiv:2505.12460v2 Announce Type: replace Abstract: Modern Gradient Boosted Decision Trees (GBDTs) accelerate split finding with histogram-based binning, which reduces complexity from $O(N\log N)$ to $O(N)$ by aggregating gradients into fixed-size bins. However, the predominant quantile binning strategy - designed to distribute data points evenly among bins -- may overlook critical boundary values that could enhance predictive performance. In this work, we consider a novel approach that replaces quantile binning with a $k$-means discretizer initialized with quantile bins, and justify the swap with a proof showing how, for any $L$-Lipschitz function, k-means maximizes the worst-case explained variance of Y obtained when treating all values in a given bin as equivalent. We test this swap against quantile and uniform binning on 33 OpenML datasets plus synthetics that control for modality, skew, and bin budget. Across 18 regression datasets, k-means shows no statistically significant losses at the 5% level and wins in three cases-most strikingly a 55% MSE drop on one particularly skewed dataset-even though k-means' mean reciprocal rank (MRR) is slightly lower (0.65 vs 0.72). On the 15 classification datasets the two methods are statistically tied (MRR 0.70 vs 0.68) with gaps $\leq$0.2 pp. Synthetic experiments confirm consistently large MSE gains - typically >20% and rising to 90% as outlier magnitude increases or bin budget drops. We find that k-means keeps error on par with exhaustive (no-binning) splitting when extra cuts add little value, yet still recovers key split points that quantile overlooks. As such, we advocate for a built-in bin_method=k-means flag, especially in regression tasks and in tight-budget settings such as the 32-64-bin GPU regime - because it is a "safe default" with large upside, yet adds only a one-off, cacheable overhead ($\approx$ 3.5s per feature to bin 10M rows on one Apple M1 thread).

Safety Subspaces are Not Linearly Distinct: A Fine-Tuning Case Study

arXiv:2505.14185v2 Announce Type: replace Abstract: Large Language Models (LLMs) rely on safety alignment to produce socially acceptable responses. However, this behavior is known to be brittle: further fine-tuning, even on benign or lightly contaminated data, can degrade safety and reintroduce harmful behaviors. A growing body of work suggests that alignment may correspond to identifiable directions in weight space, forming subspaces that could, in principle, be isolated or preserved to defend against misalignment. In this work, we conduct a comprehensive empirical study of this perspective. We examine whether safety-relevant behavior is concentrated in specific linear subspaces, whether it can be separated from general-purpose learning, and whether harmfulness arises from distinguishable patterns in activations. Across both weight and activation spaces, our findings are consistent: subspaces that amplify safe behaviors also amplify useful ones, and prompts with different safety implications activate overlapping representations. Rather than residing in distinct directions, we show that safety is highly entangled with the general learning components of the model. This suggests that subspace-based defenses face fundamental limitations and underscores the need for alternative strategies to preserve safety under continued training. We corroborate these findings with multiple experiments on five open-source LLMs from the Llama and Qwen families. Our code is publicly available at: https://github.com/CERT-Lab/safety-subspaces.

Directional Convergence, Benign Overfitting of Gradient Descent in leaky ReLU two-layer Neural Networks

arXiv:2505.16204v2 Announce Type: replace Abstract: In this paper, we study benign overfitting of fixed width leaky ReLU two-layer neural network classifiers trained on mixture data via gradient descent. We provide both, upper and lower classification error bounds, and discover a phase transition in the bound as a function of signal strength. The lower bound leads to a characterization of cases when benign overfitting provably fails even if directional convergence occurs. Our analysis allows us to considerably relax the distributional assumptions that are made in existing work on benign overfitting of leaky ReLU two-layer neural network classifiers. We can allow for non-sub-Gaussian data and do not require near orthogonality. Our results are derived by establishing directional convergence of the network parameters and studying classification error bounds for the convergent direction. Previously, directional convergence in (leaky) ReLU neural networks was established only for gradient flow. By first establishing directional convergence, we are able to study benign overfitting of fixed width leaky ReLU two-layer neural network classifiers in a much wider range of scenarios than was done before.

AdaSTaR: Adaptive Data Sampling for Training Self-Taught Reasoners

arXiv:2505.16322v3 Announce Type: replace Abstract: Self-Taught Reasoners (STaR), synonymously known as Rejection sampling Fine-Tuning (RFT), is an integral part of the training pipeline of self-improving reasoning Language Models (LMs). The self-improving mechanism often employs random observation (data) sampling. However, this results in trained observation imbalance; inefficiently over-training on solved examples while under-training on challenging ones. In response, we introduce Adaptive STaR (AdaSTaR), a novel algorithm that rectifies this by integrating two adaptive sampling principles: (1) Adaptive Sampling for Diversity: promoting balanced training across observations, and (2) Adaptive Sampling for Curriculum: dynamically adjusting data difficulty to match the model's evolving strength. Across six benchmarks, AdaSTaR achieves best test accuracy in all instances (6/6) and reduces training FLOPs by an average of 58.6% against an extensive list of baselines. These improvements in performance and efficiency generalize to different pre-trained LMs and larger models, paving the way for more efficient and effective self-improving LMs.

Computing Exact Shapley Values in Polynomial Time for Product-Kernel Methods

arXiv:2505.16516v2 Announce Type: replace Abstract: Kernel methods are widely used in machine learning due to their flexibility and expressiveness. However, their black-box nature poses significant challenges to interpretability, limiting their adoption in high-stakes applications. Shapley value-based feature attribution techniques, such as SHAP and kernel method-specific adaptation like RKHS-SHAP, offer a promising path toward explainability. Yet, computing exact Shapley values is generally intractable, leading existing methods to rely on approximations and thereby incur unavoidable error. In this work, we introduce PKeX-Shapley, a novel algorithm that utilizes the multiplicative structure of product kernels to enable the exact computation of Shapley values in polynomial time. The core of our approach is a new value function, the functional baseline value function, specifically designed for product-kernel models. This value function removes the influence of a feature subset by setting its functional component to the least informative state. Crucially, it allows a recursive thus efficient computation of Shapley values in polynomial time. As an important additional contribution, we show that our framework extends beyond predictive modeling to statistical inference. In particular, it generalizes to popular kernel-based discrepancy measures such as the Maximum Mean Discrepancy (MMD) and the Hilbert-Schmidt Independence Criterion (HSIC), thereby providing new tools for interpretable statistical inference.

Towards Coordinate- and Dimension-Agnostic Machine Learning for Partial Differential Equations

arXiv:2505.16549v2 Announce Type: replace Abstract: The machine learning methods for data-driven identification of partial differential equations (PDEs) are typically defined for a given number of spatial dimensions and a choice of coordinates the data have been collected in. This dependence prevents the learned evolution equation from generalizing to other spaces. In this work, we reformulate the problem in terms of coordinate- and dimension-independent representations, paving the way toward what we call ``spatially liberated" PDE learning. To this end, we employ a machine learning approach to predict the evolution of scalar field systems expressed in the formalism of exterior calculus, which is coordinate-free and immediately generalizes to arbitrary dimensions by construction. We demonstrate the performance of this approach in the FitzHugh-Nagumo and Barkley reaction-diffusion models, as well as the Patlak-Keller-Segel model informed by in-situ chemotactic bacteria observations. We provide extensive numerical experiments that demonstrate that our approach allows for seamless transitions across various spatial contexts. We show that the field dynamics learned in one space can be used to make accurate predictions in other spaces with different dimensions, coordinate systems, boundary conditions, and curvatures.

Advancing Brainwave Modeling with a Codebook-Based Foundation Model

arXiv:2505.16724v2 Announce Type: replace Abstract: Recent advances in large-scale pre-trained Electroencephalogram (EEG) models have shown great promise, driving progress in Brain-Computer Interfaces (BCIs) and healthcare applications. However, despite their success, many existing pre-trained models have struggled to fully capture the rich information content of neural oscillations, a limitation that fundamentally constrains their performance and generalizability across diverse BCI tasks. This limitation is frequently rooted in suboptimal architectural design choices which constrain their representational capacity. In this work, we introduce LaBraM++, an enhanced Large Brainwave Foundation Model (LBM) that incorporates principled improvements grounded in robust signal processing foundations. LaBraM++ demonstrates substantial gains across a variety of tasks, consistently outperforming its originally-based architecture and achieving competitive results when compared to other open-source LBMs. Its superior performance and training efficiency highlight its potential as a strong foundation for future advancements in LBMs.

COUNTDOWN: Contextually Sparse Activation Filtering Out Unnecessary Weights in Down Projection

arXiv:2505.17701v2 Announce Type: replace Abstract: The growing size of large language models has created significant computational inefficiencies. To address this challenge, sparse activation methods selectively deactivates non-essential parameters during inference, reducing computational costs in FFNN layers. While existing methods focus on non-linear gating mechanisms, we hypothesize that the sparsity of the FFNN layer lies globally in the form of a linear combination over its internal down projection matrix. Based on this insight, we propose two methods: M-COUNTDOWN, leveraging indirect coefficients, and D-COUNTDOWN, utilizing direct coefficients of the linear combination. Experimental results demonstrate that D-COUNTDOWN can omit 90% of computations with performance loss as low as 5.5% ideally, while M-COUNTDOWN provides a predictor-free solution with up to 29.4% better performance preservation compared to existing methods. Our specialized kernel implementations effectively realize these theoretical gains into substantial real-world acceleration.

Improved Sample Complexity For Diffusion Model Training Without Empirical Risk Minimizer Access

arXiv:2505.18344v5 Announce Type: replace Abstract: Diffusion models have demonstrated remarkable performance in generating high-dimensional samples across domains such as vision, language, and the sciences. Although continuous-state diffusion models have been extensively studied both empirically and theoretically, discrete-state diffusion models, essential for applications involving text, sequences, and combinatorial structures, they remain significantly less understood from a theoretical standpoint. In particular, all existing analyses of discrete-state models assume access to an empirical risk minimizer. In this work, we present a principled theoretical framework analyzing diffusion models, providing a state-of-the-art sample complexity bound of $\widetilde{\mathcal{O}}(\epsilon^{-4})$. Our structured decomposition of the score estimation error into statistical and optimization components offers critical insights into how diffusion models can be trained efficiently. This analysis addresses a fundamental gap in the literature and establishes the theoretical tractability and practical relevance of diffusion models.

How Many Parameters Does Your Task Really Need? Task Specific Pruning with LLM-Sieve

arXiv:2505.18350v2 Announce Type: replace Abstract: As Large Language Models (LLMs) are increasingly deployed for narrow tasks in resource-constrained settings, a central question arises: how much of an LLM is truly necessary for a given task? We present LLM-Sieve, a framework that prunes LLMs down to the minimal parameter subset needed to preserve task performance. Our approach introduces two innovations: (i) output-aligned non-orthogonal projections, which yield more faithful low-rank approximations than traditional PCA/SVD by aligning directly with layer outputs; and (ii) adaptive pruning via a Genetic Algorithm, which automatically discovers matrix-specific pruning levels and exposes the uneven distribution of task-relevant knowledge. Across models from 3.8B to 70B parameters, LLM-Sieve removes 20-75% of weights with only 1-5% accuracy loss-substantially ahead of prior pruning methods. Beyond efficiency, our framework reveals bottleneck matrices that concentrate critical knowledge, suggesting architectural implications for future LLM design. LLM-Sieve integrates seamlessly with LoRA fine-tuning and quantization, enabling both efficient deployment and deeper understanding of knowledge organization in LLMs.

How to build a consistency model: Learning flow maps via self-distillation

arXiv:2505.18825v2 Announce Type: replace Abstract: Flow-based generative models achieve state-of-the-art sample quality, but require the expensive solution of a differential equation at inference time. Flow map models, commonly known as consistency models, encompass many recent efforts to improve inference-time efficiency by learning the solution operator of this differential equation. Yet despite their promise, these models lack a unified description that clearly explains how to learn them efficiently in practice. Here, building on the methodology proposed in Boffi et. al. (2024), we present a systematic algorithmic framework for directly learning the flow map associated with a flow or diffusion model. By exploiting a relationship between the velocity field underlying a continuous-time flow and the instantaneous rate of change of the flow map, we show how to convert any distillation scheme into a direct training algorithm via self-distillation, eliminating the need for pre-trained teachers. We introduce three algorithmic families based on different mathematical characterizations of the flow map: Eulerian, Lagrangian, and Progressive methods, which we show encompass and extend all known distillation and direct training schemes for consistency models. We find that the novel class of Lagrangian methods, which avoid both spatial derivatives and bootstrapping from small steps by design, achieve significantly more stable training and higher performance than more standard Eulerian and Progressive schemes. Our methodology unifies existing training schemes under a single common framework and reveals new design principles for accelerated generative modeling. Associated code is available at https://github.com/nmboffi/flow-maps.

Behavior Injection: Preparing Language Models for Reinforcement Learning

arXiv:2505.18917v2 Announce Type: replace Abstract: Reinforcement learning (RL) has emerged as a powerful post-training technique to incentivize the reasoning ability of large language models (LLMs). However, LLMs can respond very inconsistently to RL finetuning: some show substantial performance gains, while others plateau or even degrade. To understand this divergence, we analyze the per-step influence of the RL objective and identify two key conditions for effective post-training: (1) RL-informative rollout accuracy, and (2) strong data co-influence, which quantifies how much the training data affects performance on other samples. Guided by these insights, we propose behavior injection, a task-agnostic data augmentation scheme applied prior to RL. Behavior injection enriches the supervised finetuning (SFT) data by seeding exploratory and exploitative behaviors, effectively making the model more RL-ready. We evaluate our method across two reasoning benchmarks with multiple base models. The results demonstrate that our theoretically motivated augmentation can significantly increase the performance gain from RL over the pre-RL model.

A Snapshot of Influence: A Local Data Attribution Framework for Online Reinforcement Learning

arXiv:2505.19281v2 Announce Type: replace Abstract: Online reinforcement learning (RL) excels in complex, safety-critical domains but suffers from sample inefficiency, training instability, and limited interpretability. Data attribution provides a principled way to trace model behavior back to training samples, yet existing methods assume fixed datasets, which is violated in online RL where each experience both updates the policy and shapes future data collection. In this paper, we initiate the study of data attribution for online RL, focusing on the widely used Proximal Policy Optimization (PPO) algorithm. We start by establishing a \emph{local} attribution framework, interpreting model checkpoints with respect to the records in the recent training buffer. We design two target functions, capturing agent action and cumulative return respectively, and measure each record's contribution through gradient similarity between its training loss and these targets. We demonstrate the power of this framework through three concrete applications: diagnosis of learning, temporal analysis of behavior formation, and targeted intervention during training. Leveraging this framework, we further propose an algorithm, iterative influence-based filtering (IIF), for online RL training that iteratively performs experience filtering to refine policy updates. Across standard RL benchmarks (classic control, navigation, locomotion) to RLHF for large language models, IIF reduces sample complexity, speeds up training, and achieves higher returns. Together, these results open a new direction for making online RL more interpretable, efficient, and effective.

MoESD: Unveil Speculative Decoding's Potential for Accelerating Sparse MoE

arXiv:2505.19645v3 Announce Type: replace Abstract: Large Language Models (LLMs) have achieved remarkable success across many applications, with Mixture of Experts (MoE) models demonstrating great potential. Compared to traditional dense models, MoEs achieve better performance with less computation. Speculative decoding (SD) is a widely used technique to accelerate LLM inference without accuracy loss, but it has been considered efficient only for dense models. In this work, we first demonstrate that, under medium batch sizes, MoE surprisingly benefits more from SD than dense models. Furthermore, as MoE becomes sparser -- the prevailing trend in MoE designs -- the batch size range where SD acceleration is expected to be effective becomes broader. To quantitatively understand tradeoffs involved in SD, we develop a reliable modeling based on theoretical analyses. While current SD research primarily focuses on improving acceptance rates of algorithms, changes in workload and model architecture can still lead to degraded SD acceleration even with high acceptance rates. To address this limitation, we introduce a new metric 'target efficiency' that characterizes these effects, thus helping researchers identify system bottlenecks and understand SD acceleration more comprehensively. For scenarios like private serving, this work unveils a new perspective to speed up MoE inference, where existing solutions struggle. Experiments on different GPUs show up to 2.29x speedup for Qwen2-57B-A14B at medium batch sizes and validate our theoretical predictions.

Rethinking Probabilistic Circuit Parameter Learning

arXiv:2505.19982v2 Announce Type: replace Abstract: Probabilistic Circuits (PCs) offer a computationally scalable framework for generative modeling, supporting exact and efficient inference of a wide range of probabilistic queries. While recent advances have significantly improved the expressiveness and scalability of PCs, effectively training their parameters remains a challenge. In particular, a widely used optimization method, full-batch Expectation-Maximization (EM), requires processing the entire dataset before performing a single update, making it ineffective for large datasets. Although empirical extensions to the mini-batch setting, as well as gradient-based mini-batch algorithms, converge faster than full-batch EM, they generally underperform in terms of final likelihood. We investigate this gap by establishing a novel theoretical connection between these practical algorithms and the general EM objective. Our analysis reveals a fundamental issue that existing mini-batch EM and gradient-based methods fail to properly regularize distribution changes, causing each update to effectively ``overfit'' the current mini-batch. Motivated by this insight, we introduce anemone, a new mini-batch EM algorithm for PCs. Anemone applies an implicit adaptive learning rate to each parameter, scaled by how much it contributes to the likelihood of the current batch. Across extensive experiments on language, image, and DNA datasets, anemone consistently outperforms existing optimizers in both convergence speed and final performance.

Spurious Privacy Leakage in Neural Networks

arXiv:2505.20095v2 Announce Type: replace Abstract: Neural networks trained on real-world data often exhibit biases while simultaneously being vulnerable to privacy attacks aimed at extracting sensitive information. Despite extensive research on each problem individually, their intersection remains poorly understood. In this work, we investigate the privacy impact of spurious correlation bias. We introduce \emph{spurious privacy leakage}, a phenomenon in which spurious groups are significantly more vulnerable to privacy attacks than non-spurious groups. We observe that privacy disparity between groups increases in tasks with simpler objectives (e.g. fewer classes) due to spurious features. Counterintuitively, we demonstrate that spurious robust methods, designed to reduce spurious bias, fail to mitigate privacy disparity. Our analysis reveals that this occurs because robust methods can reduce reliance on spurious features for prediction, but do not prevent their memorization during training. Finally, we systematically compare the privacy of different model architectures trained with spurious data, demonstrating that, contrary to previous work, architectural choice can affect privacy evaluation.

LUMION: Fast Fault Recovery for ML Jobs Using Programmable Optical Fabrics

arXiv:2505.23105v2 Announce Type: replace Abstract: When accelerators fail in modern ML datacenters, operators migrate the affected ML training or inference jobs to entirely new racks. This approach, while preserving network performance, is highly inefficient, requiring datacenters to reserve full racks of idle accelerators for fault tolerance. In this paper, we address this resource inefficiency by introducing LUMION, a novel reconfigurable optical fabric for connecting accelerators within a datacenter rack. Instead of migrating entire ML jobs, LUMION dynamically integrates spare accelerators into ongoing workloads as failures occur, thereby maintaining consistent performance without costly migrations. We show the benefits of LUMION by building an end-to-end hardware prototype. Our experiments fine-tune Llama 3.2 and show that LUMION swaps a failed GPU with a healthy one and restarts the ML job within ~ 1 second of the failure. LUMION achieves higher inter-GPU bandwidth compared to traditional electrical racks after replacing failed accelerators with spare ones, leading to nearly 2X improvement in fine-tuning throughput.

Inference-time Scaling of Diffusion Models through Classical Search

arXiv:2505.23614v2 Announce Type: replace Abstract: Classical search algorithms have long underpinned modern artificial intelligence. In this work, we tackle the challenge of inference-time control in diffusion models -- adapting generated outputs to meet diverse test-time objectives -- using principles from classical search. We propose a general framework that orchestrates local and global search to efficiently navigate the generative space. It employs a theoretically grounded local search via annealed Langevin MCMC and performs compute-efficient global exploration using breadth-first and depth-first tree search. We evaluate our approach on a range of challenging domains, including planning, offline reinforcement learning, and image generation. Across all tasks, we observe significant gains in both performance and efficiency. These results show that classical search provides a principled and practical foundation for inference-time scaling in diffusion models. Project page at https://diffusion-inference-scaling.github.io/.

Rethinking Exact Unlearning under Exposure: Extracting Forgotten Data under Exact Unlearning in Large Language Model

arXiv:2505.24379v2 Announce Type: replace Abstract: Large Language Models are typically trained on datasets collected from the web, which may inadvertently contain harmful or sensitive personal information. To address growing privacy concerns, unlearning methods have been proposed to remove the influence of specific data from trained models. Of these, exact unlearning -- which retrains the model from scratch without the target data -- is widely regarded the gold standard for mitigating privacy risks in deployment. In this paper, we revisit this assumption in a practical deployment setting where both the pre- and post-unlearning logits API are exposed, such as in open-weight scenarios. Targeting this setting, we introduce a novel data extraction attack that leverages signals from the pre-unlearning model to guide the post-unlearning model, uncovering patterns that reflect the removed data distribution. Combining model guidance with a token filtering strategy, our attack significantly improves extraction success rates -- doubling performance in some cases -- across common benchmarks such as MUSE, TOFU, and WMDP. Furthermore, we demonstrate our attack's effectiveness on a simulated medical diagnosis dataset to highlight real-world privacy risks associated with exact unlearning. In light of our findings, which suggest that unlearning may, in a contradictory way, increase the risk of privacy leakage during real-world deployments, we advocate for evaluation of unlearning methods to consider broader threat models that account not only for post-unlearning models but also for adversarial access to prior checkpoints. Code is publicly available at: https://github.com/Nicholas0228/unlearned_data_extraction_llm.

Cascading Adversarial Bias from Injection to Distillation in Language Models

arXiv:2505.24842v2 Announce Type: replace Abstract: Model distillation has become essential for creating smaller, deployable language models that retain larger system capabilities. However, widespread deployment raises concerns about resilience to adversarial manipulation. This paper investigates vulnerability of distilled models to adversarial injection of biased content during training. We demonstrate that adversaries can inject subtle biases into teacher models through minimal data poisoning, which propagates to student models and becomes significantly amplified. We propose two propagation modes: Untargeted Propagation, where bias affects multiple tasks, and Targeted Propagation, focusing on specific tasks while maintaining normal behavior elsewhere. With only 25 poisoned samples (0.25% poisoning rate), student models generate biased responses 76.9% of the time in targeted scenarios - higher than 69.4% in teacher models. For untargeted propagation, adversarial bias appears 6x-29x more frequently in student models on unseen tasks. We validate findings across six bias types (targeted advertisements, phishing links, narrative manipulations, insecure coding practices), various distillation methods, and different modalities spanning text and code generation. Our evaluation reveals shortcomings in current defenses - perplexity filtering, bias detection systems, and LLM-based autorater frameworks - against these attacks. Results expose significant security vulnerabilities in distilled models, highlighting need for specialized safeguards. We propose practical design principles for building effective adversarial bias mitigation strategies.

Learning Semantics, Not Addresses: Runtime Neural Prefetching for Far Memory

arXiv:2506.00384v2 Announce Type: replace Abstract: Memory prefetching has long boosted CPU caches and is increasingly vital for far-memory systems, where large portions of memory are offloaded to cheaper, remote tiers. While effective prefetching requires accurate prediction of future accesses, prior ML approaches have been limited to simulation or small-scale hardware. We introduce FarSight, the first Linux-based far-memory system to leverage deep learning by decoupling application semantics from runtime memory layout. This separation enables offline-trained models to predict access patterns over a compact ordinal vocabulary, which are resolved at runtime through lightweight mappings. Across four data-intensive workloads, FarSight delivers up to 3.6x higher performance than the state-of-the-art.

In-Context Learning for Pure Exploration

arXiv:2506.01876v2 Announce Type: replace Abstract: We study the problem active sequential hypothesis testing, also known as pure exploration: given a new task, the learner adaptively collects data from the environment to efficiently determine an underlying correct hypothesis. A classical instance of this problem is the task of identifying the best arm in a multi-armed bandit problem (a.k.a. BAI, Best-Arm Identification), where actions index hypotheses. Another important case is generalized search, a problem of determining the correct label through a sequence of strategically selected queries that indirectly reveal information about the label. In this work, we introduce In-Context Pure Exploration (ICPE), which meta-trains Transformers to map observation histories to query actions and a predicted hypothesis, yielding a model that transfers in-context. At inference time, ICPE actively gathers evidence on new tasks and infers the true hypothesis without parameter updates. Across deterministic, stochastic, and structured benchmarks, including BAI and generalized search, ICPE is competitive with adaptive baselines while requiring no explicit modeling of information structure. Our results support Transformers as practical architectures for general sequential testing.

SALAD: Systematic Assessment of Machine Unlearning on LLM-Aided Hardware Design

arXiv:2506.02089v4 Announce Type: replace Abstract: Large Language Models (LLMs) offer transformative capabilities for hardware design automation, particularly in Verilog code generation. However, they also pose significant data security challenges, including Verilog evaluation data contamination, intellectual property (IP) design leakage, and the risk of malicious Verilog generation. We introduce SALAD, a comprehensive assessment that leverages machine unlearning to mitigate these threats. Our approach enables the selective removal of contaminated benchmarks, sensitive IP and design artifacts, or malicious code patterns from pre-trained LLMs, all without requiring full retraining. Through detailed case studies, we demonstrate how machine unlearning techniques effectively reduce data security risks in LLM-aided hardware design.

Reliably Detecting Model Failures in Deployment Without Labels

arXiv:2506.05047v3 Announce Type: replace Abstract: The distribution of data changes over time; models operating operating in dynamic environments need retraining. But knowing when to retrain, without access to labels, is an open challenge since some, but not all shifts degrade model performance. This paper formalizes and addresses the problem of post-deployment deterioration (PDD) monitoring. We propose D3M, a practical and efficient monitoring algorithm based on the disagreement of predictive models, achieving low false positive rates under non-deteriorating shifts and provides sample complexity bounds for high true positive rates under deteriorating shifts. Empirical results on both standard benchmark and a real-world large-scale internal medicine dataset demonstrate the effectiveness of the framework and highlight its viability as an alert mechanism for high-stakes machine learning pipelines.

Chasing Moving Targets with Online Self-Play Reinforcement Learning for Safer Language Models

arXiv:2506.07468v3 Announce Type: replace Abstract: Conventional language model (LM) safety alignment relies on a reactive, disjoint procedure: attackers exploit a static model, followed by defensive fine-tuning to patch exposed vulnerabilities. This sequential approach creates a mismatch -- attackers overfit to obsolete defenses, while defenders perpetually lag behind emerging threats. To address this, we propose Self-RedTeam, an online self-play reinforcement learning algorithm where an attacker and defender agent co-evolve through continuous interaction. We cast safety alignment as a two-player zero-sum game, where a single model alternates between attacker and defender roles -- generating adversarial prompts and safeguarding against them -- while a reward LM adjudicates outcomes. This enables dynamic co-adaptation. Grounded in the game-theoretic framework of zero-sum games, we establish a theoretical safety guarantee which motivates the design of our method: if self-play converges to a Nash Equilibrium, the defender will reliably produce safe responses to any adversarial input. Empirically, Self-RedTeam uncovers more diverse attacks (+21.8% SBERT) compared to attackers trained against static defenders and achieves higher robustness on safety benchmarks (e.g., +65.5% on WildJailBreak) than defenders trained against static attackers. We further propose hidden Chain-of-Thought, allowing agents to plan privately, which boosts adversarial diversity and reduces over-refusals. Our results motivate a shift from reactive patching to proactive co-evolution in LM safety training, enabling scalable, autonomous, and robust self-improvement of LMs via multi-agent reinforcement learning (MARL).

PhysioWave: A Multi-Scale Wavelet-Transformer for Physiological Signal Representation

arXiv:2506.10351v3 Announce Type: replace Abstract: Physiological signals are often corrupted by motion artifacts, baseline drift, and other low-SNR disturbances, which pose significant challenges for analysis. Additionally, these signals exhibit strong non-stationarity, with sharp peaks and abrupt changes that evolve continuously, making them difficult to represent using traditional time-domain or filtering methods. To address these issues, a novel wavelet-based approach for physiological signal analysis is presented, aiming to capture multi-scale time-frequency features in various physiological signals. Leveraging this technique, two large-scale pretrained models specific to EMG and ECG are introduced for the first time, achieving superior performance and setting new baselines in downstream tasks. Additionally, a unified multi-modal framework is constructed by integrating pretrained EEG model, where each modality is guided through its dedicated branch and fused via learnable weighted fusion. This design effectively addresses challenges such as low signal-to-noise ratio, high inter-subject variability, and device mismatch, outperforming existing methods on multi-modal tasks. The proposed wavelet-based architecture lays a solid foundation for analysis of diverse physiological signals, while the multi-modal design points to next-generation physiological signal processing with potential impact on wearable health monitoring, clinical diagnostics, and broader biomedical applications. Code and data are available at: github.com/ForeverBlue816/PhysioWave

Can We Infer Confidential Properties of Training Data from LLMs?

arXiv:2506.10364v3 Announce Type: replace Abstract: Large language models (LLMs) are increasingly fine-tuned on domain-specific datasets to support applications in fields such as healthcare, finance, and law. These fine-tuning datasets often have sensitive and confidential dataset-level properties -- such as patient demographics or disease prevalence -- that are not intended to be revealed. While prior work has studied property inference attacks on discriminative models (e.g., image classification models) and generative models (e.g., GANs for image data), it remains unclear if such attacks transfer to LLMs. In this work, we introduce PropInfer, a benchmark task for evaluating property inference in LLMs under two fine-tuning paradigms: question-answering and chat-completion. Built on the ChatDoctor dataset, our benchmark includes a range of property types and task configurations. We further propose two tailored attacks: a prompt-based generation attack and a shadow-model attack leveraging word frequency signals. Empirical evaluations across multiple pretrained LLMs show the success of our attacks, revealing a previously unrecognized vulnerability in LLMs.

Measurement-Aligned Sampling for Inverse Problem

arXiv:2506.11893v2 Announce Type: replace Abstract: Diffusion models provide a powerful way to incorporate complex prior information for solving inverse problems. However, existing methods struggle to correctly incorporate guidance from conflicting signals in the prior and measurement, and often failed to maximizing the consistency to the measurement, especially in the challenging setting of non-Gaussian or unknown noise. To address these issues, we propose Measurement-Aligned Sampling (MAS), a novel framework for linear inverse problem solving that flexibly balances prior and measurement information. MAS unifies and extends existing approaches such as DDNM, TMPD, while generalizing to handle both known Gaussian noise and unknown or non-Gaussian noise types. Extensive experiments demonstrate that MAS consistently outperforms state-of-the-art methods across a variety of tasks, while maintaining relatively low computational cost.

Curating art exhibitions using machine learning

arXiv:2506.19813v2 Announce Type: replace Abstract: Here we present a series of artificial models - a total of four related models - based on machine learning techniques that attempt to learn from existing exhibitions which have been curated by human experts, in order to be able to do similar curatorship work. Out of our four artificial intelligence models, three achieve a reasonable ability at imitating these various curators responsible for all those exhibitions, with various degrees of precision and curatorial coherence. In particular, we can conclude two key insights: first, that there is sufficient information in these exhibitions to construct an artificial intelligence model that replicates past exhibitions with an accuracy well above random choices; and second, that using feature engineering and carefully designing the architecture of modest size models can make them almost as good as those using the so-called large language models such as GPT in a brute force approach.

Grokking in LLM Pretraining? Monitor Memorization-to-Generalization without Test

arXiv:2506.21551v3 Announce Type: replace Abstract: This paper presents the first study of grokking in practical LLM pretraining. Specifically, we investigate when an LLM memorizes the training data, when its generalization on downstream tasks starts to improve, and what happens if there is a lag between the two. Unlike existing works studying when a small model generalizes to limited and specified tasks during thousands epochs' training on algorithmic data, we focus on a practical setting for LLMs, i.e., one-epoch pretraining of next-token prediction on a cross-domain, large-scale corpus, and generalization on diverse benchmark tasks covering math/commonsense reasoning, code generation, and domain-specific retrieval. Our study, for the first time, verifies that grokking still emerges in pretraining mixture-of-experts (MoE) LLMs, though different local data groups may enter their grokking stages asynchronously due to the heterogeneity of their distributions and attributions to others. To find a mechanistic interpretation of this local grokking, we investigate the dynamics of training data's pathways (i.e., expert choices across layers in MoE). Our primary discovery is that the pathways evolve from random, non-smooth across layers, instance-specific to more structured and transferable across samples, despite the converged pretraining loss. This depicts a transition from memorization to generalization. Two novel metrics are developed to quantify these patterns: one computes the pathway similarity between samples, while the other measures the consistency of aggregated experts between subsequent layers for each sample. These training data based metrics induce zero cost but can faithfully track and monitor the generalization of LLMs on downstream tasks, which, in conventional settings, requires costly instruction tuning and benchmark evaluation.

Optimas: Optimizing Compound AI Systems with Globally Aligned Local Rewards

arXiv:2507.03041v3 Announce Type: replace Abstract: Compound AI systems integrating multiple components, such as Large Language Models, specialized tools, and traditional machine learning models, are increasingly deployed to solve complex real-world tasks. However, optimizing compound systems remains challenging due to their non-differentiable structures and diverse configuration types across components, including prompts, hyperparameters, and model parameters. To address this challenge, we propose Optimas, a unified framework for effective optimization of compound systems. The core idea of Optimas is to maintain one Local Reward Function (LRF) per component, each satisfying a local-global alignment property, i.e., each component's local reward correlates with the global system performance. In each iteration, Optimas efficiently adapts the LRFs to maintain this property while simultaneously maximizing each component's local reward. This approach enables independent updates of heterogeneous configurations using the designated optimization method, while ensuring that local improvements consistently lead to performance gains. We present extensive evaluations across five real-world compound systems to demonstrate that Optimas outperforms strong baselines by an average improvement of 11.92%, offering a general and effective approach for improving compound systems. Our website is at https://optimas.stanford.edu.

Sampling-aware Adversarial Attacks Against Large Language Models

arXiv:2507.04446v3 Announce Type: replace Abstract: To guarantee safe and robust deployment of large language models (LLMs) at scale, it is critical to accurately assess their adversarial robustness. Existing adversarial attacks typically target harmful responses in single-point greedy generations, overlooking the inherently stochastic nature of LLMs and overestimating robustness. We show that for the goal of eliciting harmful responses, repeated sampling of model outputs during the attack complements prompt optimization and serves as a strong and efficient attack vector. By casting attacks as a resource allocation problem between optimization and sampling, we determine compute-optimal trade-offs and show that integrating sampling into existing attacks boosts success rates by up to 37\% and improves efficiency by up to two orders of magnitude. We further analyze how distributions of output harmfulness evolve during an adversarial attack, discovering that many common optimization strategies have little effect on output harmfulness. Finally, we introduce a label-free proof-of-concept objective based on entropy maximization, demonstrating how our sampling-aware perspective enables new optimization targets. Overall, our findings establish the importance of sampling in attacks to accurately assess and strengthen LLM safety at scale.

Inference-Time Scaling of Diffusion Language Models with Particle Gibbs Sampling

arXiv:2507.08390v2 Announce Type: replace Abstract: Discrete diffusion models have recently emerged as strong alternatives to autoregressive language models, matching their performance through large-scale training. However, inference-time control remains relatively underexplored. In this work, we study how to steer generation toward desired rewards without retraining the models. Prior methods typically resample or filter within a single denoising trajectory, optimizing rewards step-by-step without trajectory-level refinement. We introduce particle Gibbs sampling for diffusion language models (PG-DLM), a novel inference-time algorithm enabling trajectory-level refinement while preserving generation perplexity under reward optimization. PG-DLM constructs a Markov chain over full denoising trajectories and applies a conditional sequential Monte Carlo kernel to resample them. We derive theoretical guarantees for convergence, including asymptotic consistency and variance bounds. Within this framework, we further analyze trade-offs across four key axes for inference-time scaling under fixed budgets: iterations, samples, denoising steps, and reward estimation. Our analysis shows scaling iterations achieves the best reward-perplexity trade-off. Empirically, PG-DLM consistently outperforms prior methods using MDLM and LLaDA-8B as base models across a wide range of compute budgets for reward-guided generation tasks including toxicity and sentiment control as well as linguistic acceptability.

GUIDE: Towards Scalable Advising for Research Ideas

arXiv:2507.08870v2 Announce Type: replace Abstract: The field of AI research is advancing at an unprecedented pace, enabling automated hypothesis generation and experimental design across diverse domains such as biology, mathematics, and artificial intelligence. Despite these advancements, there remains a significant gap in the availability of scalable advising systems capable of providing high-quality, well-reasoned feedback to refine proposed hypotheses and experimental designs. To address this challenge, we explore key factors that underlie the development of robust advising systems, including model size, context length, confidence estimation, and structured reasoning processes. Our findings reveal that a relatively small model, when equipped with a well-compressed literature database and a structured reasoning framework, can outperform powerful general-purpose language models such as Deepseek-R1 in terms of acceptance rates for self-ranked top-30% submissions to ICLR 2025. Moreover, when limited to high-confidence predictions, our system achieves an acceptance rate exceeding 90% on the ICLR 2025 test set, underscoring its potential to significantly enhance the quality and efficiency of hypothesis generation and experimental design. The code is released at https://github.com/HowardLiu0830/GUIDE-Research-Idea-Evaluation.

TolerantECG: A Foundation Model for Imperfect Electrocardiogram

arXiv:2507.09887v4 Announce Type: replace Abstract: The electrocardiogram (ECG) is an essential and effective tool for diagnosing heart diseases. However, its effectiveness can be compromised by noise or unavailability of one or more leads of the standard 12-lead recordings, resulting in diagnostic errors or uncertainty. To address these challenges, we propose TolerantECG, a foundation model for ECG signals that is robust to noise and capable of functioning with arbitrary subsets of the standard 12-lead ECG. TolerantECG training combines contrastive and self-supervised learning frameworks to jointly learn ECG signal representations alongside their corresponding knowledge-retrieval-based text report descriptions and corrupted or lead-missing signals. Comprehensive benchmarking results demonstrate that TolerantECG consistently ranks as the best or second-best performer across various ECG signal conditions and class levels in the PTB-XL dataset, and achieves the highest performance on the MIT-BIH Arrhythmia Database.

Thought Purity: A Defense Framework For Chain-of-Thought Attack

arXiv:2507.12314v2 Announce Type: replace Abstract: While reinforcement learning-trained Large Reasoning Models (LRMs, e.g., Deepseek-R1) demonstrate advanced reasoning capabilities in the evolving Large Language Models (LLMs) domain, their susceptibility to security threats remains a critical vulnerability. This weakness is particularly evident in Chain-of-Thought (CoT) generation processes, where adversarial methods like backdoor prompt attacks can systematically subvert the model's core reasoning mechanisms. The emerging Chain-of-Thought Attack (CoTA) reveals this vulnerability through exploiting prompt controllability, simultaneously degrading both CoT safety and task performance with low-cost interventions. To address this compounded security-performance vulnerability, we propose Thought Purity (TP): a defense framework that systematically strengthens resistance to malicious content while preserving operational efficacy. Our solution achieves this through three synergistic components: (1) a safety-optimized data processing pipeline (2) reinforcement learning-enhanced rule constraints (3) adaptive monitoring metrics. Our approach establishes the first comprehensive defense mechanism against CoTA vulnerabilities in reinforcement learning-aligned reasoning systems, significantly advancing the security-functionality equilibrium for next-generation AI architectures.

Solar Photovoltaic Assessment with Large Language Model

arXiv:2507.19144v2 Announce Type: replace Abstract: Accurate detection and localization of solar photovoltaic (PV) panels in satellite imagery is essential for optimizing microgrids and active distribution networks (ADNs), which are critical components of renewable energy systems. Existing methods lack transparency regarding their underlying algorithms or training datasets, rely on large, high-quality PV training data, and struggle to generalize to new geographic regions or varied environmental conditions without extensive re-training. These limitations lead to inconsistent detection outcomes, hindering large-scale deployment and data-driven grid optimization. In this paper, we investigate how large language models (LLMs) can be leveraged to overcome these challenges. Despite their promise, LLMs face several challenges in solar panel detection, including difficulties with multi-step logical processes, inconsistent output formatting, frequent misclassification of visually similar objects (e.g., shadows, parking lots), and low accuracy in complex tasks such as spatial localization and quantification. To overcome these issues, we propose the PV Assessment with LLMs (PVAL) framework, which incorporates task decomposition for more efficient workflows, output standardization for consistent and scalable formatting, few-shot prompting to enhance classification accuracy, and fine-tuning using curated PV datasets with detailed annotations. PVAL ensures transparency, scalability, and adaptability across heterogeneous datasets while minimizing computational overhead. By combining open-source accessibility with robust methodologies, PVAL establishes an automated and reproducible pipeline for solar panel detection, paving the way for large-scale renewable energy integration and optimized grid management.

First Hallucination Tokens Are Different from Conditional Ones

arXiv:2507.20836v4 Announce Type: replace Abstract: Large Language Models (LLMs) hallucinate, and detecting these cases is key to ensuring trust. While many approaches address hallucination detection at the response or span level, recent work explores token-level detection, enabling more fine-grained intervention. However, the distribution of hallucination signal across sequences of hallucinated tokens remains unexplored. We leverage token-level annotations from the RAGTruth corpus and find that the first hallucinated token is far more detectable than later ones. This structural property holds across models, suggesting that first hallucination tokens play a key role in token-level hallucination detection. Our code is available at https://github.com/jakobsnl/RAGTruth_Xtended.

VITA: Variational Pretraining of Transformers for Climate-Robust Crop Yield Forecasting

arXiv:2508.03589v2 Announce Type: replace Abstract: Accurate crop yield forecasting is essential for global food security. However, current AI models systematically underperform when yields deviate from historical trends. We attribute this to the lack of rich, physically grounded datasets directly linking atmospheric states to yields. To address this, we introduce VITA (Variational Inference Transformer for Asymmetric data), a variational pretraining framework that learns representations from large satellite-based weather datasets and transfers to the ground-based limited measurements available for yield prediction. VITA is trained using detailed meteorological variables as proxy targets during pretraining and learns to predict latent atmospheric states under a seasonality-aware sinusoidal prior. This allows the model to be fine-tuned using limited weather statistics during deployment. Applied to 763 counties in the U.S. Corn Belt, VITA achieves state-of-the-art performance in predicting corn and soybean yields across all evaluation scenarios, particularly during extreme years, with statistically significant improvements (paired t-test, $p < 0.0001$). Importantly, VITA outperforms prior frameworks like GNN-RNN without soil data, and bigger foundational models (e.g., Chronos-Bolt) with less compute, making it practical for real-world use--especially in data-scarce regions. This work highlights how domain-aware AI design can overcome data limitations and support resilient agricultural forecasting in a changing climate.

Synaptic Pruning: A Biological Inspiration for Deep Learning Regularization

arXiv:2508.09330v2 Announce Type: replace Abstract: Synaptic pruning in biological brains removes weak connections to improve efficiency. In contrast, dropout regularization in artificial neural networks randomly deactivates neurons without considering activity-dependent pruning. We propose a magnitude-based synaptic pruning method that better reflects biology by progressively removing low-importance connections during training. Integrated directly into the training loop as a dropout replacement, our approach computes weight importance from absolute magnitudes across layers and applies a cubic schedule to gradually increase global sparsity. At fixed intervals, pruning masks permanently remove low-importance weights while maintaining gradient flow for active ones, eliminating the need for separate pruning and fine-tuning phases. Experiments on multiple time series forecasting models including RNN, LSTM, and Patch Time Series Transformer across four datasets show consistent gains. Our method ranked best overall, with statistically significant improvements confirmed by Friedman tests (p < 0.01). In financial forecasting, it reduced Mean Absolute Error by up to 20% over models with no or standard dropout, and up to 52% in select transformer models. This dynamic pruning mechanism advances regularization by coupling weight elimination with progressive sparsification, offering easy integration into diverse architectures. Its strong performance, especially in financial time series forecasting, highlights its potential as a practical alternative to conventional dropout techniques.

Depth-Breadth Synergy in RLVR: Unlocking LLM Reasoning Gains with Adaptive Exploration

arXiv:2508.13755v4 Announce Type: replace Abstract: Reinforcement Learning with Verifiable Reward (RLVR) has emerged as a powerful paradigm for unlocking reasoning capabilities in large language models, yet its full potential is hindered by two under-explored dimensions: Depth-the hardest problem a model can sample; Breadth-the number of instances consumed in a single iteration. We dissect the popular GRPO algorithm and reveal a systematic bias: the cumulative-advantage disproportionately weights samples with medium accuracy, while down-weighting the low-accuracy instances that are crucial for pushing reasoning boundaries. To rectify the depth neglect, we introduce Difficulty Adaptive Rollout Sampling (DARS), which re-weights hard problems through targeted multi-stage rollouts, thereby increasing the number of positive rollouts for hard problems. Empirically, naively enlarging rollout size only accelerates convergence and even hurts Pass@K. Our DARS, in contrast, delivers consistent Pass@K gains without extra inference cost at convergence. Just as we adaptively expanded the depth of exploration, we now ask whether aggressively scaling the breadth of training data can further amplify reasoning gains. To this end, we intensely scale batch size and replace PPO's mini-batch iterations with full-batch updates over multiple epochs. Increasing breadth significantly enhances Pass@1 performance. Large-breadth training sustains high token-level entropy, indicating continued exploration and reduced gradient noise. We further present DARS-B, which augments DARS with large breadth, and demonstrate simultaneous gains in Pass@K and Pass@1. The results confirm that breadth and adaptive exploration across depth operate as orthogonal dimensions in RLVR, which are key to unleashing the reasoning power of RLVR.

On Zero-Shot Reinforcement Learning

arXiv:2508.16496v2 Announce Type: replace Abstract: Modern reinforcement learning (RL) systems capture deep truths about general, human problem-solving. In domains where new data can be simulated cheaply, these systems uncover sequential decision-making policies that far exceed the ability of any human. Society faces many problems whose solutions require this skill, but they are often in domains where new data cannot be cheaply simulated. In such scenarios, we can learn simulators from existing data, but these will only ever be approximately correct, and can be pathologically incorrect when queried outside of their training distribution. As a result, a misalignment between the environments in which we train our agents and the real-world in which we wish to deploy our agents is inevitable. Dealing with this misalignment is the primary concern of zero-shot reinforcement learning, a problem setting where the agent must generalise to a new task or domain with zero practice shots. Whilst impressive progress has been made on methods that perform zero-shot RL in idealised settings, new work is needed if these results are to be replicated in real-world settings. In this thesis, we argue that doing so requires us to navigate (at least) three constraints. First, the data quality constraint: real-world datasets are small and homogeneous. Second, the observability constraint: states, dynamics and rewards in the real-world are often only partially observed. And third, the data availability constraint: a priori access to data cannot always be assumed. This work proposes a suite of methods that perform zero-shot RL subject to these constraints. In a series of empirical studies we expose the failings of existing methods, and justify our techniques for remedying them. We believe these designs take us a step closer to RL methods that can be deployed to solve real-world problems.

Owen Sampling Accelerates Contribution Estimation in Federated Learning

arXiv:2508.21261v2 Announce Type: replace Abstract: Federated Learning (FL) aggregates information from multiple clients to train a shared global model without exposing raw data. Accurately estimating each client's contribution is essential not just for fair rewards, but for selecting the most useful clients so the global model converges faster. The Shapley value is a principled choice, yet exact computation scales exponentially with the number of clients, making it infeasible for large federations. We propose FedOwen, an efficient framework that uses Owen sampling to approximate Shapley values under the same total evaluation budget as existing methods while keeping the approximation error small. In addition, FedOwen uses an adaptive client selection strategy that balances exploiting high-value clients with exploring under-sampled ones, reducing bias and uncovering rare but informative data. Under a fixed valuation cost, FedOwen achieves up to 23 percent higher final accuracy within the same number of communication rounds compared to state-of-the-art baselines on non-IID benchmarks.

Attention as an Adaptive Filter

arXiv:2509.04154v2 Announce Type: replace Abstract: We introduce Adaptive Filter Attention (AFA), a novel attention mechanism that incorporates a learnable dynamics model directly into the computation of attention weights. Rather than comparing queries and keys directly, we model the input sequence as discrete observations of a linear stochastic differential equation (SDE). By imposing a linear dynamics model with simultaneously diagonalizable state matrices and noise covariances, we can make use of a closed-form solution to the differential Lyapunov equation to efficiently propagate pairwise uncertainties through the dynamics. Attention naturally arises as the maximum likelihood solution for this linear SDE, with attention weights corresponding to robust residual-based reweightings of the propagated pairwise precisions. Imposing an additional constraint on the state matrix's eigenvalues leads to a simplified variant with the same computational and memory complexity as standard attention. In the limit of vanishing dynamics and process noise, and using a small-angle approximation, we recover ordinary dot-product attention.

time2time: Causal Intervention in Hidden States to Simulate Rare Events in Time Series Foundation Models

arXiv:2509.05801v2 Announce Type: replace Abstract: While transformer-based foundation models excel at forecasting routine patterns, two questions remain: do they internalize semantic concepts such as market regimes, or merely fit curves? And can their internal representations be leveraged to simulate rare, high-stakes events such as market crashes? To investigate this, we introduce activation transplantation, a causal intervention that manipulates hidden states by imposing the statistical moments of one event (e.g., a historical crash) onto another (e.g., a calm period) during the forward pass. This procedure deterministically steers forecasts: injecting crash semantics induces downturn predictions, while injecting calm semantics suppresses crashes and restores stability. Beyond binary control, we find that models encode a graded notion of event severity, with the latent vector norm directly correlating with the magnitude of systemic shocks. Validated across two architecturally distinct TSFMs, Toto (decoder only) and Chronos (encoder-decoder), our results demonstrate that steerable, semantically grounded representations are a robust property of large time series transformers. Our findings provide evidence for a latent concept space that governs model predictions, shifting interpretability from post-hoc attribution to direct causal intervention, and enabling semantic "what-if" analysis for strategic stress-testing.

Physics-informed Value Learner for Offline Goal-Conditioned Reinforcement Learning

arXiv:2509.06782v2 Announce Type: replace Abstract: Offline Goal-Conditioned Reinforcement Learning (GCRL) holds great promise for domains such as autonomous navigation and locomotion, where collecting interactive data is costly and unsafe. However, it remains challenging in practice due to the need to learn from datasets with limited coverage of the state-action space and to generalize across long-horizon tasks. To improve on these challenges, we propose a \emph{Physics-informed (Pi)} regularized loss for value learning, derived from the Eikonal Partial Differential Equation (PDE) and which induces a geometric inductive bias in the learned value function. Unlike generic gradient penalties that are primarily used to stabilize training, our formulation is grounded in continuous-time optimal control and encourages value functions to align with cost-to-go structures. The proposed regularizer is broadly compatible with temporal-difference-based value learning and can be integrated into existing Offline GCRL algorithms. When combined with Hierarchical Implicit Q-Learning (HIQL), the resulting method, Eikonal-regularized HIQL (Eik-HIQL), yields significant improvements in both performance and generalization, with pronounced gains in stitching regimes and large-scale navigation tasks.

FRAUDGUESS: Spotting and Explaining New Types of Fraud in Million-Scale Financial Data

arXiv:2509.15493v2 Announce Type: replace Abstract: Given a set of financial transactions (who buys from whom, when, and for how much), as well as prior information from buyers and sellers, how can we find fraudulent transactions? If we have labels for some transactions for known types of fraud, we can build a classifier. However, we also want to find new types of fraud, still unknown to the domain experts ('Detection'). Moreover, we also want to provide evidence to experts that supports our opinion ('Justification'). In this paper, we propose FRAUDGUESS, to achieve two goals: (a) for 'Detection', it spots new types of fraud as micro-clusters in a carefully designed feature space; (b) for 'Justification', it uses visualization and heatmaps for evidence, as well as an interactive dashboard for deep dives. FRAUDGUESS is used in real life and is currently considered for deployment in an Anonymous Financial Institution (AFI). Thus, we also present the three new behaviors that FRAUDGUESS discovered in a real, million-scale financial dataset. Two of these behaviors are deemed fraudulent or suspicious by domain experts, catching hundreds of fraudulent transactions that would otherwise go un-noticed.

Reinforced Generation of Combinatorial Structures: Applications to Complexity Theory

arXiv:2509.18057v4 Announce Type: replace Abstract: We explore whether techniques from AI can help discover new combinatorial structures that improve on known limits on efficient algorithms. Specifically, we use AlphaEvolve (an LLM coding agent) to study two settings: a) Average-case hardness for MAX-CUT and MAX-Independent Set: We improve a recent result of Kunisky and Yu to obtain near-optimal upper and (conditional) lower bounds on certification algorithms for MAX-CUT and MAX-Independent Set on random 3- and 4-regular graphs. Our improved lower bounds are obtained by constructing nearly extremal Ramanujan graphs on as many as $163$ nodes, using AlphaEvolve. Additionally, via analytical arguments we strengthen the upper bounds to settle the computational hardness of these questions up to an error in the third decimal place. b) Worst-case Hardness of Approximation for MAX-k-CUT: We obtain new inapproximability results, proving that it is NP-hard to approximate MAX-4-CUT and MAX-3-CUT within factors of $0.987$ and $0.9649$ respectively, using AlphaEvolve to discover new gadget reductions. Our MAX-4-CUT result improves upon the SOTA of $0.9883$, and our MAX-3-CUT result improves on the current best gadget-based inapproximability result of $0.9853$, but falls short of improving the SOTA of $16/17$ that relies on a custom PCP, rather than a gadget reduction from "standard" H{\aa}stad-style PCPs. A key technical challenge we faced: verifying a candidate construction produced by AlphaEvolve is costly (often requiring exponential time). In both settings above, our results were enabled by using AlphaEvolve itself to evolve the verification procedure to be faster (sometimes by $10,000\times$). We conclude with a discussion of norms by which to assess the assistance from AI in developing proofs.

Analyzing Uncertainty Quantification in Statistical and Deep Learning Models for Probabilistic Electricity Price Forecasting

arXiv:2509.19417v2 Announce Type: replace Abstract: Precise probabilistic forecasts are fundamental for energy risk management, and there is a wide range of both statistical and machine learning models for this purpose. Inherent to these probabilistic models is some form of uncertainty quantification. However, most models do not capture the full extent of uncertainty, which arises not only from the data itself but also from model and distributional choices. In this study, we examine uncertainty quantification in state-of-the-art statistical and deep learning probabilistic forecasting models for electricity price forecasting in the German market. In particular, we consider deep distributional neural networks (DDNNs) and augment them with an ensemble approach, Monte Carlo (MC) dropout, and conformal prediction to account for model uncertainty. Additionally, we consider the LASSO-estimated autoregressive (LEAR) approach combined with quantile regression averaging (QRA), generalized autoregressive conditional heteroskedasticity (GARCH), and conformal prediction. Across a range of performance metrics, we find that the LEAR-based models perform well in terms of probabilistic forecasting, irrespective of the uncertainty quantification method. Furthermore, we find that DDNNs benefit from incorporating both data and model uncertainty, improving both point and probabilistic forecasting. Uncertainty itself appears to be best captured by the models using conformal prediction. Overall, our extensive study shows that all models under consideration perform competitively. However, their relative performance depends on the choice of metrics for point and probabilistic forecasting.

Frame-based Equivariant Diffusion Models for 3D Molecular Generation

arXiv:2509.19506v2 Announce Type: replace Abstract: Recent methods for molecular generation face a trade-off: they either enforce strict equivariance with costly architectures or relax it to gain scalability and flexibility. We propose a frame-based diffusion paradigm that achieves deterministic E(3)-equivariance while decoupling symmetry handling from the backbone. Building on this paradigm, we investigate three variants: Global Frame Diffusion (GFD), which assigns a shared molecular frame; Local Frame Diffusion (LFD), which constructs node-specific frames and benefits from additional alignment constraints; and Invariant Frame Diffusion (IFD), which relies on pre-canonicalized invariant representations. To enhance expressivity, we further utilize EdgeDiT, a Diffusion Transformer with edge-aware attention. On the QM9 dataset, GFD with EdgeDiT achieves state-of-the-art performance, with a test NLL of -137.97 at standard scale and -141.85 at double scale, alongside atom stability of 98.98%, and molecular stability of 90.51%. These results surpass all equivariant baselines while maintaining high validity and uniqueness and nearly 2x faster sampling compared to EDM. Altogether, our study establishes frame-based diffusion as a scalable, flexible, and physically grounded paradigm for molecular generation, highlighting the critical role of global structure preservation.

When Judgment Becomes Noise: How Design Failures in LLM Judge Benchmarks Silently Undermine Validity

arXiv:2509.20293v2 Announce Type: replace Abstract: LLM-judged benchmarks are increasingly used to evaluate complex model behaviors, yet their design introduces failure modes absent in conventional ground-truth based benchmarks. We argue that without tight objectives and verifiable constructions, benchmark rankings can produce high-confidence rankings that are in fact largely noise. We introduce two mechanisms to diagnose these issues. Schematic adherence quantifies how much of a judge's overall verdict is explained by the explicit evaluation schema, revealing unexplained variance when judges deviate from their own rubric. Psychometric validity aggregates internal consistency and discriminant validity signals to quantify irreducible uncertainty in any benchmarking run. Applying these tools to Arena-Hard Auto, we find severe schema incoherence and factor collapse across popular judges: for example, unexplained variance exceeding 90 percent for DeepSeek-R1-32B and factor correlations above 0.93 for most criteria. We also show that the ELO-style aggregation used by Arena-Hard Auto collapses and masks genuine ranking uncertainty. Our results highlight design failures that undermine validity and offer actionable principles for building better-scoped, reliability-aware LLM-judged benchmarks. We released our code and dataset at https://github.com/penfever/judgment-to-noise

CE-GPPO: Coordinating Entropy via Gradient-Preserving Clipping Policy Optimization in Reinforcement Learning

arXiv:2509.20712v3 Announce Type: replace Abstract: Reinforcement learning (RL) has become a powerful paradigm for optimizing large language models (LLMs) to handle complex reasoning tasks. A core challenge in this process lies in managing policy entropy, which reflects the balance between exploration and exploitation during training. Existing methods, such as proximal policy optimization (PPO) and its variants, discard valuable gradient signals from low-probability tokens due to the clipping mechanism. We systematically analyze the entropy dynamics and reveal that these clipped tokens play a critical yet overlooked role in regulating entropy evolution. We propose \textbf{C}oordinating \textbf{E}ntropy via \textbf{G}radient-\textbf{P}reserving \textbf{P}olicy \textbf{O}ptimization (CE-GPPO), a novel algorithm that reintroduces gradients from clipped tokens in native PPO in a gentle and bounded manner. By controlling the magnitude of gradients from tokens outside the clipping interval, CE-GPPO is able to achieve an exploration-exploitation trade-off. We provide theoretical justification and empirical evidence showing that CE-GPPO effectively mitigates entropy instability. Extensive experiments on mathematical reasoning benchmarks show that CE-GPPO consistently outperforms strong baselines across different model scales.

RL Grokking Recipe: How Does RL Unlock and Transfer New Algorithms in LLMs?

arXiv:2509.21016v2 Announce Type: replace Abstract: It remains an open question whether LLMs can acquire or generalize genuinely new reasoning strategies, beyond the sharpened skills encoded in their parameters during pre-training or post-training. To attempt to answer this debate, we introduce DELTA-Code -- Distributional Evaluation of Learnability and Transferrability in Algorithmic Coding -- a controlled benchmark of synthetic coding problem families designed to probe two fundamental aspects: learnability -- can LLMs, through reinforcement learning (RL), solve problem families where pretrained models exhibit failure with large enough attempts (pass@K=0)? -- and transferrability -- if learnability happens, can such skills transfer systematically to out-of-distribution (OOD) test sets? Unlike prior public coding datasets, DELTA isolates reasoning skills through templated problem generators and introduces fully OOD problem families that demand novel strategies rather than tool invocation or memorized patterns. Our experiments reveal a striking grokking phase transition: after an extended period with near-zero reward, RL-trained models abruptly climb to near-perfect accuracy. To enable learnability on previously unsolvable problem families, we explore key training ingredients such as staged warm-up with dense rewards, experience replay, curriculum training, and verification-in-the-loop. Beyond learnability, we use DELTA to evaluate transferability or generalization along exploratory, compositional, and transformative axes, as well as cross-family transfer. Results show solid gains within families and for recomposed skills, but persistent weaknesses in transformative cases. DELTA thus offers a clean testbed for probing the limits of RL-driven reasoning and for understanding how models can move beyond existing priors to acquire new algorithmic skills.

Active Attacks: Red-teaming LLMs via Adaptive Environments

arXiv:2509.21947v2 Announce Type: replace Abstract: We address the challenge of generating diverse attack prompts for large language models (LLMs) that elicit harmful behaviors (e.g., insults, sexual content) and are used for safety fine-tuning. Rather than relying on manual prompt engineering, attacker LLMs can be trained with reinforcement learning (RL) to automatically generate such prompts using only a toxicity classifier as a reward. However, capturing a wide range of harmful behaviors is a significant challenge that requires explicit diversity objectives. Existing diversity-seeking RL methods often collapse to limited modes: once high-reward prompts are found, exploration of new regions is discouraged. Inspired by the active learning paradigm that encourages adaptive exploration, we introduce \textit{Active Attacks}, a novel RL-based red-teaming algorithm that adapts its attacks as the victim evolves. By periodically safety fine-tuning the victim LLM with collected attack prompts, rewards in exploited regions diminish, which forces the attacker to seek unexplored vulnerabilities. This process naturally induces an easy-to-hard exploration curriculum, where the attacker progresses beyond easy modes toward increasingly difficult ones. As a result, Active Attacks uncovers a wide range of local attack modes step by step, and their combination achieves wide coverage of the multi-mode distribution. Active Attacks, a simple plug-and-play module that seamlessly integrates into existing RL objectives, unexpectedly outperformed prior RL-based methods -- including GFlowNets, PPO, and REINFORCE -- by improving cross-attack success rates against GFlowNets, the previous state-of-the-art, from 0.07% to 31.28% (a relative gain greater than $400\ \times$) with only a 6% increase in computation. Our code is publicly available \href{https://github.com/dbsxodud-11/active_attacks}{here}.

Boundary on the Table: Efficient Black-Box Decision-Based Attacks for Structured Data

arXiv:2509.22850v2 Announce Type: replace Abstract: Adversarial robustness in structured data remains an underexplored frontier compared to vision and language domains. In this work, we introduce a novel black-box, decision-based adversarial attack tailored for tabular data. Our approach combines gradient-free direction estimation with an iterative boundary search, enabling efficient navigation of discrete and continuous feature spaces under minimal oracle access. Extensive experiments demonstrate that our method successfully compromises nearly the entire test set across diverse models, ranging from classical machine learning classifiers to large language model (LLM)-based pipelines. Remarkably, the attack achieves success rates consistently above 90%, while requiring only a small number of queries per instance. These results highlight the critical vulnerability of tabular models to adversarial perturbations, underscoring the urgent need for stronger defenses in real-world decision-making systems.

Understanding Catastrophic Interference On the Identifibility of Latent Representations

arXiv:2509.23027v2 Announce Type: replace Abstract: Catastrophic interference, also known as catastrophic forgetting, is a fundamental challenge in machine learning, where a trained learning model progressively loses performance on previously learned tasks when adapting to new ones. In this paper, we aim to better understand and model the catastrophic interference problem from a latent representation learning point of view, and propose a novel theoretical framework that formulates catastrophic interference as an identification problem. Our analysis demonstrates that the forgetting phenomenon can be quantified by the distance between partial-task aware (PTA) and all-task aware (ATA) setups. Building upon recent advances in identifiability theory, we prove that this distance can be minimized through identification of shared latent variables between these setups. When learning, we propose our method \ourmeos with two-stage training strategy: First, we employ maximum likelihood estimation to learn the latent representations from both PTA and ATA configurations. Subsequently, we optimize the KL divergence to identify and learn the shared latent variables. Through theoretical guarantee and empirical validations, we establish that identifying and learning these shared representations can effectively mitigate catastrophic interference in machine learning systems. Our approach provides both theoretical guarantees and practical performance improvements across both synthetic and benchmark datasets.

NanoFlux: Adversarial Dual-LLM Evaluation and Distillation For Multi-Domain Reasoning

arXiv:2509.23252v2 Announce Type: replace Abstract: We present NanoFlux, a novel adversarial framework for generating targeted training data to improve LLM reasoning, where adversarially-generated datasets containing fewer than 200 examples outperform conventional fine-tuning approaches. The framework employs a competitive dynamic between models alternating as Attacker and Defender, supervised by a tool-augmented Judge, synthesizing multi-step questions with explanatory annotations that target specific reasoning capabilities. Fine-tuning a 4B-parameter model on NanoFlux-generated data yields performance gains across diverse domains compared to full-benchmark fine-tuning: +5.9% on mathematical reasoning (GSMHard), +3.6% on scientific reasoning (GenomeBench), and +16.6% on medical reasoning (MultiMedQA), while reducing computational requirements by 3-14x. Ablation studies reveal a non-monotonic relationship between dataset characteristics and model performance, uncovering domain-specific optimal points for question complexity and reasoning quality. NanoFlux automates training data generation through embedding-based novelty filtering, tool-augmented evaluation, and multi-hop reasoning, suggesting that future model improvements may lie in the intelligent synthesis of small, precisely targeted training datasets.

Emergence of Superposition: Unveiling the Training Dynamics of Chain of Continuous Thought

arXiv:2509.23365v2 Announce Type: replace Abstract: Previous work shows that the chain of continuous thought (continuous CoT) improves the reasoning capability of large language models (LLMs) by enabling implicit parallel thinking, and a subsequent work provided theoretical insight by showing that a two-layer transformer equipped with continuous CoT can efficiently solve directed graph reachability by maintaining a superposition of multiple reasoning traces in the continuous thought. However, it remains unclear how the superposition mechanism is naturally learned from gradient-based training methods. To fill this gap, we theoretically analyze the training dynamics of a simplified two-layer transformer on the directed graph reachability problem to unveil how the superposition mechanism emerges during training in two training stages -- (i) a thought-generation stage that autoregressively expands the continuous thought, and (ii) a prediction stage that converts the thought into the final answer. Our analysis reveals that during training using continuous thought, the index-matching logit, an important quantity which reflects the strength of the model's local search ability, will first increase and then remain bounded under mild assumptions. The bounded index-matching logit effectively balances exploration and exploitation during the reasoning process: the model will exploit local problem structures to identify plausible search traces, and assign comparable weights to multiple such traces to explore when it is uncertain about which solution is correct, which results in superposition. Our experimental results tracking the growth of logits further validate our theory.

InfMasking: Unleashing Synergistic Information by Contrastive Multimodal Interactions

arXiv:2509.25270v2 Announce Type: replace Abstract: In multimodal representation learning, synergistic interactions between modalities not only provide complementary information but also create unique outcomes through specific interaction patterns that no single modality could achieve alone. Existing methods may struggle to effectively capture the full spectrum of synergistic information, leading to suboptimal performance in tasks where such interactions are critical. This is particularly problematic because synergistic information constitutes the fundamental value proposition of multimodal representation. To address this challenge, we introduce InfMasking, a contrastive synergistic information extraction method designed to enhance synergistic information through an Infinite Masking strategy. InfMasking stochastically occludes most features from each modality during fusion, preserving only partial information to create representations with varied synergistic patterns. Unmasked fused representations are then aligned with masked ones through mutual information maximization to encode comprehensive synergistic information. This infinite masking strategy enables capturing richer interactions by exposing the model to diverse partial modality combinations during training. As computing mutual information estimates with infinite masking is computationally prohibitive, we derive an InfMasking loss to approximate this calculation. Through controlled experiments, we demonstrate that InfMasking effectively enhances synergistic information between modalities. In evaluations on large-scale real-world datasets, InfMasking achieves state-of-the-art performance across seven benchmarks. Code is released at https://github.com/brightest66/InfMasking.

Uncertainty-Aware Generative Oversampling Using an Entropy-Guided Conditional Variational Autoencoder

arXiv:2509.25334v3 Announce Type: replace Abstract: Class imbalance remains a major challenge in machine learning, especially for high-dimensional biomedical data where nonlinear manifold structures dominate. Traditional oversampling methods such as SMOTE rely on local linear interpolation, often producing implausible synthetic samples. Deep generative models like Conditional Variational Autoencoders (CVAEs) better capture nonlinear distributions, but standard variants treat all minority samples equally, neglecting the importance of uncertain, boundary-region examples emphasized by heuristic methods like Borderline-SMOTE and ADASYN. We propose Local Entropy-Guided Oversampling with a CVAE (LEO-CVAE), a generative oversampling framework that explicitly incorporates local uncertainty into both representation learning and data generation. To quantify uncertainty, we compute Shannon entropy over the class distribution in a sample's neighborhood: high entropy indicates greater class overlap, serving as a proxy for uncertainty. LEO-CVAE leverages this signal through two mechanisms: (i) a Local Entropy-Weighted Loss (LEWL) that emphasizes robust learning in uncertain regions, and (ii) an entropy-guided sampling strategy that concentrates generation in these informative, class-overlapping areas. Applied to clinical genomics datasets (ADNI and TCGA lung cancer), LEO-CVAE consistently improves classifier performance, outperforming both traditional oversampling and generative baselines. These results highlight the value of uncertainty-aware generative oversampling for imbalanced learning in domains governed by complex nonlinear structures, such as omics data.

Scalable Disk-Based Approximate Nearest Neighbor Search with Page-Aligned Graph

arXiv:2509.25487v2 Announce Type: replace Abstract: Approximate Nearest Neighbor Search (ANNS), as the core of vector databases (VectorDBs), has become widely used in modern AI and ML systems, powering applications from information retrieval to bio-informatics. While graph-based ANNS methods achieve high query efficiency, their scalability is constrained by the available host memory. Recent disk-based ANNS approaches mitigate memory usage by offloading data to Solid-State Drives (SSDs). However, they still suffer from issues such as long I/O traversal path, misalignment with storage I/O granularity, and high in-memory indexing overhead, leading to significant I/O latency and ultimately limiting scalability for large-scale vector search. In this paper, we propose PageANN, a disk-based approximate nearest neighbor search (ANNS) framework designed for high performance and scalability. PageANN introduces a page-node graph structure that aligns logical graph nodes with physical SSD pages, thereby shortening I/O traversal paths and reducing I/O operations. Specifically, similar vectors are clustered into page nodes, and a co-designed disk data layout leverages this structure with a merging technique to store only representative vectors and topology information, avoiding unnecessary reads. To further improve efficiency, we design a memory management strategy that combines lightweight indexing with coordinated memory-disk data allocation, maximizing host memory utilization while minimizing query latency and storage overhead. Experimental results show that PageANN significantly outperforms state-of-the-art (SOTA) disk-based ANNS methods, achieving 1.85x-10.83x higher throughput and 51.7%-91.9% lower latency across different datasets and memory budgets, while maintaining comparable high recall accuracy.

Layer-wise dynamic rank for compressing large language models

arXiv:2509.25622v2 Announce Type: replace Abstract: Large language models (LLMs) have rapidly scaled in size, bringing severe memory and computational challenges that hinder their deployment. Singular Value Decomposition (SVD)-based compression has emerged as an appealing post-training compression technique for LLMs, yet most existing methods apply a uniform compression ratio across all layers, implicitly assuming homogeneous information included in various layers. This overlooks the substantial intra-layer heterogeneity observed in LLMs, where middle layers tend to encode richer information while early and late layers are more redundant. In this work, we revisit the existing SVD-based compression method and propose D-Rank, a framework with layer-wise balanced Dynamic Rank allocation for LLMs compression. We first introduce effective rank as a principled metric to measure the information density of weight matrices, and then allocate ranks via a Lagrange multiplier-based optimization scheme to adaptively assign more capacity to groups with higher information density under a fixed compression ratio. Moreover, we rebalance the allocated ranks across attention layers to account for their varying importance and extend D-Rank to latest LLMs with grouped-query attention. Extensive experiments on various LLMs with different scales across multiple compression ratios demonstrate that D-Rank consistently outperforms SVD-LLM, ASVD, and Basis Sharing, achieving more than 15 lower perplexity with LLaMA-3-8B model on C4 datasets at 20% compression ratio and up to 5% higher zero-shot reasoning accuracy with LLaMA-7B model at 40% compression ratio while achieving even higher throughput.

Autonomy-Aware Clustering: When Local Decisions Supersede Global Prescriptions

arXiv:2509.25775v2 Announce Type: replace Abstract: Clustering arises in a wide range of problem formulations, yet most existing approaches assume that the entities under clustering are passive and strictly conform to their assigned groups. In reality, entities often exhibit local autonomy, overriding prescribed associations in ways not fully captured by feature representations. Such autonomy can substantially reshape clustering outcomes -- altering cluster compositions, geometry, and cardinality -- with significant downstream effects on inference and decision-making. We introduce autonomy-aware clustering, a reinforcement learning (RL) framework that learns and accounts for the influence of local autonomy without requiring prior knowledge of its form. Our approach integrates RL with a Deterministic Annealing (DA) procedure, where, to determine underlying clusters, DA naturally promotes exploration in early stages of annealing and transitions to exploitation later. We also show that the annealing procedure exhibits phase transitions that enable design of efficient annealing schedules. To further enhance adaptability, we propose the Adaptive Distance Estimation Network (ADEN), a transformer-based attention model that learns dependencies between entities and cluster representatives within the RL loop, accommodates variable-sized inputs and outputs, and enables knowledge transfer across diverse problem instances. Empirical results show that our framework closely aligns with underlying data dynamics: even without explicit autonomy models, it achieves solutions close to the ground truth (gap ~3-4%), whereas ignoring autonomy leads to substantially larger gaps (~35-40%). The code and data are publicly available at https://github.com/salar96/AutonomyAwareClustering.

Learning to Reason as Action Abstractions with Scalable Mid-Training RL

arXiv:2509.25810v2 Announce Type: replace Abstract: Large language models excel with reinforcement learning (RL), but fully unlocking this potential requires a mid-training stage. An effective mid-training phase should identify a compact set of useful actions and enable fast selection among them through online RL. We formalize this intuition by presenting the first theoretical result on how mid-training shapes post-training: it characterizes an action subspace that minimizes both the value approximation error from pruning and the RL error during subsequent planning. Our analysis reveals two key determinants of mid-training effectiveness: pruning efficiency, which shapes the prior of the initial RL policy, and its impact on RL convergence, which governs the extent to which that policy can be improved via online interactions. These results suggest that mid-training is most effective when the decision space is compact and the effective horizon is short, highlighting the importance of operating in the space of action abstractions rather than primitive actions. Building on these insights, we propose Reasoning as Action Abstractions (RA3), a scalable mid-training algorithm. Specifically, we derive a sequential variational lower bound and optimize it by iteratively discovering temporally-consistent latent structures via RL, followed by fine-tuning on the bootstrapped data. Experiments on code generation tasks demonstrate the effectiveness of our approach. Across multiple base models, RA3 improves the average performance on HumanEval and MBPP by 8 and 4 points over the base model and the next-token prediction baseline. Furthermore, RA3 achieves faster convergence and higher asymptotic performance in RLVR on HumanEval+, MBPP+, LiveCodeBench, and Codeforces.

Muon Outperforms Adam in Tail-End Associative Memory Learning

arXiv:2509.26030v2 Announce Type: replace Abstract: The Muon optimizer is consistently faster than Adam in training Large Language Models (LLMs), yet the mechanism underlying its success remains unclear. This paper demystifies this mechanism through the lens of associative memory. By ablating the transformer components optimized by Muon, we reveal that the associative memory parameters of LLMs, namely the Value and Output (VO) attention weights and Feed-Forward Networks (FFNs), are the primary contributors to Muon's superiority. Motivated by this associative memory view, we then explain Muon's superiority on real-world corpora, which are intrinsically heavy-tailed: a few classes (tail classes) appear far less frequently than others. The superiority is explained through two key properties: (i) its update rule consistently yields a more isotropic singular spectrum than Adam; and as a result, (ii) on heavy-tailed data, it optimizes tail classes more effectively than Adam. Beyond empirical evidence, we theoretically confirm these findings by analyzing a one-layer associative memory model under class-imbalanced data. We prove that Muon consistently achieves balanced learning across classes regardless of feature embeddings, whereas Adam can induce large disparities in learning errors depending on embedding properties. In summary, our empirical observations and theoretical analyses reveal Muon's core advantage: its update rule aligns with the outer-product structure of linear associative memories, enabling more balanced and effective learning of tail classes in heavy-tailed distributions than Adam.

Learning Passive Continuous-Time Dynamics with Multistep Port-Hamiltonian Gaussian Processes

arXiv:2510.00384v2 Announce Type: replace Abstract: We propose the multistep port-Hamiltonian Gaussian process (MS-PHS GP) to learn physically consistent continuous-time dynamics and a posterior over the Hamiltonian from noisy, irregularly-sampled trajectories. By placing a GP prior on the Hamiltonian surface $H$ and encoding variable-step multistep integrator constraints as finite linear functionals, MS-PHS GP enables closed-form conditioning of both the vector field and the Hamiltonian surface without latent states, while enforcing energy balance and passivity by design. We state a finite-sample vector-field bound that separates the estimation and variable-step discretization terms. Lastly, we demonstrate improved vector-field recovery and well-calibrated Hamiltonian uncertainty on mass-spring, Van der Pol, and Duffing benchmarks.

MG2FlowNet: Accelerating High-Reward Sample Generation via Enhanced MCTS and Greediness Control

arXiv:2510.00805v2 Announce Type: replace Abstract: Generative Flow Networks (GFlowNets) have emerged as a powerful tool for generating diverse and high-reward structured objects by learning to sample from a distribution proportional to a given reward function. Unlike conventional reinforcement learning (RL) approaches that prioritize optimization of a single trajectory, GFlowNets seek to balance diversity and reward by modeling the entire trajectory distribution. This capability makes them especially suitable for domains such as molecular design and combinatorial optimization. However, existing GFlowNets sampling strategies tend to overexplore and struggle to consistently generate high-reward samples, particularly in large search spaces with sparse high-reward regions. Therefore, improving the probability of generating high-reward samples without sacrificing diversity remains a key challenge under this premise. In this work, we integrate an enhanced Monte Carlo Tree Search (MCTS) into the GFlowNets sampling process, using MCTS-based policy evaluation to guide the generation toward high-reward trajectories and Polynomial Upper Confidence Trees (PUCT) to balance exploration and exploitation adaptively, and we introduce a controllable mechanism to regulate the degree of greediness. Our method enhances exploitation without sacrificing diversity by dynamically balancing exploration and reward-driven guidance. The experimental results show that our method can not only accelerate the speed of discovering high-reward regions but also continuously generate high-reward samples, while preserving the diversity of the generative distribution. All implementations are available at https://github.com/ZRNB/MG2FlowNet.

PEL-NAS: Search Space Partitioned Architecture Prompt Co-Evolutionary LLM-driven Hardware-Aware Neural Architecture Search

arXiv:2510.01472v2 Announce Type: replace Abstract: Hardware-Aware Neural Architecture Search (HW-NAS) requires joint optimization of accuracy and latency under device constraints. Traditional supernet-based methods require multiple GPU days per dataset. Large Language Model (LLM)-driven approaches avoid training a large supernet and can provide quick feedback, but we observe an exploration bias: the LLM repeatedly proposes neural network designs within limited search space and fails to discover architectures across different latency ranges in the entire search space. To address this issue, we propose PEL-NAS: a search space Partitioned, architecture prompt co-Evolutionary and LLM-driven Neural Architecture Search that can generate neural networks with high accuracy and low latency with reduced search cost. Our proposed PEL-NAS has three key components: 1) a complexity-driven partitioning engine that divides the search space by complexity to enforce diversity and mitigate exploration bias; 2) an LLM-powered architecture prompt co-evolution operator, in which the LLM first updates a knowledge base of design heuristics based on results from the previous round, then performs a guided evolution algorithm on architectures with prompts that incorporate this knowledge base. Prompts and designs improve together across rounds which avoids random guesswork and improve efficiency; 3) a zero-cost predictor to avoid training a large number of candidates from scratch. Experimental results show that on HW-NAS-Bench, PEL-NAS can achieve overall higher HV, lower IGD, and up to 54% lower latency than baselines at similar accuracy. Meanwhile, the search cost drops from days to minutes compared with traditional supernet baselines.

Rethinking KL Regularization in RLHF: From Value Estimation to Gradient Optimization

arXiv:2510.01555v2 Announce Type: replace Abstract: Reinforcement Learning from Human Feedback (RLHF) leverages a Kullback-Leibler (KL) divergence loss to stabilize training and prevent overfitting. However, in methods such as GRPO, its implementation may be guided by principles from numerical value estimation-a practice that overlooks the term's functional role as an optimization loss. To analyze this issue, we establish a unified framework that connects two seemingly distinct implementation styles: using the mathematical term $k_n$ as a detached coefficient for the policy's score function ('$k_n$ in reward') or as a direct loss function through which gradients are propagated ('$k_n$ as loss'). We show that the latter can always be analyzed via an equivalent gradient coefficient in the former, unifying the two perspectives. Through this framework, we prove that the conventional '$k_1$ in reward' (like in PPO) is the principled loss for Reverse KL (RKL) regularization. We further establish a key finding: under on-policy conditions, the '$k_2$ as loss' formulation is, in fact, gradient-equivalent to '$k_1$ in reward'. This equivalence, first proven in our work, identifies both as the theoretically sound implementations of the RKL objective. In contrast, we show that the recently adopted '$k_3$ as loss' (like in GRPO) is merely a first-order, biased approximation of the principled loss. Furthermore, we argue that common off-policy implementations of '$k_n$ as loss' methods are biased due to neglected importance sampling, and we propose a principled correction. Our findings provide a comprehensive, gradient-based rationale for choosing and correctly implementing KL regularization, paving the way for more robust and effective RLHF systems.

Learning Representations Through Contrastive Neural Model Checking

arXiv:2510.01853v2 Announce Type: replace Abstract: Model checking is a key technique for verifying safety-critical systems against formal specifications, where recent applications of deep learning have shown promise. However, while ubiquitous for vision and language domains, representation learning remains underexplored in formal verification. We introduce Contrastive Neural Model Checking (CNML), a novel method that leverages the model checking task as a guiding signal for learning aligned representations. CNML jointly embeds logical specifications and systems into a shared latent space through a self-supervised contrastive objective. On industry-inspired retrieval tasks, CNML considerably outperforms both algorithmic and neural baselines in cross-modal and intra-modal settings. We further show that the learned representations effectively transfer to downstream tasks and generalize to more complex formulas. These findings demonstrate that model checking can serve as an objective for learning representations for formal languages.

PENEX: AdaBoost-Inspired Neural Network Regularization

arXiv:2510.02107v2 Announce Type: replace Abstract: AdaBoost sequentially fits so-called weak learners to minimize an exponential loss, which penalizes mislabeled data points more severely than other loss functions like cross-entropy. Paradoxically, AdaBoost generalizes well in practice as the number of weak learners grows. In the present work, we introduce Penalized Exponential Loss (PENEX), a new formulation of the multi-class exponential loss that is theoretically grounded and, in contrast to the existing formulation, amenable to optimization via first-order methods. We demonstrate both empirically and theoretically that PENEX implicitly maximizes margins of data points. Also, we show that gradient increments on PENEX implicitly parameterize weak learners in the boosting framework. Across computer vision and language tasks, we show that PENEX exhibits a regularizing effect often better than established methods with similar computational cost. Our results highlight PENEX's potential as an AdaBoost-inspired alternative for effective training and fine-tuning of deep neural networks.

Diffusion^2: Turning 3D Environments into Radio Frequency Heatmaps

arXiv:2510.02274v2 Announce Type: replace Abstract: Modeling radio frequency (RF) signal propagation is essential for understanding the environment, as RF signals offer valuable insights beyond the capabilities of RGB cameras, which are limited by the visible-light spectrum, lens coverage, and occlusions. It is also useful for supporting wireless diagnosis, deployment, and optimization. However, accurately predicting RF signals in complex environments remains a challenge due to interactions with obstacles such as absorption and reflection. We introduce Diffusion^2, a diffusion-based approach that uses 3D point clouds to model the propagation of RF signals across a wide range of frequencies, from Wi-Fi to millimeter waves. To effectively capture RF-related features from 3D data, we present the RF-3D Encoder, which encapsulates the complexities of 3D geometry along with signal-specific details. These features undergo multi-scale embedding to simulate the actual RF signal dissemination process. Our evaluation, based on synthetic and real-world measurements, demonstrates that Diffusion^2 accurately estimates the behavior of RF signals in various frequency bands and environmental conditions, with an error margin of just 1.9 dB and 27x faster than existing methods, marking a significant advancement in the field. Refer to https://rfvision-project.github.io/ for more information.

MINERVA: Mutual Information Neural Estimation for Supervised Feature Selection

arXiv:2510.02610v2 Announce Type: replace Abstract: Existing feature filters rely on statistical pair-wise dependence metrics to model feature-target relationships, but this approach may fail when the target depends on higher-order feature interactions rather than individual contributions. We introduce Mutual Information Neural Estimation Regularized Vetting Algorithm (MINERVA), a novel approach to supervised feature selection based on neural estimation of mutual information between features and targets. We paramaterize the approximation of mutual information with neural networks and perform feature selection using a carefully designed loss function augmented with sparsity-inducing regularizers. Our method is implemented in a two-stage process to decouple representation learning from feature selection, ensuring better generalization and a more accurate expression of feature importance. We present examples of ubiquitous dependency structures that are rarely captured in literature and show that our proposed method effectively captures these complex feature-target relationships by evaluating feature subsets as an ensemble. Experimental results on synthetic and real-life fraud datasets demonstrate the efficacy of our method and its ability to perform exact solutions.

Relevance-Aware Thresholding in Online Conformal Prediction for Time Series

arXiv:2510.02809v2 Announce Type: replace Abstract: Uncertainty quantification has received considerable interest in recent works in Machine Learning. In particular, Conformal Prediction (CP) gains ground in this field. For the case of time series, Online Conformal Prediction (OCP) becomes an option to address the problem of data distribution shift over time. Indeed, the idea of OCP is to update a threshold of some quantity (whether the miscoverage level or the quantile) based on the distribution observation. To evaluate the performance of OCP methods, two key aspects are typically considered: the coverage validity and the prediction interval width minimization. Recently, new OCP methods have emerged, offering long-run coverage guarantees and producing more informative intervals. However, during the threshold update step, most of these methods focus solely on the validity of the prediction intervals~--~that is, whether the ground truth falls inside or outside the interval~--~without accounting for their relevance. In this paper, we aim to leverage this overlooked aspect. Specifically, we propose enhancing the threshold update step by replacing the binary evaluation (inside/outside) with a broader class of functions that quantify the relevance of the prediction interval using the ground truth. This approach helps prevent abrupt threshold changes, potentially resulting in narrower prediction intervals. Indeed, experimental results on real-world datasets suggest that these functions can produce tighter intervals compared to existing OCP methods while maintaining coverage validity.

Distributional Inverse Reinforcement Learning

arXiv:2510.03013v2 Announce Type: replace Abstract: We propose a distributional framework for offline Inverse Reinforcement Learning (IRL) that jointly models uncertainty over reward functions and full distributions of returns. Unlike conventional IRL approaches that recover a deterministic reward estimate or match only expected returns, our method captures richer structure in expert behavior, particularly in learning the reward distribution, by minimizing first-order stochastic dominance (FSD) violations and thus integrating distortion risk measures (DRMs) into policy learning, enabling the recovery of both reward distributions and distribution-aware policies. This formulation is well-suited for behavior analysis and risk-aware imitation learning. Empirical results on synthetic benchmarks, real-world neurobehavioral data, and MuJoCo control tasks demonstrate that our method recovers expressive reward representations and achieves state-of-the-art imitation performance.

Mixtures of Gaussian Process Experts with SMC$^2$

arXiv:2208.12830v3 Announce Type: replace-cross Abstract: Gaussian processes are a key component of many flexible statistical and machine learning models. However, they exhibit cubic computational complexity and high memory constraints due to the need of inverting and storing a full covariance matrix. To circumvent this, mixtures of Gaussian process experts have been considered where data points are assigned to independent experts, reducing the complexity by allowing inference based on smaller, local covariance matrices. Moreover, mixtures of Gaussian process experts substantially enrich the model's flexibility, allowing for behaviors such as non-stationarity, heteroscedasticity, and discontinuities. In this work, we construct a novel inference approach based on nested sequential Monte Carlo samplers to simultaneously infer both the gating network and Gaussian process expert parameters. This greatly improves inference compared to importance sampling, particularly in settings when a stationary Gaussian process is inappropriate, while still being thoroughly parallelizable.

Quick Adaptive Ternary Segmentation: An Efficient Decoding Procedure For Hidden Markov Models

arXiv:2305.18578v2 Announce Type: replace-cross Abstract: Hidden Markov models (HMMs) are characterized by an unobservable Markov chain and an observable process -- a noisy version of the hidden chain. Decoding the original signal from the noisy observations is one of the main goals in nearly all HMM based data analyses. Existing decoding algorithms such as Viterbi and the pointwise maximum a posteriori (PMAP) algorithm have computational complexity at best linear in the length of the observed sequence, and sub-quadratic in the size of the state space of the hidden chain. We present Quick Adaptive Ternary Segmentation (QATS), a divide-and-conquer procedure with computational complexity polylogarithmic in the length of the sequence, and cubic in the size of the state space, hence particularly suited for large scale HMMs with relatively few states. It also suggests an effective way of data storage as specific cumulative sums. In essence, the estimated sequence of states sequentially maximizes local likelihood scores among all local paths with at most three segments, and is meanwhile admissible. The maximization is performed only approximately using an adaptive search procedure. Our simulations demonstrate the speedups offered by QATS in comparison to Viterbi and PMAP, along with a precision analysis. An implementation of QATS is in the R-package QATS on GitHub.

Fitted value iteration methods for bicausal optimal transport

arXiv:2306.12658v3 Announce Type: replace-cross Abstract: We develop a fitted value iteration (FVI) method to compute bicausal optimal transport (OT) where couplings have an adapted structure. Based on the dynamic programming formulation, FVI adopts a function class to approximate the value functions in bicausal OT. Under the concentrability condition and approximate completeness assumption, we prove the sample complexity using (local) Rademacher complexity. Furthermore, we demonstrate that multilayer neural networks with appropriate structures satisfy the crucial assumptions required in sample complexity proofs. Numerical experiments reveal that FVI outperforms linear programming and adapted Sinkhorn methods in scalability as the time horizon increases, while still maintaining acceptable accuracy.

AgentBench: Evaluating LLMs as Agents

arXiv:2308.03688v3 Announce Type: replace-cross Abstract: The potential of Large Language Model (LLM) as agents has been widely acknowledged recently. Thus, there is an urgent need to quantitatively \textit{evaluate LLMs as agents} on challenging tasks in interactive environments. We present AgentBench, a multi-dimensional benchmark that consists of 8 distinct environments to assess LLM-as-Agent's reasoning and decision-making abilities. Our extensive test over \num API-based and open-sourced (OSS) LLMs shows that, while top commercial LLMs present a strong ability of acting as agents in complex environments, there is a significant disparity in performance between them and many OSS competitors that are no larger than 70B. We identify the typical reasons of failures in environments and LLMs, showing that poor long-term reasoning, decision-making, and instruction following abilities are the main obstacles for developing usable LLM agents. Improving instruction following and training on high quality multi-round alignment data could improve agent performance. And different from existing assumptions, training on code present ambivalent impacts on different agent tasks. Datasets, environments, and an integrated evaluation package for AgentBench are released at https://github.com/THUDM/AgentBench.

Robust MRI Reconstruction by Smoothed Unrolling (SMUG)

arXiv:2312.07784v3 Announce Type: replace-cross Abstract: As the popularity of deep learning (DL) in the field of magnetic resonance imaging (MRI) continues to rise, recent research has indicated that DL-based MRI reconstruction models might be excessively sensitive to minor input disturbances, including worst-case additive perturbations. This sensitivity often leads to unstable, aliased images. This raises the question of how to devise DL techniques for MRI reconstruction that can be robust to train-test variations. To address this problem, we propose a novel image reconstruction framework, termed Smoothed Unrolling (SMUG), which advances a deep unrolling-based MRI reconstruction model using a randomized smoothing (RS)-based robust learning approach. RS, which improves the tolerance of a model against input noises, has been widely used in the design of adversarial defense approaches for image classification tasks. Yet, we find that the conventional design that applies RS to the entire DL-based MRI model is ineffective. In this paper, we show that SMUG and its variants address the above issue by customizing the RS process based on the unrolling architecture of a DL-based MRI reconstruction model. Compared to the vanilla RS approach, we show that SMUG improves the robustness of MRI reconstruction with respect to a diverse set of instability sources, including worst-case and random noise perturbations to input measurements, varying measurement sampling rates, and different numbers of unrolling steps. Furthermore, we theoretically analyze the robustness of our method in the presence of perturbations.

Impact of Dataset Properties on Membership Inference Vulnerability of Deep Transfer Learning

arXiv:2402.06674v5 Announce Type: replace-cross Abstract: Membership inference attacks (MIAs) are used to test practical privacy of machine learning models. MIAs complement formal guarantees from differential privacy (DP) under a more realistic adversary model. We analyse MIA vulnerability of fine-tuned neural networks both empirically and theoretically, the latter using a simplified model of fine-tuning. We show that the vulnerability of non-DP models when measured as the attacker advantage at a fixed false positive rate reduces according to a simple power law as the number of examples per class increases. A similar power-law applies even for the most vulnerable points, but the dataset size needed for adequate protection of the most vulnerable points is very large.

RealKIE: Five Novel Datasets for Enterprise Key Information Extraction

arXiv:2403.20101v2 Announce Type: replace-cross Abstract: We introduce RealKIE, a benchmark of five challenging datasets aimed at advancing key information extraction methods, with an emphasis on enterprise applications. The datasets include a diverse range of documents including SEC S1 Filings, US Non-disclosure Agreements, UK Charity Reports, FCC Invoices, and Resource Contracts. Each presents unique challenges: poor text serialization, sparse annotations in long documents, and complex tabular layouts. These datasets provide a realistic testing ground for key information extraction tasks like investment analysis and contract analysis. In addition to presenting these datasets, we offer an in-depth description of the annotation process, document processing techniques, and baseline modeling approaches. This contribution facilitates the development of NLP models capable of handling practical challenges and supports further research into information extraction technologies applicable to industry-specific problems. The annotated data, OCR outputs, and code to reproduce baselines are available to download at https://indicodatasolutions.github.io/RealKIE/.

Data-Driven Performance Guarantees for Classical and Learned Optimizers

arXiv:2404.13831v3 Announce Type: replace-cross Abstract: We introduce a data-driven approach to analyze the performance of continuous optimization algorithms using generalization guarantees from statistical learning theory. We study classical and learned optimizers to solve families of parametric optimization problems. We build generalization guarantees for classical optimizers, using a sample convergence bound, and for learned optimizers, using the Probably Approximately Correct (PAC)-Bayes framework. To train learned optimizers, we use a gradient-based algorithm to directly minimize the PAC-Bayes upper bound. Numerical experiments in signal processing, control, and meta-learning showcase the ability of our framework to provide strong generalization guarantees for both classical and learned optimizers given a fixed budget of iterations. For classical optimizers, our bounds are much tighter than those that worst-case guarantees provide. For learned optimizers, our bounds outperform the empirical outcomes observed in their non-learned counterparts.

Learning Safe Strategies for Value Maximizing Buyers in Uniform Price Auctions

arXiv:2406.03674v4 Announce Type: replace-cross Abstract: We study the bidding problem in repeated uniform price multi-unit auctions from the perspective of a value-maximizing buyer. The buyer aims to maximize their cumulative value over $T$ rounds while adhering to per-round return-on-investment (RoI) constraints in a strategic (or adversarial) environment. Using an $m$-uniform bidding format, the buyer submits $m$ bid-quantity pairs $(b_i, q_i)$ to demand $q_i$ units at bid $b_i$, with $m \ll M$ in practice, where $M$ denotes the maximum demand of the buyer. We introduce the notion of safe bidding strategies as those that satisfy the RoI constraints irrespective of competing bids. Despite the stringent requirement, we show that these strategies satisfy a mild no-overbidding condition, depend only on the valuation curve of the bidder, and the bidder can focus on a finite subset without loss of generality. Though the subset size is $O(M^m)$, we design a polynomial-time learning algorithm that achieves sublinear regret, both in full-information and bandit settings, relative to the hindsight-optimal safe strategy. We assess the robustness of safe strategies against the hindsight-optimal strategy from a richer class. We define the richness ratio $\alpha \in (0,1]$ as the minimum ratio of the value of the optimal safe strategy to that of the optimal strategy from richer class and construct hard instances showing the tightness of $\alpha$. Our algorithm achieves $\alpha$-approximate sublinear regret against these stronger benchmarks. Simulations on semi-synthetic auction data show that empirical richness ratios significantly outperform the theoretical worst-case bounds. The proposed safe strategies and learning algorithm extend naturally to more nuanced buyer and competitor models.

CHARME: A chain-based reinforcement learning approach for the minor embedding problem

arXiv:2406.07124v2 Announce Type: replace-cross Abstract: Quantum annealing (QA) has great potential to solve combinatorial optimization problems efficiently. However, the effectiveness of QA algorithms is heavily based on the embedding of problem instances, represented as logical graphs, into the quantum processing unit (QPU) whose topology is in the form of a limited connectivity graph, known as the minor embedding problem. Because the minor embedding problem is an NP-hard problem~\mbox{\cite{Goodrich2018}}, existing methods for the minor embedding problem suffer from scalability issues when faced with larger problem sizes. In this paper, we propose a novel approach utilizing Reinforcement Learning (RL) techniques to address the minor embedding problem, named CHARME. CHARME includes three key components: a Graph Neural Network (GNN) architecture for policy modeling, a state transition algorithm that ensures solution validity, and an order exploration strategy for effective training. Through comprehensive experiments on synthetic and real-world instances, we demonstrate the efficiency of our proposed order exploration strategy as well as our proposed RL framework, CHARME. In particular, CHARME yields superior solutions in terms of qubit usage compared to fast embedding methods such as Minorminer and ATOM. Moreover, our method surpasses the OCT-based approach, known for its slower runtime but high-quality solutions, in several cases. In addition, our proposed exploration enhances the efficiency of the training of the CHARME framework by providing better solutions compared to the greedy strategy.

A Deterministic Information Bottleneck Method for Clustering Mixed-Type Data

arXiv:2407.03389v4 Announce Type: replace-cross Abstract: In this paper, we present an information-theoretic method for clustering mixed-type data, that is, data consisting of both continuous and categorical variables. The proposed approach extends the Information Bottleneck principle to heterogeneous data through generalised product kernels, integrating continuous, nominal, and ordinal variables within a unified optimization framework. We address the following challenges: developing a systematic bandwidth selection strategy that equalises contributions across variable types, and proposing an adaptive hyperparameter updating scheme that ensures a valid solution into a predetermined number of potentially imbalanced clusters. Through simulations on 28,800 synthetic data sets and ten publicly available benchmarks, we demonstrate that the proposed method, named DIBmix, achieves superior performance compared to four established methods (KAMILA, K-Prototypes, FAMD with K-Means, and PAM with Gower's dissimilarity). Results show DIBmix particularly excels when clusters exhibit size imbalances, data contain low or moderate cluster overlap, and categorical and continuous variables are equally represented. The method presents a significant advantage over traditional centroid-based algorithms, establishing DIBmix as a competitive and theoretically grounded alternative for mixed-type data clustering.

Sharp Generalization for Nonparametric Regression in Interpolation Space by Over-Parameterized Neural Networks Trained with Preconditioned Gradient Descent and Early Stopping

arXiv:2407.11353v4 Announce Type: replace-cross Abstract: We study nonparametric regression using an over-parameterized two-layer neural networks trained with algorithmic guarantees in this paper. We consider the setting where the training features are drawn uniformly from the unit sphere in $\RR^d$, and the target function lies in an interpolation space commonly studied in statistical learning theory. We demonstrate that training the neural network with a novel Preconditioned Gradient Descent (PGD) algorithm, equipped with early stopping, achieves a sharp regression rate of $\cO(n^{-\frac{2\alpha s'}{2\alpha s'+1}})$ when the target function is in the interpolation space $\bth{\cH_K}^{s'}$ with $s' \ge 3$. This rate is even sharper than the currently known nearly-optimal rate of $\cO(n^{-\frac{2\alpha s'}{2\alpha s'+1}})\log^2(1/\delta)$~\citep{Li2024-edr-general-domain}, where $n$ is the size of the training data and $\delta \in (0,1)$ is a small probability. This rate is also sharper than the standard kernel regression rate of $\cO(n^{-\frac{2\alpha}{2\alpha+1}})$ obtained under the regular Neural Tangent Kernel (NTK) regime when training the neural network with the vanilla gradient descent (GD), where $2\alpha = d/(d-1)$. Our analysis is based on two key technical contributions. First, we present a principled decomposition of the network output at each PGD step into a function in the reproducing kernel Hilbert space (RKHS) of a newly induced integral kernel, and a residual function with small $L^{\infty}$-norm. Second, leveraging this decomposition, we apply local Rademacher complexity theory to tightly control the complexity of the function class comprising all the neural network functions obtained in the PGD iterates. Our results further suggest that PGD enables the neural network to escape the linear NTK regime and achieve improved generalization.

ClinicRealm: Re-evaluating Large Language Models with Conventional Machine Learning for Non-Generative Clinical Prediction Tasks

arXiv:2407.18525v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly deployed in medicine. However, their utility in non-generative clinical prediction, often presumed inferior to specialized models, remains under-evaluated, leading to ongoing debate within the field and potential for misuse, misunderstanding, or over-reliance due to a lack of systematic benchmarking. Our ClinicRealm study addresses this by benchmarking 15 GPT-style LLMs, 5 BERT-style models, and 11 traditional methods on unstructured clinical notes and structured Electronic Health Records (EHR), while also assessing their reasoning, reliability, and fairness. Key findings reveal a significant shift: for clinical note predictions, leading LLMs (e.g., DeepSeek-V3.1-Think, GPT-5) in zero-shot settings now decisively outperform finetuned BERT models. On structured EHRs, while specialized models excel with ample data, advanced LLMs (e.g., GPT-5, DeepSeek-V3.1-Think) show potent zero-shot capabilities, often surpassing conventional models in data-scarce settings. Notably, leading open-source LLMs can match or exceed proprietary counterparts. These results provide compelling evidence that modern LLMs are competitive tools for non-generative clinical prediction, particularly with unstructured text and offering data-efficient structured data options, thus necessitating a re-evaluation of model selection strategies. This research should serve as an important insight for medical informaticists, AI developers, and clinical researchers, potentially prompting a reassessment of current assumptions and inspiring new approaches to LLM application in predictive healthcare.

Universality of Kernel Random Matrices and Kernel Regression in the Quadratic Regime

arXiv:2408.01062v2 Announce Type: replace-cross Abstract: Kernel ridge regression (KRR) is a popular class of machine learning models that has become an important tool for understanding deep learning. Much of the focus thus far has been on studying the proportional asymptotic regime, $n \asymp d$, where $n$ is the number of training samples and $d$ is the dimension of the dataset. In the proportional regime, under certain conditions on the data distribution, the kernel random matrix involved in KRR exhibits behavior akin to that of a linear kernel. In this work, we extend the study of kernel regression to the quadratic asymptotic regime, where $n \asymp d^2$. In this regime, we demonstrate that a broad class of inner-product kernels exhibits behavior similar to a quadratic kernel. Specifically, we establish an operator norm approximation bound for the difference between the original kernel random matrix and a quadratic kernel random matrix with additional correction terms compared to the Taylor expansion of the kernel functions. The approximation works for general data distributions under a Gaussian-moment-matching assumption with a covariance structure. This new approximation is utilized to obtain a limiting spectral distribution of the original kernel matrix and characterize the precise asymptotic training and test errors for KRR in the quadratic regime when $n/d^2$ converges to a non-zero constant. The generalization errors are obtained for (i) a random teacher model, (ii) a deterministic teacher model where the weights are perfectly aligned with the covariance of the data. Under the random teacher model setting, we also verify that the generalized cross-validation (GCV) estimator can consistently estimate the generalization error in the quadratic regime for anisotropic data. Our proof techniques combine moment methods, Wick's formula, orthogonal polynomials, and resolvent analysis of random matrices with correlated entries.

On-Demand Growth of Semiconductor Heterostructures Guided by Physics-Informed Machine Learning

arXiv:2408.03508v4 Announce Type: replace-cross Abstract: Developing tailored semiconductor heterostructures on demand represents a critical capability for addressing the escalating performance demands in electronic and optoelectronic devices. However, traditional fabrication methods remain constrained by simulation-based design and iterative trial-and-error optimization. Here, we introduce SemiEpi, a self-driving platform designed for molecular beam epitaxy (MBE) to perform multi-step semiconductor heterostructure growth through in-situ monitoring and on-the-fly feedback control. By integrating standard MBE reactors, physics-informed machine learning (ML) models, and parameter initialization, SemiEpi identifies optimal initial conditions and proposes experiments for heterostructure growth, eliminating the need for extensive expertise in MBE processes. As a proof of concept, we demonstrate the optimization of high-density InAs quantum dot (QD) growth with a target emission wavelength of 1240 nm, showcasing the power of SemiEpi. We achieve a QD density of 5 x 10^10 cm^-2, a 1.6-fold increase in photoluminescence (PL) intensity, and a reduced full width at half maximum (FWHM) of 29.13 meV, leveraging in-situ reflective high-energy electron diffraction monitoring with feedback control for adjusting growth temperatures. Taken together, our results highlight the potential of ML-guided systems to address challenges in multi-step heterostructure growth, facilitate the development of a hardware-independent framework, and enhance process repeatability and stability, even without exhaustive knowledge of growth parameters.

Exploring Applications of State Space Models and Advanced Training Techniques in Sequential Recommendations: A Comparative Study on Efficiency and Performance

arXiv:2408.05606v2 Announce Type: replace-cross Abstract: Recommender systems aim to estimate the dynamically changing user preferences and sequential dependencies between historical user behaviour and metadata. Although transformer-based models have proven to be effective in sequential recommendations, their state growth is proportional to the length of the sequence that is being processed, which makes them expensive in terms of memory and inference costs. Our research focused on three promising directions in sequential recommendations: enhancing speed through the use of State Space Models (SSM), as they can achieve SOTA results in the sequential recommendations domain with lower latency, memory, and inference costs, as proposed by arXiv:2403.03900 improving the quality of recommendations with Large Language Models (LLMs) via Monolithic Preference Optimization without Reference Model (ORPO); and implementing adaptive batch- and step-size algorithms to reduce costs and accelerate training processes.

Approximation Bounds for Recurrent Neural Networks with Application to Regression

arXiv:2409.05577v2 Announce Type: replace-cross Abstract: We study the approximation capacity of deep ReLU recurrent neural networks (RNNs) and explore the convergence properties of nonparametric least squares regression using RNNs. We derive upper bounds on the approximation error of RNNs for H\"older smooth functions, in the sense that the output at each time step of an RNN can approximate a H\"older function that depends only on past and current information, termed a past-dependent function. This allows a carefully constructed RNN to simultaneously approximate a sequence of past-dependent H\"older functions. We apply these approximation results to derive non-asymptotic upper bounds for the prediction error of the empirical risk minimizer in regression problem. Our error bounds achieve minimax optimal rate under both exponentially $\beta$-mixing and i.i.d. data assumptions, improving upon existing ones. Our results provide statistical guarantees on the performance of RNNs.

Conformal Fields from Neural Networks

arXiv:2409.12222v2 Announce Type: replace-cross Abstract: We use the embedding formalism to construct conformal fields in $D$ dimensions, by restricting Lorentz-invariant ensembles of homogeneous neural networks in $(D+2)$ dimensions to the projective null cone. Conformal correlators may be computed using the parameter space description of the neural network. Exact four-point correlators are computed in a number of examples, and we perform a 4D conformal block decomposition that elucidates the spectrum. In some examples the analysis is facilitated by recent approaches to Feynman integrals. Generalized free CFTs are constructed using the infinite-width Gaussian process limit of the neural network, enabling a realization of the free boson. The extension to deep networks constructs conformal fields at each subsequent layer, with recursion relations relating their conformal dimensions and four-point functions. Numerical approaches are discussed.

Machine Learning for Inverse Problems and Data Assimilation

arXiv:2410.10523v2 Announce Type: replace-cross Abstract: The aim of these notes is to demonstrate the potential for ideas in machine learning to impact on the fields of inverse problems and data assimilation. The perspective is one that is primarily aimed at researchers from inverse problems and/or data assimilation who wish to see a mathematical presentation of machine learning as it pertains to their fields. As a by-product, we include a succinct mathematical treatment of various fundamental underpinning topics in machine learning, and adjacent areas of (computational) mathematics.

Fast constrained sampling in pre-trained diffusion models

arXiv:2410.18804v3 Announce Type: replace-cross Abstract: Large denoising diffusion models, such as Stable Diffusion, have been trained on billions of image-caption pairs to perform text-conditioned image generation. As a byproduct of this training, these models have acquired general knowledge about image statistics, which can be useful for other inference tasks. However, when confronted with sampling an image under new constraints, e.g. generating the missing parts of an image, using large pre-trained text-to-image diffusion models is inefficient and often unreliable. Previous approaches either utilized backpropagation through the denoiser network, making them significantly slower and more memory-demanding than simple text-to-image generation, or only enforced the constraint locally, failing to capture critical long-range correlations in the sampled image. In this work, we propose an algorithm that enables fast, high-quality generation under arbitrary constraints. We show that in denoising diffusion models, we can employ an approximation to Newton's optimization method that allows us to speed up inference and avoid the expensive backpropagation operations. Our approach produces results that rival or surpass the state-of-the-art training-free inference methods while requiring a fraction of the time. We demonstrate the effectiveness of our algorithm under both linear (inpainting, super-resolution) and non-linear (style-guided generation) constraints. An implementation is provided at https://github.com/cvlab-stonybrook/fast-constrained-sampling.

Probabilistic Language-Image Pre-Training

arXiv:2410.18857v4 Announce Type: replace-cross Abstract: Vision-language models (VLMs) embed aligned image-text pairs into a joint space but often rely on deterministic embeddings, assuming a one-to-one correspondence between images and texts. This oversimplifies real-world relationships, which are inherently many-to-many, with multiple captions describing a single image and vice versa. We introduce Probabilistic Language-Image Pre-training (ProLIP), the first probabilistic VLM pre-trained on a billion-scale image-text dataset using only probabilistic objectives, achieving a strong zero-shot capability (e.g., 74.6% ImageNet zero-shot accuracy with ViT-B/16). ProLIP efficiently estimates uncertainty by an "uncertainty token" without extra parameters. We also introduce a novel inclusion loss that enforces distributional inclusion relationships between image-text pairs and between original and masked inputs. Experiments demonstrate that, by leveraging uncertainty estimates, ProLIP benefits downstream tasks and aligns with intuitive notions of uncertainty, e.g., shorter texts being more uncertain and more general inputs including specific ones. Utilizing text uncertainties, we further improve ImageNet accuracy from 74.6% to 75.8% (under a few-shot setting), supporting the practical advantages of our probabilistic approach. The code is available at https://github.com/naver-ai/prolip

Multi-Turn Human-LLM Interaction Through the Lens of a Two-Way Intelligibility Protocol

arXiv:2410.20600v3 Announce Type: replace-cross Abstract: Our interest is in the design of software systems involving a human-expert interacting -- using natural language -- with a large language model (LLM) on data analysis tasks. For complex problems, it is possible that LLMs can harness human expertise and creativity to find solutions that were otherwise elusive. On one level, this interaction takes place through multiple turns of prompts from the human and responses from the LLM. Here we investigate a more structured approach based on an abstract protocol described in [3] for interaction between agents. The protocol is motivated by a notion of "two-way intelligibility" and is modelled by a pair of communicating finite-state machines. We provide an implementation of the protocol, and provide empirical evidence of using the implementation to mediate interactions between an LLM and a human-agent in two areas of scientific interest (radiology and drug design). We conduct controlled experiments with a human proxy (a database), and uncontrolled experiments with human subjects. The results provide evidence in support of the protocol's capability of capturing one- and two-way intelligibility in human-LLM interaction; and for the utility of two-way intelligibility in the design of human-machine systems. Our code is available at https://github.com/karannb/interact.

Don't Pay Attention, PLANT It: Pretraining Attention via Learning-to-Rank

arXiv:2410.23066v2 Announce Type: replace-cross Abstract: State-of-the-art Extreme Multi-Label Text Classification models rely on multi-label attention to focus on key tokens in input text, but learning good attention weights is challenging. We introduce PLANT - Pretrained and Leveraged Attention - a plug-and-play strategy for initializing attention. PLANT works by planting label-specific attention using a pretrained Learning-to-Rank model guided by mutual information gain. This architecture-agnostic approach integrates seamlessly with large language model backbones such as Mistral-7B, LLaMA3-8B, DeepSeek-V3, and Phi-3. PLANT outperforms state-of-the-art methods across tasks including ICD coding, legal topic classification, and content recommendation. Gains are especially pronounced in few-shot settings, with substantial improvements on rare labels. Ablation studies confirm that attention initialization is a key driver of these gains. For code and trained models, see https://github.com/debjyotiSRoy/xcube/tree/plant

Counterfactual explainability and analysis of variance

arXiv:2411.01625v2 Announce Type: replace-cross Abstract: Existing tools for explaining complex models and systems are associational rather than causal and do not provide mechanistic understanding. We propose a new notion called counterfactual explainability for causal attribution that is motivated by the concept of genetic heritability in twin studies. Counterfactual explainability extends methods for global sensitivity analysis (including the functional analysis of variance and Sobol's indices), which assumes independent explanatory variables, to dependent explanations by using a directed acyclic graphs to describe their causal relationship. Therefore, this explanability measure directly incorporates causal mechanisms by construction. Under a comonotonicity assumption, we discuss methods for estimating counterfactual explainability and apply them to a real dataset dataset to explain income inequality by gender, race, and educational attainment.

SEE-DPO: Self Entropy Enhanced Direct Preference Optimization

arXiv:2411.04712v2 Announce Type: replace-cross Abstract: Direct Preference Optimization (DPO) has been successfully used to align large language models (LLMs) according to human preferences, and more recently it has also been applied to improving the quality of text-to-image diffusion models. However, DPO-based methods such as SPO, Diffusion-DPO, and D3PO are highly susceptible to overfitting and reward hacking, especially when the generative model is optimized to fit out-of-distribution during prolonged training. To overcome these challenges and stabilize the training of diffusion models, we introduce a self-entropy regularization mechanism in reinforcement learning from human feedback. This enhancement improves DPO training by encouraging broader exploration and greater robustness. Our regularization technique effectively mitigates reward hacking, leading to improved stability and enhanced image quality across the latent space. Extensive experiments demonstrate that integrating human feedback with self-entropy regularization can significantly boost image diversity and specificity, achieving state-of-the-art results on key image generation metrics.

H3Fusion: Helpful, Harmless, Honest Fusion of Aligned LLMs

arXiv:2411.17792v2 Announce Type: replace-cross Abstract: Alignment of pretrained LLMs using instruction-based datasets is critical for creating fine-tuned models that reflect human preference. A growing number of alignment-based fine-tuning algorithms and benchmarks emerged recently, fueling the efforts on effective alignments of pre-trained LLMs to ensure helpful, harmless, and honest answers from both open-source and closed-source LLMs. This paper tackles this problem by developing an alignment fusion approach, coined as $H^3$Fusion, with three unique characteristics. First, $H^3$Fusion ensembles multiple individually aligned LLMs to create a final fine-tuned alignment model with enhanced capabilities beyond those of individual models, delivering robust alignment through promoting helpful, harmless, honest fusion. Second, $H^3$Fusion leverages the mixture-of-experts (MoE) methodology in two steps. We first freeze the multi-head attention weights of each individual model while tuning the FFN layer during alignment fusion. Then we merge the aligned model weights with an expert router according to the type of input instruction and dynamically select a subset of experts that are best suited for producing the output response. Finally, we boost the performance of the resulting $H^3$3Fusion model by introducing gating loss and regularization terms. The former penalizes the selection errors of the expert-router, and the latter mediates the expert weights drifting during fine-tuning and dynamically adjusts the fusion behavior of the resulting model by canalizing the activations on the experts. Extensive evaluations on three benchmark datasets show that $H^3$3Fusion is more helpful, less harmful, and more honest from two aspects: it outperforms each individually aligned model by $11.37\%$, and it provides stronger robustness compared to the state-of-the-art LLM ensemble approaches by $13.77\%$. Code is available at github.com/sftekin/h3fusion.

Another look at inference after prediction

arXiv:2411.19908v5 Announce Type: replace-cross Abstract: From structural biology to epidemiology, predictions from machine learning (ML) models increasingly complement costly gold-standard data to enable faster, more affordable, and scalable scientific inquiry. In response, prediction-based (PB) inference has emerged to accommodate statistical analysis using a large volume of predictions together with a small amount of gold-standard data. The goals of PB inference are two-fold: (i) to mitigate bias from errors in predictions and (ii) to improve efficiency relative to classical inference using only the gold-standard data. While early PB inference methods focused on bias, their ability to enhance efficiency remains a focus of ongoing research. We revisit a foundational PB inference method and show that a simple modification can be applied to guarantee provable improvements in efficiency. In doing so, we establish new connections between augmented inverse probability weighted estimators (AIPW) and several recently proposed PB inference methods with a similar focus. The utility of our proposal, which leverages prediction-based outcomes to enhance efficiency, is demonstrated through extensive simulation studies and an application to real data from the UK Biobank. Further, we contextualize PB inference by drawing connections to historical literature from economics and statistics, highlighting how classic methods directly inform this contemporary problem.

STIV: Scalable Text and Image Conditioned Video Generation

arXiv:2412.07730v2 Announce Type: replace-cross Abstract: The field of video generation has made remarkable advancements, yet there remains a pressing need for a clear, systematic recipe that can guide the development of robust and scalable models. In this work, we present a comprehensive study that systematically explores the interplay of model architectures, training recipes, and data curation strategies, culminating in a simple and scalable text-image-conditioned video generation method, named STIV. Our framework integrates image condition into a Diffusion Transformer (DiT) through frame replacement, while incorporating text conditioning via a joint image-text conditional classifier-free guidance. This design enables STIV to perform both text-to-video (T2V) and text-image-to-video (TI2V) tasks simultaneously. Additionally, STIV can be easily extended to various applications, such as video prediction, frame interpolation, multi-view generation, and long video generation, etc. With comprehensive ablation studies on T2I, T2V, and TI2V, STIV demonstrate strong performance, despite its simple design. An 8.7B model with 512 resolution achieves 83.1 on VBench T2V, surpassing both leading open and closed-source models like CogVideoX-5B, Pika, Kling, and Gen-3. The same-sized model also achieves a state-of-the-art result of 90.1 on VBench I2V task at 512 resolution. By providing a transparent and extensible recipe for building cutting-edge video generation models, we aim to empower future research and accelerate progress toward more versatile and reliable video generation solutions.

A Statistical Hypothesis Testing Framework for Data Misappropriation Detection in Large Language Models

arXiv:2501.02441v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are rapidly gaining enormous popularity in recent years. However, the training of LLMs has raised significant privacy and legal concerns, particularly regarding the distillation and inclusion of copyrighted materials in their training data without proper attribution or licensing, an issue that falls under the broader concern of data misappropriation. In this article, we focus on a specific problem of data misappropriation detection, namely, to determine whether a given LLM has incorporated the data generated by another LLM. We propose embedding watermarks into the copyrighted training data and formulating the detection of data misappropriation as a hypothesis testing problem. We develop a general statistical testing framework, construct test statistics, determine optimal rejection thresholds, and explicitly control type I and type II errors. Furthermore, we establish the asymptotic optimality properties of the proposed tests, and demonstrate the empirical effectiveness through intensive numerical experiments.

Neural Deconstruction Search for Vehicle Routing Problems

arXiv:2501.03715v2 Announce Type: replace-cross Abstract: Autoregressive construction approaches generate solutions to vehicle routing problems in a step-by-step fashion, leading to high-quality solutions that are nearing the performance achieved by handcrafted operations research techniques. In this work, we challenge the conventional paradigm of sequential solution construction and introduce an iterative search framework where solutions are instead deconstructed by a neural policy. Throughout the search, the neural policy collaborates with a simple greedy insertion algorithm to rebuild the deconstructed solutions. Our approach matches or surpasses the performance of state-of-the-art operations research methods across three challenging vehicle routing problems of various problem sizes.

Unlocking In-Context Learning for Natural Datasets Beyond Language Modelling

arXiv:2501.06256v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) exhibit In-Context Learning (ICL), which enables the model to perform new tasks conditioning only on the examples provided in the context without updating the model's weights. While ICL offers fast adaptation across natural language tasks and domains, its emergence is less straightforward for modalities beyond text. In this work, we systematically uncover properties present in LLMs that support the emergence of ICL for autoregressive models and various modalities by promoting the learning of the needed mechanisms for ICL. We identify exact token repetitions in the training data sequences as an important factor for ICL. Such repetitions further improve stability and reduce transiency in ICL performance. Moreover, we emphasise the significance of training task difficulty for the emergence of ICL. Finally, by applying our novel insights on ICL emergence, we unlock ICL capabilities for various visual datasets and a more challenging EEG classification task.

Evolutionary Optimization of Physics-Informed Neural Networks: Evo-PINN Frontiers and Opportunities

arXiv:2501.06572v4 Announce Type: replace-cross Abstract: Deep learning models trained on finite data lack a complete understanding of the physical world. On the other hand, physics-informed neural networks (PINNs) are infused with such knowledge through the incorporation of mathematically expressible laws of nature into their training loss function. By complying with physical laws, PINNs provide advantages over purely data-driven models in limited-data regimes and present as a promising route towards Physical AI. This feature has propelled them to the forefront of scientific machine learning, a domain characterized by scarce and costly data. However, the vision of accurate physics-informed learning comes with significant challenges. This work examines PINNs in terms of model optimization and generalization, shedding light on the need for new algorithmic advances to overcome issues pertaining to the training speed, precision, and generalizability of today's PINN models. Of particular interest are gradient-free evolutionary algorithms (EAs) for optimizing the uniquely complex loss landscapes arising in PINN training. Methods synergizing gradient descent and EAs for discovering bespoke neural architectures and balancing multiple terms in physics-informed learning objectives are positioned as important avenues for future research. Another exciting track is to cast EAs as a meta-learner of generalizable PINN models. To substantiate these proposed avenues, we further highlight results from recent literature to showcase the early success of such approaches in addressing the aforementioned challenges in PINN optimization and generalization.

Characterizing Mobile SoC for Accelerating Heterogeneous LLM Inference

arXiv:2501.14794v2 Announce Type: replace-cross Abstract: With the rapid advancement of artificial intelligence technologies such as ChatGPT, AI agents, and video generation, contemporary mobile systems have begun integrating these AI capabilities on local devices to enhance privacy and reduce response latency. To meet the computational demands of AI tasks, current mobile SoCs are equipped with diverse AI accelerators, including GPUs and Neural Processing Units (NPUs). However, there has not been a comprehensive characterization of these heterogeneous processors, and existing designs typically only leverage a single AI accelerator for LLM inference, leading to suboptimal use of computational resources and memory bandwidth. In this paper, we first summarize key performance characteristics of heterogeneous processors, SoC memory bandwidth, etc. Drawing on these observations, we propose different heterogeneous parallel mechanisms to fully exploit both GPU and NPU computational power and memory bandwidth. We further design a fast synchronization mechanism between heterogeneous processors that leverages the unified memory architecture. By employing these techniques, we present HeteroInfer, the fastest LLM inference engine in mobile devices which supports GPU-NPU heterogeneous execution. Evaluation shows that HeteroInfer delivers a 1.34x to 6.02x end-to-end speedup over state-of-the-art GPU-only and NPU-only LLM engines, while maintaining negligible interference with other applications.

Controllable Video Generation with Provable Disentanglement

arXiv:2502.02690v3 Announce Type: replace-cross Abstract: Controllable video generation remains a significant challenge, despite recent advances in generating high-quality and consistent videos. Most existing methods for controlling video generation treat the video as a whole, neglecting intricate fine-grained spatiotemporal relationships, which limits both control precision and efficiency. In this paper, we propose Controllable Video Generative Adversarial Networks (CoVoGAN) to disentangle the video concepts, thus facilitating efficient and independent control over individual concepts. Specifically, following the minimal change principle, we first disentangle static and dynamic latent variables. We then leverage the sufficient change property to achieve component-wise identifiability of dynamic latent variables, enabling disentangled control of video generation. To establish the theoretical foundation, we provide a rigorous analysis demonstrating the identifiability of our approach. Building on these theoretical insights, we design a Temporal Transition Module to disentangle latent dynamics. To enforce the minimal change principle and sufficient change property, we minimize the dimensionality of latent dynamic variables and impose temporal conditional independence. To validate our approach, we integrate this module as a plug-in for GANs. Extensive qualitative and quantitative experiments on various video generation benchmarks demonstrate that our method significantly improves generation quality and controllability across diverse real-world scenarios.

Summaries as Centroids for Interpretable and Scalable Text Clustering

arXiv:2502.09667v3 Announce Type: replace-cross Abstract: We introduce k-NLPmeans and k-LLMmeans, text-clustering variants of k-means that periodically replace numeric centroids with textual summaries. The key idea, summary-as-centroid, retains k-means assignments in embedding space while producing human-readable, auditable cluster prototypes. The method is LLM-optional: k-NLPmeans uses lightweight, deterministic summarizers, enabling offline, low-cost, and stable operation; k-LLMmeans is a drop-in upgrade that uses an LLM for summaries under a fixed per-iteration budget whose cost does not grow with dataset size. We also present a mini-batch extension for real-time clustering of streaming text. Across diverse datasets, embedding models, and summarization strategies, our approach consistently outperforms classical baselines and approaches the accuracy of recent LLM-based clustering-without extensive LLM calls. Finally, we provide a case study on sequential text streams and release a StackExchange-derived benchmark for evaluating streaming text clustering.

DiffSampling: Enhancing Diversity and Accuracy in Neural Text Generation

arXiv:2502.14037v4 Announce Type: replace-cross Abstract: Despite their growing capabilities, language models still frequently reproduce content from their training data, generate repetitive text, and favor common grammatical patterns and vocabulary. A possible cause is the decoding strategy: the most common strategies either consider only the most probable tokens, which reduces output diversity, or increase the likelihood of unlikely tokens, compromising output accuracy and correctness. In this paper, we propose DiffSampling, a new decoding method that leverages a mathematical analysis of the token probability distribution to ensure the generation of contextually appropriate text. In particular, the difference between consecutive, sorted probabilities can be used to truncate incorrect tokens. In addition, we also propose two variations of the proposed method that aim to correct the subtle inconsistencies of common sampling strategies. Experiments involving four different text-generation tasks demonstrate that our approach consistently performs at least on par with the existing methods it builds upon in terms of quality, while potentially improving output diversity.

Sample Complexity of Linear Quadratic Regulator Without Initial Stability

arXiv:2502.14210v3 Announce Type: replace-cross Abstract: Inspired by REINFORCE, we introduce a novel receding-horizon algorithm for the Linear Quadratic Regulator (LQR) problem with unknown dynamics. Unlike prior methods, our algorithm avoids reliance on two-point gradient estimates while maintaining the same order of sample complexity. Furthermore, it eliminates the restrictive requirement of starting with a stable initial policy, broadening its applicability. Beyond these improvements, we introduce a refined analysis of error propagation through the contraction of the Riccati operator under the Riemannian distance. This refinement leads to a better sample complexity and ensures improved convergence guarantees.

On Pruning State-Space LLMs

arXiv:2502.18886v2 Announce Type: replace-cross Abstract: Recent work proposed state-space models (SSMs) as an efficient alternative to transformer-based LLMs. Can these models be pruned to further reduce their computation costs? We adapt several pruning methods to the SSM structure, and apply them to four SSM-based LLMs across multiple tasks. We find that such models are quite robust to some pruning methods (e.g. WANDA), while using other methods lead to fast performance degradation.

Vector Copula Variational Inference and Dependent Block Posterior Approximations

arXiv:2503.01072v2 Announce Type: replace-cross Abstract: The key to VI is the selection of a tractable density to approximate the Bayesian posterior. For large and complex models a common choice is to assume independence between multivariate blocks in a partition of the parameter space. While this simplifies the problem it can reduce accuracy. This paper proposes using vector copulas to capture dependence between the blocks parsimoniously. Tailored multivariate marginals are constructed using learnable transport maps. We call the resulting joint distribution a ``dependent block posterior'' approximation. Vector copula models are suggested that make tractable and flexible variational approximations. They allow for differing marginals, numbers of blocks, block sizes and forms of between block dependence. They also allow for solution of the variational optimization using efficient stochastic gradient methods. The approach is demonstrated using four different statistical models and 16 datasets which have posteriors that are challenging to approximate. This includes models that use global-local shrinkage priors for regularization, and hierarchical models for smoothing and heteroscedastic time series. In all cases, our method produces more accurate posterior approximations than benchmark VI methods that either assume block independence or factor-based dependence, at limited additional computational cost. A python package implementing the method is available on GitHub at https://github.com/YuFuOliver/VCVI_Rep_PyPackage.

Evolutionary Guided Decoding: Iterative Value Refinement for LLMs

arXiv:2503.02368v3 Announce Type: replace-cross Abstract: While guided decoding, especially value-guided methods, has emerged as a cost-effective alternative for controlling language model outputs without re-training models, its effectiveness is limited by the accuracy of the value function. We identify that this inaccuracy stems from a core distributional gap: existing methods train static value functions on trajectories sampled exclusively from the base policy, which inherently confines their training to a narrow and suboptimal view of the potential output space. We propose Iterative Value Refinement, a novel framework designed to bridge this gap. It employs Value Exploration to provide a more comprehensive and robust training signal, complemented by Iterative Self-Refinement, which uses the improved value function from one iteration to guide the generation of higher-quality data for the next. Extensive experiments on text summarization, multi-turn dialogue, and instruction following demonstrate the effectiveness of our framework in aligning language models. Our approach not only achieves alignment but also significantly reduces computational costs by leveraging principled value function optimization for efficient and effective control.

Seeded Poisson Factorization: leveraging domain knowledge to fit topic models

arXiv:2503.02741v2 Announce Type: replace-cross Abstract: Topic models are widely used for discovering latent thematic structures in large text corpora, yet traditional unsupervised methods often struggle to align with pre-defined conceptual domains. This paper introduces seeded Poisson Factorization (SPF), a novel approach that extends the Poisson Factorization (PF) framework by incorporating domain knowledge through seed words. SPF enables a structured topic discovery by modifying the prior distribution of topic-specific term intensities, assigning higher initial rates to pre-defined seed words. The model is estimated using variational inference with stochastic gradient optimization, ensuring scalability to large datasets. We present in detail the results of applying SPF to an Amazon customer feedback dataset, leveraging pre-defined product categories as guiding structures. SPF achieves superior performance compared to alternative guided probabilistic topic models in terms of computational efficiency and classification performance. Robustness checks highlight SPF's ability to adaptively balance domain knowledge and data-driven topic discovery, even in case of imperfect seed word selection. Further applications of SPF to four additional benchmark datasets, where the corpus varies in size and the number of topics differs, demonstrate its general superior classification performance compared to the unseeded PF model.

Graph Alignment via Birkhoff Relaxation

arXiv:2503.05323v2 Announce Type: replace-cross Abstract: We consider the graph alignment problem, wherein the objective is to find a vertex correspondence between two graphs that maximizes the edge overlap. The graph alignment problem is an instance of the quadratic assignment problem (QAP), known to be NP-hard in the worst case even to approximately solve. In this paper, we analyze Birkhoff relaxation, a tight convex relaxation of QAP, and present theoretical guarantees on its performance when the inputs follow the Gaussian Wigner Model. More specifically, the weighted adjacency matrices are correlated Gaussian Orthogonal Ensemble with correlation $1/\sqrt{1+\sigma^2}$. Denote the optimal solutions of the QAP and Birkhoff relaxation by $\Pi^\star$ and $X^\star$ respectively. We show that $|X^\star-\Pi^\star|_F^2 = o(n)$ when $\sigma = o(n^{-1.25})$ and $|X^\star-\Pi^\star|_F^2 = \Omega(n)$ when $\sigma = \Omega(n^{-0.5})$. Thus, the optimal solution $X^\star$ transitions from a small perturbation of $\Pi^\star$ for small $\sigma$ to being well separated from $\Pi^\star$ as $\sigma$ becomes larger than $n^{-0.5}$. This result allows us to guarantee that simple rounding procedures on $X^\star$ align $1-o(1)$ fraction of vertices correctly whenever $\sigma = o(n^{-1.25})$. This condition on $\sigma$ to ensure the success of the Birkhoff relaxation is state-of-the-art.

PRO-VPT: Distribution-Adaptive Visual Prompt Tuning via Prompt Relocation

arXiv:2503.06901v2 Announce Type: replace-cross Abstract: Visual prompt tuning (VPT), i.e., fine-tuning some lightweight prompt tokens, provides an efficient and effective approach for adapting pre-trained models to various downstream tasks. However, most prior art indiscriminately uses a fixed prompt distribution across different tasks, neglecting the importance of each block varying depending on the task. In this paper, we introduce adaptive distribution optimization (ADO) by tackling two key questions: (1) How to appropriately and formally define ADO, and (2) How to design an adaptive distribution strategy guided by this definition? Through empirical analysis, we first confirm that properly adjusting the distribution significantly improves VPT performance, and further uncover a key insight that a nested relationship exists between ADO and VPT. Based on these findings, we propose a new VPT framework, termed PRO-VPT (iterative Prompt RelOcation-based VPT), which adaptively adjusts the distribution built upon a nested optimization formulation. Specifically, we develop a prompt relocation strategy derived from this formulation, comprising two steps: pruning idle prompts from prompt-saturated blocks, followed by allocating these prompts to the most prompt-needed blocks. By iteratively performing prompt relocation and VPT, our proposal can adaptively learn the optimal prompt distribution in a nested optimization-based manner, thereby unlocking the full potential of VPT. Extensive experiments demonstrate that our proposal significantly outperforms advanced VPT methods, e.g., PRO-VPT surpasses VPT by 1.6 pp and 2.0 pp average accuracy, leading prompt-based methods to state-of-the-art performance on VTAB-1k and FGVC benchmarks. The code is available at https://github.com/ckshang/PRO-VPT.

FEB-Cache: Frequency-Guided Exposure Bias Reduction for Enhancing Diffusion Transformer Caching

arXiv:2503.07120v3 Announce Type: replace-cross Abstract: Diffusion Transformer (DiT) has exhibited impressive generation capabilities but faces great challenges due to its high computational complexity. To address this issue, various methods, notably feature caching, have been introduced. However, these approaches focus on aligning non-cache diffusion without analyzing why caching damage the generation processes. In this paper, we first confirm that the cache greatly amplifies the exposure bias, resulting in a decline in the generation quality. However, directly applying noise scaling is challenging for this issue due to the non-smoothness of exposure bias. We found that this phenomenon stems from the mismatch between its frequency response characteristics and the simple cache of Attention and MLP. Since these two components exhibit unique preferences for frequency signals, which provides us with a caching strategy to separate Attention and MLP to achieve an enhanced fit of exposure bias and reduce it. Based on this, we introduced FEB-Cache, a joint caching strategy that aligns with the non-exposed bias diffusion process (which gives us a higher performance cap) of caching Attention and MLP based on the frequency-guided cache table. Our approach combines a comprehensive understanding of the caching mechanism and offers a new perspective on leveraging caching to accelerate the diffusion process. Empirical results indicate that FEB-Cache optimizes model performance while concurrently facilitating acceleration. Code is available at https://github.com/aSleepyTree/EB-Cache.

Learning to Play Piano in the Real World

arXiv:2503.15481v2 Announce Type: replace-cross Abstract: Towards the grand challenge of achieving human-level manipulation in robots, playing piano is a compelling testbed that requires strategic, precise, and flowing movements. Over the years, several works demonstrated hand-designed controllers on real world piano playing, while other works evaluated robot learning approaches on simulated piano scenarios. In this paper, we develop the first piano playing robotic system that makes use of learning approaches while also being deployed on a real world dexterous robot. Specifically, we make use of Sim2Real to train a policy in simulation using reinforcement learning before deploying the learned policy on a real world dexterous robot. In our experiments, we thoroughly evaluate the interplay between domain randomization and the accuracy of the dynamics model used in simulation. Moreover, we evaluate the robot's performance across multiple songs with varying complexity to study the generalization of our learned policy. By providing a proof-of-concept of learning to play piano in the real world, we want to encourage the community to adopt piano playing as a compelling benchmark towards human-level manipulation. We open-source our code and show additional videos at https://lasr.org/research/learning-to-play-piano .

Benchmark Dataset for Pore-Scale CO2-Water Interaction

arXiv:2503.17592v2 Announce Type: replace-cross Abstract: Accurately capturing the complex interaction between CO2 and water in porous media at the pore scale is essential for various geoscience applications, including carbon capture and storage (CCS). We introduce a comprehensive dataset generated from high-fidelity numerical simulations to capture the intricate interaction between CO2 and water at the pore scale. The dataset consists of 624 2D samples, each of size 512x512 with a resolution of 35 {\mu}m, covering 100 time steps under a constant CO2 injection rate. It includes various levels of heterogeneity, represented by different grain sizes with random variation in spacing, offering a robust testbed for developing predictive models. This dataset provides high-resolution temporal and spatial information crucial for benchmarking machine learning models.

RAG over Tables: Hierarchical Memory Index, Multi-Stage Retrieval, and Benchmarking

arXiv:2504.01346v4 Announce Type: replace-cross Abstract: Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by integrating them with an external knowledge base to improve the answer relevance and accuracy. In real-world scenarios, beyond pure text, a substantial amount of knowledge is stored in tables, and user questions often require retrieving answers that are distributed across multiple tables. Retrieving knowledge from a table corpora (i.e., various individual tables) for a question remains nascent, at least, for (i) how to understand intra- and inter-table knowledge effectively, (ii) how to filter unnecessary tables and how to retrieve the most relevant tables efficiently, (iii) how to prompt LLMs to infer over the retrieval, (iv) how to evaluate the corresponding performance in a realistic setting. Facing the above challenges, in this paper, we first propose a table-corpora-aware RAG framework, named T-RAG, which consists of the hierarchical memory index, multi-stage retrieval, and graph-aware prompting for effective and efficient table knowledge retrieval and inference. Further, we first develop a multi-table question answering benchmark named MultiTableQA, which spans 3 different task types, 57,193 tables, and 23,758 questions in total, and the sources are all from real-world scenarios. Based on MultiTableQA, we did the holistic comparison over table retrieval methods, RAG methods, and table-to-graph representation learning methods, where T-RAG shows the leading accuracy, recall, and running time performance. Also, under T-RAG, we evaluate the inference ability upgrade of different LLMs. Code and Data are available at https://github.com/jiaruzouu/T-RAG

MAD: A Magnitude And Direction Policy Parametrization for Stability Constrained Reinforcement Learning

arXiv:2504.02565v2 Announce Type: replace-cross Abstract: We introduce magnitude and direction (MAD) policies, a policy parameterization for reinforcement learning (RL) that preserves Lp closed-loop stability for nonlinear dynamical systems. Despite their completeness in describing all stabilizing controllers, methods based on nonlinear Youla and system-level synthesis are significantly impacted by the difficulty of parametrizing Lp-stable operators. In contrast, MAD policies introduce explicit feedback on state-dependent features - a key element behind the success of reinforcement learning pipelines - without jeopardizing closed-loop stability. This is achieved by letting the magnitude of the control input be described by a disturbance-feedback Lp-stable operator, while selecting its direction based on state-dependent features through a universal function approximator. We further characterize the robust stability properties of MAD policies under model mismatch. Unlike existing disturbance-feedback policy parametrizations, MAD policies introduce state-feedback components compatible with model-free RL pipelines, ensuring closed-loop stability with no model information beyond assuming open-loop stability. Numerical experiments show that MAD policies trained with deep deterministic policy gradient (DDPG) methods generalize to unseen scenarios - matching the performance of standard neural network policies while guaranteeing closed-loop stability by design.

On the Effectiveness and Generalization of Race Representations for Debiasing High-Stakes Decisions

arXiv:2504.06303v2 Announce Type: replace-cross Abstract: Understanding and mitigating biases is critical for the adoption of large language models (LLMs) in high-stakes decision-making. We introduce Admissions and Hiring, decision tasks with hypothetical applicant profiles where a person's race can be inferred from their name, as simplified test beds for racial bias. We show that Gemma 2B Instruct and LLaMA 3.2 3B Instruct exhibit strong biases. Gemma grants admission to 26% more White than Black applicants, and LLaMA hires 60% more Asian than White applicants. We demonstrate that these biases are resistant to prompt engineering: multiple prompting strategies all fail to promote fairness. In contrast, using distributed alignment search, we can identify "race subspaces" within model activations and intervene on them to debias model decisions. Averaging the representation across all races within the subspaces reduces Gemma's bias by 37-57%. Finally, we examine the generalizability of Gemma's race subspaces, and find limited evidence for generalization, where changing the prompt format can affect the race representation. Our work suggests mechanistic approaches may provide a promising venue for improving the fairness of LLMs, but a universal race representation remains elusive.

Conformalized Generative Bayesian Imaging: An Uncertainty Quantification Framework for Computational Imaging

arXiv:2504.07696v2 Announce Type: replace-cross Abstract: Uncertainty quantification plays an important role in achieving trustworthy and reliable learning-based computational imaging. Recent advances in generative modeling and Bayesian neural networks have enabled the development of uncertainty-aware image reconstruction methods. Current generative model-based methods seek to quantify the inherent (aleatoric) uncertainty on the underlying image for given measurements by learning to sample from the posterior distribution of the underlying image. On the other hand, Bayesian neural network-based approaches aim to quantify the model (epistemic) uncertainty on the parameters of a deep neural network-based reconstruction method by approximating the posterior distribution of those parameters. Unfortunately, an ongoing need for an inversion method that can jointly quantify complex aleatoric uncertainty and epistemic uncertainty patterns still persists. In this paper, we present a scalable framework that can quantify both aleatoric and epistemic uncertainties. The proposed framework accepts an existing generative model-based posterior sampling method as an input and introduces an epistemic uncertainty quantification capability through Bayesian neural networks with latent variables and deep ensembling. Furthermore, by leveraging the conformal prediction methodology, the proposed framework can be easily calibrated to ensure rigorous uncertainty quantification. We evaluated the proposed framework on magnetic resonance imaging, computed tomography, and image inpainting problems and showed that the epistemic and aleatoric uncertainty estimates produced by the proposed framework display the characteristic features of true epistemic and aleatoric uncertainties. Furthermore, our results demonstrated that the use of conformal prediction on top of the proposed framework enables marginal coverage guarantees consistent with frequentist principles.

J1: Incentivizing Thinking in LLM-as-a-Judge via Reinforcement Learning

arXiv:2505.10320v2 Announce Type: replace-cross Abstract: The progress of AI is bottlenecked by the quality of evaluation, making powerful LLM-as-a-Judge models a core solution. The efficacy of these judges depends on their chain-of-thought reasoning, creating a critical need for methods that can effectively optimize this reasoning process. In this work, we introduce J1, a reinforcement learning framework for teaching LLM judges to think before making decisions. Our core contribution lies in converting all judgment tasks for non-verifiable and verifiable prompts into a unified format with verifiable rewards, enabling direct optimization of evaluation quality while mitigating positional bias. We then use RL to train thinking-judges at scales of 8B, 32B, and 70B and show that they obtain state-of-the-art performance across multiple benchmarks. In particular, J1-Qwen-32B, our multitasked pointwise and pairwise judge also outperforms o1-mini, o3, and a much larger 671B DeepSeek-R1 on some benchmarks, while only training on synthetic data. Through comprehensive ablations of pairwise, pointwise, and multitask J1 variants, we demonstrate the effectiveness of our approach across seed prompts, reward strategies, and training recipes. Qualitative analysis reveals that J1 develops systematic evaluation strategies, including dynamic criteria generation, reference answer creation, iterative self-correction of initial assessments, and feedback generation for low-quality responses.

A Hybrid Strategy for Probabilistic Forecasting and Trading of Aggregated Wind-Solar Power: Design and Analysis in HEFTCom2024

arXiv:2505.10367v2 Announce Type: replace-cross Abstract: Obtaining accurate probabilistic energy forecasts and making effective decisions amid diverse uncertainties are routine challenges in future energy systems. This paper presents the winning solution of team GEB, which ranked 3rd in trading, 4th in forecasting, and 1st among student teams in the IEEE Hybrid Energy Forecasting and Trading Competition 2024 (HEFTCom2024). The solution provides accurate probabilistic forecasts for a wind-solar hybrid system, and achieves substantial trading revenue in the day-ahead electricity market. Key components include: (1) a stacking-based approach combining sister forecasts from various Numerical Weather Predictions (NWPs) to provide wind power forecasts, (2) an online solar post-processing model to address the distribution shift in the online test set caused by increased solar capacity, (3) a probabilistic aggregation method for accurate quantile forecasts of hybrid generation, and (4) a stochastic trading strategy to maximize expected trading revenue considering uncertainties in electricity prices. This paper also explores the potential of end-to-end learning to further enhance the trading revenue by shifting the distribution of forecast errors. Detailed case studies are provided to validate the effectiveness of these proposed methods. Code for all mentioned methods is available for reproduction and further research in both industry and academia.

TranSUN: A Preemptive Paradigm to Eradicate Retransformation Bias Intrinsically from Regression Models in Recommender Systems

arXiv:2505.13881v4 Announce Type: replace-cross Abstract: Regression models are crucial in recommender systems. However, retransformation bias problem has been conspicuously neglected within the community. While many works in other fields have devised effective bias correction methods, all of them are post-hoc cures externally to the model, facing practical challenges when applied to real-world recommender systems. Hence, we propose a preemptive paradigm to eradicate the bias intrinsically from the models via minor model refinement. Specifically, a novel TranSUN method is proposed with a joint bias learning manner to offer theoretically guaranteed unbiasedness under empirical superior convergence. It is further generalized into a novel generic regression model family, termed Generalized TranSUN (GTS), which not only offers more theoretical insights but also serves as a generic framework for flexibly developing various bias-free models. Comprehensive experimental results demonstrate the superiority of our methods across data from various domains, which have been successfully deployed in two real-world industrial recommendation scenarios, i.e. product and short video recommendation scenarios in Guess What You Like business domain in the homepage of Taobao App (a leading e-commerce platform with DAU > 300M), to serve the major online traffic.

Critical Points of Random Neural Networks

arXiv:2505.17000v2 Announce Type: replace-cross Abstract: This work investigates the expected number of critical points of random neural networks with different activation functions as the depth increases in the infinite-width limit. Under suitable regularity conditions, we derive precise asymptotic formulas for the expected number of critical points of fixed index and those exceeding a given threshold. Our analysis reveals three distinct regimes depending on the value of the first derivative of the covariance evaluated at 1: the expected number of critical points may converge, grow polynomially, or grow exponentially with depth. The theoretical predictions are supported by numerical experiments. Moreover, we provide numerical evidence suggesting that, when the regularity condition is not satisfied (e.g. for neural networks with ReLU as activation function), the number of critical points increases as the map resolution increases, indicating a potential divergence in the number of critical points.

A quantitative analysis of semantic information in deep representations of text and images

arXiv:2505.17101v3 Announce Type: replace-cross Abstract: Deep neural networks are known to develop similar representations for semantically related data, even when they belong to different domains, such as an image and its description, or the same text in different languages. We present a method for quantitatively investigating this phenomenon by measuring the relative information content of the representations of semantically related data and probing how it is encoded into multiple tokens of large language models (LLMs) and vision transformers. Looking first at how LLMs process pairs of translated sentences, we identify inner ``semantic'' layers containing the most language-transferable information. We find moreover that, on these layers, a larger LLM (DeepSeek-V3) extracts significantly more general information than a smaller one (Llama3.1-8B). Semantic information of English text is spread across many tokens and it is characterized by long-distance correlations between tokens and by a causal left-to-right (i.e., past-future) asymmetry. We also identify layers encoding semantic information within visual transformers. We show that caption representations in the semantic layers of LLMs predict visual representations of the corresponding images. We observe significant and model-dependent information asymmetries between image and text representations.

From Compression to Expression: A Layerwise Analysis of In-Context Learning

arXiv:2505.17322v2 Announce Type: replace-cross Abstract: In-context learning (ICL) enables large language models (LLMs) to adapt to new tasks without weight updates by learning from demonstration sequences. While ICL shows strong empirical performance, its internal representational mechanisms are not yet well understood. In this work, we conduct a statistical geometric analysis of ICL representations to investigate how task-specific information is captured across layers. Our analysis reveals an intriguing phenomenon, which we term Layerwise Compression-Expression: early layers progressively produce compact and discriminative representations that encode task information from the input demonstrations, while later layers express these representations to incorporate the query and generate the prediction. This phenomenon is observed consistently across diverse tasks and a range of contemporary LLM architectures. We demonstrate that it has important implications for ICL performance -- improving with model size and the number of demonstrations -- and for robustness in the presence of noisy examples. To further understand the effect of the compact task representation, we propose a bias-variance decomposition and provide a theoretical analysis showing how attention mechanisms contribute to reducing both variance and bias, thereby enhancing performance as the number of demonstrations increases. Our findings reveal an intriguing layerwise dynamic in ICL, highlight how structured representations emerge within LLMs, and showcase that analyzing internal representations can facilitate a deeper understanding of model behavior.

AutoMiSeg: Automatic Medical Image Segmentation via Test-Time Adaptation of Foundation Models

arXiv:2505.17931v2 Announce Type: replace-cross Abstract: Medical image segmentation is vital for clinical diagnosis, yet current deep learning methods often demand extensive expert effort, i.e., either through annotating large training datasets or providing prompts at inference time for each new case. This paper introduces a zero-shot and automatic segmentation pipeline that combines off-the-shelf vision-language and segmentation foundation models. Given a medical image and a task definition (e.g., "segment the optic disc in an eye fundus image"), our method uses a grounding model to generate an initial bounding box, followed by a visual prompt boosting module that enhance the prompts, which are then processed by a promptable segmentation model to produce the final mask. To address the challenges of domain gap and result verification, we introduce a test-time adaptation framework featuring a set of learnable adaptors that align the medical inputs with foundation model representations. Its hyperparameters are optimized via Bayesian Optimization, guided by a proxy validation model without requiring ground-truth labels. Our pipeline offers an annotation-efficient and scalable solution for zero-shot medical image segmentation across diverse tasks. Our pipeline is evaluated on seven diverse medical imaging datasets and shows promising results. By proper decomposition and test-time adaptation, our fully automatic pipeline not only substantially surpasses the previously best-performing method, yielding a 69\% relative improvement in accuracy (Dice Score from 42.53 to 71.81), but also performs competitively with weakly-prompted interactive foundation models.

Uniform convergence of the smooth calibration error and its relationship with functional gradient

arXiv:2505.19396v4 Announce Type: replace-cross Abstract: Calibration is a critical requirement for reliable probabilistic prediction, especially in high-risk applications. However, the theoretical understanding of which learning algorithms can simultaneously achieve high accuracy and good calibration remains limited, and many existing studies provide empirical validation or a theoretical guarantee in restrictive settings. To address this issue, in this work, we focus on the smooth calibration error (CE) and provide a uniform convergence bound, showing that the smooth CE is bounded by the sum of the smooth CE over the training dataset and a generalization gap. We further prove that the functional gradient of the loss function can effectively control the training smooth CE. Based on this framework, we analyze three representative algorithms: gradient boosting trees, kernel boosting, and two-layer neural networks. For each, we derive conditions under which both classification and calibration performances are simultaneously guaranteed. Our results offer new theoretical insights and practical guidance for designing reliable probabilistic models with provable calibration guarantees.

Do We Need All the Synthetic Data? Targeted Synthetic Image Augmentation via Diffusion Models

arXiv:2505.21574v2 Announce Type: replace-cross Abstract: Synthetically augmenting training datasets with diffusion models has been an effective strategy for improving generalization of image classifiers. However, existing techniques struggle to ensure the diversity of generation and increase the size of the data by up to 10-30x to improve the in-distribution performance. In this work, we show that synthetically augmenting part of the data that is not learned early in training with faithful images-containing same features but different noise-outperforms augmenting the entire dataset. By analyzing a two-layer CNN, we prove that this strategy improves generalization by promoting homogeneity in feature learning speed without amplifying noise. Our extensive experiments show that by augmenting only 30%-40% of the data, our method boosts generalization by up to 2.8% in a variety of scenarios, including training ResNet, ViT, ConvNeXt, and Swin Transformer on CIFAR-10/100, and TinyImageNet, with various optimizers including SGD and SAM. Notably, our method applied with SGD outperforms the SOTA optimizer, SAM, on CIFAR-100 and TinyImageNet.

A Copula Based Supervised Filter for Feature Selection in Diabetes Risk Prediction Using Machine Learning

arXiv:2505.22554v3 Announce Type: replace-cross Abstract: Effective feature selection is vital for robust and interpretable medical prediction, especially for identifying risk factors concentrated in extreme patient strata. Standard methods emphasize average associations and may miss predictors whose importance lies in the tails of the distribution. We propose a computationally efficient supervised filter that ranks features using the Gumbel copula upper tail dependence coefficient ($\lambda_U$), prioritizing variables that are simultaneously extreme with the positive class. We benchmarked against Mutual Information, mRMR, ReliefF, and $L_1$ Elastic Net across four classifiers on two diabetes datasets: a large public health survey (CDC, N=253,680) and a clinical benchmark (PIMA, N=768). Evaluation included paired statistical tests, permutation importance, and robustness checks with label flips, feature noise, and missingness. On CDC, our method was the fastest selector and reduced the feature space by about 52% while retaining strong discrimination. Although using all 21 features yielded the highest AUC, our filter significantly outperformed Mutual Information and mRMR and was statistically indistinguishable from ReliefF. On PIMA, with only eight predictors, our ranking produced the numerically highest ROC AUC, and no significant differences were found versus strong baselines. Across both datasets, the upper tail criterion consistently identified clinically coherent, impactful predictors. We conclude that copula based feature selection via upper tail dependence is a powerful, efficient, and interpretable approach for building risk models in public health and clinical medicine.

Query Drift Compensation: Enabling Compatibility in Continual Learning of Retrieval Embedding Models

arXiv:2506.00037v2 Announce Type: replace-cross Abstract: Text embedding models enable semantic search, powering several NLP applications like Retrieval Augmented Generation by efficient information retrieval (IR). However, text embedding models are commonly studied in scenarios where the training data is static, thus limiting its applications to dynamic scenarios where new training data emerges over time. IR methods generally encode a huge corpus of documents to low-dimensional embeddings and store them in a database index. During retrieval, a semantic search over the corpus is performed and the document whose embedding is most similar to the query embedding is returned. When updating an embedding model with new training data, using the already indexed corpus is suboptimal due to the non-compatibility issue, since the model which was used to obtain the embeddings of the corpus has changed. While re-indexing of old corpus documents using the updated model enables compatibility, it requires much higher computation and time. Thus, it is critical to study how the already indexed corpus can still be effectively used without the need of re-indexing. In this work, we establish a continual learning benchmark with large-scale datasets and continually train dense retrieval embedding models on query-document pairs from new datasets in each task and observe forgetting on old tasks due to significant drift of embeddings. We employ embedding distillation on both query and document embeddings to maintain stability and propose a novel query drift compensation method during retrieval to project new model query embeddings to the old embedding space. This enables compatibility with previously indexed corpus embeddings extracted using the old model and thus reduces the forgetting. We show that the proposed method significantly improves performance without any re-indexing. Code is available at https://github.com/dipamgoswami/QDC.

UniSim: A Unified Simulator for Time-Coarsened Dynamics of Biomolecules

arXiv:2506.03157v3 Announce Type: replace-cross Abstract: Molecular Dynamics (MD) simulations are essential for understanding the atomic-level behavior of molecular systems, giving insights into their transitions and interactions. However, classical MD techniques are limited by the trade-off between accuracy and efficiency, while recent deep learning-based improvements have mostly focused on single-domain molecules, lacking transferability to unfamiliar molecular systems. Therefore, we propose \textbf{Uni}fied \textbf{Sim}ulator (UniSim), which leverages cross-domain knowledge to enhance the understanding of atomic interactions. First, we employ a multi-head pretraining approach to learn a unified atomic representation model from a large and diverse set of molecular data. Then, based on the stochastic interpolant framework, we learn the state transition patterns over long timesteps from MD trajectories, and introduce a force guidance module for rapidly adapting to different chemical environments. Our experiments demonstrate that UniSim achieves highly competitive performance across small molecules, peptides, and proteins.

RhoDARTS: Differentiable Quantum Architecture Search with Density Matrix Simulations

arXiv:2506.03697v2 Announce Type: replace-cross Abstract: Variational Quantum Algorithms (VQAs) are a promising approach to leverage Noisy Intermediate-Scale Quantum (NISQ) computers. However, choosing optimal quantum circuits that efficiently solve a given VQA problem is a non-trivial task. Quantum Architecture Search (QAS) algorithms enable automatic generation of quantum circuits tailored to the provided problem. Existing QAS approaches typically adapt classical neural architecture search techniques, training machine learning models to sample relevant circuits, but often overlook the inherent quantum nature of the circuits they produce. By reformulating QAS from a quantum perspective, we propose a sampling-free differentiable QAS algorithm that models the search process as the evolution of a quantum mixed state, which emerges from the search space of quantum circuits. The mixed state formulation also enables our method to incorporate generic noise models, for example the depolarizing channel, which cannot be modeled by state vector simulation. We validate our method by finding circuits for state initialization and Hamiltonian optimization tasks, namely the variational quantum eigensolver and the unweighted max-cut problems. We show our approach to be comparable to, if not outperform, existing QAS techniques while requiring significantly fewer quantum simulations during training, and also show improved robustness levels to noise.

Contextual Integrity in LLMs via Reasoning and Reinforcement Learning

arXiv:2506.04245v2 Announce Type: replace-cross Abstract: As the era of autonomous agents making decisions on behalf of users unfolds, ensuring contextual integrity (CI) -- what is the appropriate information to share while carrying out a certain task -- becomes a central question to the field. We posit that CI demands a form of reasoning where the agent needs to reason about the context in which it is operating. To test this, we first prompt LLMs to reason explicitly about CI when deciding what information to disclose. We then extend this approach by developing a reinforcement learning (RL) framework that further instills in models the reasoning necessary to achieve CI. Using a synthetic, automatically created, dataset of only $\sim700$ examples but with diverse contexts and information disclosure norms, we show that our method substantially reduces inappropriate information disclosure while maintaining task performance across multiple model sizes and families. Importantly, improvements transfer from this synthetic dataset to established CI benchmarks such as PrivacyLens that has human annotations and evaluates privacy leakage of AI assistants in actions and tool calls.

MedAgentGym: A Scalable Agentic Training Environment for Code-Centric Reasoning in Biomedical Data Science

arXiv:2506.04405v2 Announce Type: replace-cross Abstract: We introduce MedAgentGym, a scalable and interactive training environment designed to enhance coding-based biomedical reasoning capabilities in large language model (LLM) agents. MedAgentGym comprises 72,413 task instances across 129 categories derived from 12 authentic real-world biomedical scenarios. Tasks are encapsulated within executable sandbox environments, each featuring detailed task specifications, interactive feedback mechanisms, verifiable ground truth annotations, and scalable training trajectory generation. Extensive benchmarking of 29 LLMs reveals substantial performance disparities in biomedical data science between commercial and open-source LLMs. Leveraging efficient multi-threaded and multi-turn trajectory sampling in MedAgentGym, Med-Copilot achieves performance gains of +43.02% and +45.28% from offline and online reinforcement learning, respectively, demonstrating MedAgentGym as an effective training ground while establishing itself as a cost-effective, privacy-preserving alternative competitive with proprietary LLMs (gpt-4o). By offering a unified execution environment with a comprehensive benchmark and accessible, extensible training resources, MedAgentGym delivers an integrated platform to develop LLM-based coding assistants for advanced biomedical data science.

Beyond RLHF and NLHF: Population-Proportional Alignment under an Axiomatic Framework

arXiv:2506.05619v2 Announce Type: replace-cross Abstract: Conventional preference learning methods often prioritize opinions held more widely when aggregating preferences from multiple evaluators. This may result in policies that are biased in favor of some types of opinions or groups and susceptible to strategic manipulation. To address this issue, we develop a novel preference learning framework capable of aligning aggregate opinions and policies proportionally with the true population distribution of evaluator preferences. Grounded in social choice theory, our approach infers the feasible set of evaluator population distributions directly from pairwise comparison data. Using these estimates, the algorithm constructs a policy that satisfies foundational axioms from social choice theory, namely monotonicity and Pareto efficiency, as well as our newly-introduced axioms of population-proportional alignment and population-bounded manipulability. Moreover, we propose a soft-max relaxation method that smoothly trade-offs population-proportional alignment with the selection of the Condorcet winner (which beats all other options in pairwise comparisons). Finally, we validate the effectiveness and scalability of our approach through experiments on both tabular recommendation tasks and large language model alignment.

Learning What Reinforcement Learning Can't: Interleaved Online Fine-Tuning for Hardest Questions

arXiv:2506.07527v2 Announce Type: replace-cross Abstract: Recent advances in large language model (LLM) reasoning have shown that sophisticated behaviors such as planning and self-reflection can emerge through reinforcement learning (RL). However, despite these successes, RL in its current form remains insufficient to induce capabilities that exceed the limitations of the base model, as it is primarily optimized based on existing knowledge of the model rather than facilitating the acquisition of new information. To address this limitation, we employ supervised fine-tuning (SFT) to learn what RL cannot, which enables the incorporation of new knowledge and reasoning patterns by leveraging high-quality demonstration data. We analyze the training dynamics of RL and SFT for LLM reasoning and find that RL excels at maintaining and improving performance on questions within the model's original capabilities, while SFT is more effective at enabling progress on questions beyond the current scope of the model. Motivated by the complementary strengths of RL and SFT, we introduce a novel training approach, \textbf{ReLIFT} (\textbf{Re}inforcement \textbf{L}earning \textbf{I}nterleaved with Online \textbf{F}ine-\textbf{T}uning). In ReLIFT, the model is primarily trained using RL, but when it encounters challenging questions, high-quality solutions are collected for fine-tuning, and the training process alternates between RL and fine-tuning to enhance the model's reasoning abilities. ReLIFT achieves an average improvement of over +5.2 points across five competition-level benchmarks and one out-of-distribution benchmark compared to other zero-RL models. Furthermore, we demonstrate that ReLIFT outperforms both RL and SFT while using only 13\% of the detailed demonstration data, highlighting its scalability. These results provide compelling evidence that ReLIFT overcomes the fundamental limitations of RL and underscores the significant potential.

Probably Approximately Correct Labels

arXiv:2506.10908v2 Announce Type: replace-cross Abstract: Obtaining high-quality labeled datasets is often costly, requiring either human annotation or expensive experiments. In theory, powerful pre-trained AI models provide an opportunity to automatically label datasets and save costs. Unfortunately, these models come with no guarantees on their accuracy, making wholesale replacement of manual labeling impractical. In this work, we propose a method for leveraging pre-trained AI models to curate cost-effective and high-quality datasets. In particular, our approach results in probably approximately correct labels: with high probability, the overall labeling error is small. Our method is nonasymptotically valid under minimal assumptions on the dataset or the AI model being studied, and thus enables rigorous yet efficient dataset curation using modern AI models. We demonstrate the benefits of the methodology through text annotation with large language models, image labeling with pre-trained vision models, and protein folding analysis with AlphaFold.

Machine Learning as Iterated Belief Change a la Darwiche and Pearl

arXiv:2506.13157v2 Announce Type: replace-cross Abstract: Artificial Neural Networks (ANNs) are powerful machine-learning models capable of capturing intricate non-linear relationships. They are widely used nowadays across numerous scientific and engineering domains, driving advancements in both research and real-world applications. In our recent work, we focused on the statics and dynamics of a particular subclass of ANNs, which we refer to as binary ANNs. A binary ANN is a feed-forward network in which both inputs and outputs are restricted to binary values, making it particularly suitable for a variety of practical use cases. Our previous study approached binary ANNs through the lens of belief-change theory, specifically the Alchourron, Gardenfors and Makinson (AGM) framework, yielding several key insights. Most notably, we demonstrated that the knowledge embodied in a binary ANN (expressed through its input-output behaviour) can be symbolically represented using a propositional logic language. Moreover, the process of modifying a belief set (through revision or contraction) was mapped onto a gradual transition through a series of intermediate belief sets. Analogously, the training of binary ANNs was conceptualized as a sequence of such belief-set transitions, which we showed can be formalized using full-meet AGM-style belief change. In the present article, we extend this line of investigation by addressing some critical limitations of our previous study. Specifically, we show that Dalal's method for belief change naturally induces a structured, gradual evolution of states of belief. More importantly, given the known shortcomings of full-meet belief change, we demonstrate that the training dynamics of binary ANNs can be more effectively modelled using robust AGM-style change operations -- namely, lexicographic revision and moderate contraction -- that align with the Darwiche-Pearl framework for iterated belief change.

Making Logic a First-Class Citizen in Network Data Generation with ML

arXiv:2506.23964v2 Announce Type: replace-cross Abstract: Generative ML models are increasingly popular in networking for tasks such as telemetry imputation, prediction, and synthetic trace generation. Despite their capabilities, they suffer from two shortcomings: (i) their output is often visibly violating well-known networking rules, which undermines their trustworthiness; and (ii) they are difficult to control, frequently requiring retraining even for minor changes. To address these limitations and unlock the benefits of generative models for networking, we propose a new paradigm for integrating explicit network knowledge in the form of first-order logic rules into ML models used for networking tasks. Rules capture well-known relationships among used signals, e.g., that increased latency precedes packet loss. While the idea is conceptually straightforward, its realization is challenging: networking knowledge is rarely formalized into rules, and naively injecting them into ML models often hampers ML's effectiveness. This paper introduces NetNomos a multi-stage framework that (1) learns rules directly from data (e.g., measurements); (2) filters them to distinguish semantically meaningful ones; and (3) enforces them through a collaborative generation between an ML model and an SMT solver.

Self-Correction Bench: Uncovering and Addressing the Self-Correction Blind Spot in Large Language Models

arXiv:2507.02778v2 Announce Type: replace-cross Abstract: Although large language models (LLMs) have transformed AI, they still make mistakes and can explore unproductive reasoning paths. Self-correction capability is essential for deploying LLMs in safety-critical applications. We uncover a systematic failure: LLMs cannot correct errors in their own outputs while successfully correcting identical errors from external sources - a limitation we term the Self-Correction Blind Spot. To study this phenomenon, we introduce Self-Correction Bench, an evaluation framework to measure this phenomenon through controlled error injection at three complexity levels. Testing 14 open-source non-reasoning models, we find an average 64.5% blind spot rate. We provide multiple lines of evidence suggesting this limitation may be influenced by training data: human demonstrations rarely include error-correction sequences (favoring error-free responses), whereas reinforcement learning (RL) trained models learn error correction via outcome feedback. Remarkably, appending a minimal "Wait" prompt activates a 89.3% reduction in blind spots, suggesting dormant capabilities that require triggering. Our work highlights a critical limitation potentially influenced by training distribution and offers a practical approach to enhance LLM reliability and trustworthiness - vital for safety-critical domains.

Conformal Prediction for Long-Tailed Classification

arXiv:2507.06867v2 Announce Type: replace-cross Abstract: Many real-world classification problems, such as plant identification, have extremely long-tailed class distributions. In order for prediction sets to be useful in such settings, they should (i) provide good class-conditional coverage, ensuring that rare classes are not systematically omitted from the prediction sets, and (ii) be a reasonable size, allowing users to easily verify candidate labels. Unfortunately, existing conformal prediction methods, when applied to the long-tailed setting, force practitioners to make a binary choice between small sets with poor class-conditional coverage or sets with very good class-conditional coverage but that are extremely large. We propose methods with guaranteed marginal coverage that smoothly trade off between set size and class-conditional coverage. First, we introduce a new conformal score function called prevalence-adjusted softmax that targets macro-coverage, a relaxed notion of class-conditional coverage. Second, we propose a new procedure that interpolates between marginal and class-conditional conformal prediction by linearly interpolating their conformal score thresholds. We demonstrate our methods on Pl@ntNet-300K and iNaturalist-2018, two long-tailed image datasets with 1,081 and 8,142 classes, respectively.

Divergence Minimization Preference Optimization for Diffusion Model Alignment

arXiv:2507.07510v2 Announce Type: replace-cross Abstract: Diffusion models have achieved remarkable success in generating realistic and versatile images from text prompts. Inspired by the recent advancements of language models, there is an increasing interest in further improving the models by aligning with human preferences. However, we investigate alignment from a divergence minimization perspective and reveal that existing preference optimization methods are typically trapped in suboptimal mean-seeking optimization. In this paper, we introduce Divergence Minimization Preference Optimization (DMPO), a novel and principled method for aligning diffusion models by minimizing reverse KL divergence, which asymptotically enjoys the same optimization direction as original RL. We provide rigorous analysis to justify the effectiveness of DMPO and conduct comprehensive experiments to validate its empirical strength across both human evaluations and automatic metrics. Our extensive results show that diffusion models fine-tuned with DMPO can consistently outperform or match existing techniques, specifically consistently outperforming all baseline models across different base models and test sets, achieving the best PickScore in every case, demonstrating the method's superiority in aligning generative behavior with desired outputs. Overall, DMPO unlocks a robust and elegant pathway for preference alignment, bridging principled theory with practical performance in diffusion models.

MapIQ: Evaluating Multimodal Large Language Models for Map Question Answering

arXiv:2507.11625v2 Announce Type: replace-cross Abstract: Recent advancements in multimodal large language models (MLLMs) have driven researchers to explore how well these models read data visualizations, e.g., bar charts, scatter plots. More recently, attention has shifted to visual question answering with maps (Map-VQA). However, Map-VQA research has primarily focused on choropleth maps, which cover only a limited range of thematic categories and visual analytical tasks. To address these gaps, we introduce MapIQ, a benchmark dataset comprising 14,706 question-answer pairs across three map types: choropleth maps, cartograms, and proportional symbol maps spanning topics from six distinct themes (e.g., housing, crime). We evaluate multiple MLLMs using six visual analytical tasks, comparing their performance against one another and a human baseline. An additional experiment examining the impact of map design changes (e.g., altered color schemes, modified legend designs, and removal of map elements) provides insights into the robustness and sensitivity of MLLMs, their reliance on internal geographic knowledge, and potential avenues for improving Map-VQA performance.

Doubly robust outlier resistant inference on causal treatment effect

arXiv:2507.17439v3 Announce Type: replace-cross Abstract: Outliers can severely distort causal effect estimation in observational studies, especially in small samples. We develop a doubly robust estimator of the ATE under a contaminated-data model that explicitly accommodates outliers. Robustness to outliers is delivered via a bounded-influence estimating equation for the outcome model and covariate balancing propensity scores (CBPS) for treatment assignment. To mitigate overfitting in high dimensions, we incorporate variable selection and unify all components within a penalized empirical likelihood framework. For further inference, we derive an optimal finite-sample confidence interval (CI) whose endpoints are invariant to outliers under the contaminated model. Across extensive simulations and two gene-expression applications (Golub; Khan pediatric tumor), the proposed ATE estimator and finite-sample CI outperform state-of-the-art competitors in bias, mean squared error, empirical coverage, and interval length over a wide range of contamination levels and sample sizes.

Riemannian Optimization on Tree Tensor Networks with Application in Machine Learning

arXiv:2507.21726v2 Announce Type: replace-cross Abstract: Tree tensor networks (TTNs) are widely used in low-rank approximation and quantum many-body simulation. In this work, we present a formal analysis of the differential geometry underlying TTNs. Building on this foundation, we develop efficient first- and second-order optimization algorithms that exploit the intrinsic quotient structure of TTNs. Additionally, we devise a backpropagation algorithm for training TTNs in a kernel learning setting. We validate our methods through numerical experiments on a representative machine learning task.

MIBoost: A Gradient Boosting Algorithm for Variable Selection After Multiple Imputation

arXiv:2507.21807v4 Announce Type: replace-cross Abstract: Statistical learning methods for automated variable selection, such as LASSO, elastic nets, or gradient boosting, have become increasingly popular tools for building powerful prediction models. Yet, in practice, analyses are often complicated by missing data. The most widely used approach to address missingness is multiple imputation, which involves creating several completed datasets. However, there is an ongoing debate on how to perform model selection in the presence of multiple imputed datasets. Simple strategies, such as pooling models across datasets, have been shown to have suboptimal properties. Although more sophisticated methods exist, they are often difficult to implement and therefore not widely applied. In contrast, two recent approaches modify the regularization methods LASSO and elastic nets by defining a single loss function, resulting in a unified set of coefficients across imputations. Our key contribution is to extend this principle to the framework of component-wise gradient boosting by proposing MIBoost, a novel algorithm that employs a uniform variable-selection mechanism across imputed datasets. Simulation studies suggest that our approach yields prediction performance comparable to that of these recently proposed methods.

Trainable Dynamic Mask Sparse Attention

arXiv:2508.02124v4 Announce Type: replace-cross Abstract: In large language models, the demand for modeling long contexts is ever-increasing, yet the quadratic complexity of standard self-attention presents a significant bottleneck. While existing sparse attention mechanisms enhance efficiency, they often suffer from limitations such as static patterns and information loss. This paper introduces a Trainable Dynamic Mask Sparse Attention mechanism that addresses these challenges through three key innovations. First, it leverages value vectors to dynamically generate content-aware sparse masks, enabling the model to adaptively identify and focus on crucial information. Second, it implements a position-aware sparse attention computation that effectively skips unnecessary computational regions. Finally, we ensure that the introduced dynamic masks and sparse weights do not obstruct gradients, thereby supporting end-to-end training. This dual-sparsity design allows the model to retain complete information while significantly reducing computational complexity, achieving an excellent balance between efficiency and performance. We validate the performance of Dynamic Mask Attention through comprehensive experiments. Comparative studies demonstrate that our method consistently achieves Pareto dominance across various tasks, including scaling laws, multi-query associative recall, general benchmarks, and needle-in-a-haystack tests, delivering up to 10 times acceleration. These results highlight its capability to effectively balance model efficiency with long-context modeling. Our computational kernel is open-sourced at https://github.com/SmallDoges/flash-dmattn to facilitate further research and application within the community.

The Cosine Schedule is Fisher-Rao-Optimal for Masked Discrete Diffusion Models

arXiv:2508.04884v2 Announce Type: replace-cross Abstract: In this work, we study the problem of choosing the discretisation schedule for sampling from masked discrete diffusion models in terms of the information geometry of the induced probability path. Specifically, we show that the optimal schedule under the Fisher-Rao geometry recovers the popularly-used cosine schedule.

LPI-RIT at LeWiDi-2025: Improving Distributional Predictions via Metadata and Loss Reweighting with DisCo

arXiv:2508.08163v2 Announce Type: replace-cross Abstract: The Learning With Disagreements (LeWiDi) 2025 shared task aims to model annotator disagreement through soft label distribution prediction and perspectivist evaluation, which focuses on modeling individual annotators. We adapt DisCo (Distribution from Context), a neural architecture that jointly models item-level and annotator-level label distributions, and present detailed analysis and improvements. In this paper, we extend DisCo by introducing annotator metadata embeddings, enhancing input representations, and multi-objective training losses to capture disagreement patterns better. Through extensive experiments, we demonstrate substantial improvements in both soft and perspectivist evaluation metrics across three datasets. We also conduct in-depth calibration and error analyses that reveal when and why disagreement-aware modeling improves. Our findings show that disagreement can be better captured by conditioning on annotator demographics and by optimizing directly for distributional metrics, yielding consistent improvements across datasets.

TSLA: A Task-Specific Learning Adaptation for Semantic Segmentation on Autonomous Vehicles Platform

arXiv:2508.12279v2 Announce Type: replace-cross Abstract: Autonomous driving platforms encounter diverse driving scenarios, each with varying hardware resources and precision requirements. Given the computational limitations of embedded devices, it is crucial to consider computing costs when deploying on target platforms like the NVIDIA\textsuperscript{\textregistered} DRIVE PX 2. Our objective is to customize the semantic segmentation network according to the computing power and specific scenarios of autonomous driving hardware. We implement dynamic adaptability through a three-tier control mechanism -- width multiplier, classifier depth, and classifier kernel -- allowing fine-grained control over model components based on hardware constraints and task requirements. This adaptability facilitates broad model scaling, targeted refinement of the final layers, and scenario-specific optimization of kernel sizes, leading to improved resource allocation and performance. Additionally, we leverage Bayesian Optimization with surrogate modeling to efficiently explore hyperparameter spaces under tight computational budgets. Our approach addresses scenario-specific and task-specific requirements through automatic parameter search, accommodating the unique computational complexity and accuracy needs of autonomous driving. It scales its Multiply-Accumulate Operations (MACs) for Task-Specific Learning Adaptation (TSLA), resulting in alternative configurations tailored to diverse self-driving tasks. These TSLA customizations maximize computational capacity and model accuracy, optimizing hardware utilization.

OptimalThinkingBench: Evaluating Over and Underthinking in LLMs

arXiv:2508.13141v2 Announce Type: replace-cross Abstract: Thinking LLMs solve complex tasks at the expense of increased compute and overthinking on simpler problems, while non-thinking LLMs are faster and cheaper but underthink on harder reasoning problems. This has led to the development of separate thinking and non-thinking LLM variants, leaving the onus of selecting the optimal model for each query on the end user. We introduce OptimalThinkingBench, a unified benchmark that jointly evaluates overthinking and underthinking in LLMs and also encourages the development of optimally-thinking models that balance performance and efficiency. Our benchmark comprises two sub-benchmarks: OverthinkingBench, featuring simple math and general queries in 72 domains, and UnderthinkingBench, containing 11 challenging reasoning tasks along with harder math problems. Using novel thinking-adjusted accuracy metrics, we extensively evaluate 33 different thinking and non-thinking models and show that no model is able to optimally think on our benchmark. Thinking models often overthink for hundreds of tokens on the simplest user queries without improving performance. In contrast, large non-thinking models underthink, often falling short of much smaller thinking models. We further explore several methods to encourage optimal thinking, but find that these approaches often improve on one sub-benchmark at the expense of the other, highlighting the need for better unified and optimal models in the future.

Understanding and Improving Shampoo and SOAP via Kullback-Leibler Minimization

arXiv:2509.03378v3 Announce Type: replace-cross Abstract: Shampoo and its efficient variant, SOAP, employ structured second-moment estimations and have shown strong performance for training neural networks (NNs). In practice, however, Shampoo typically requires step-size grafting with Adam to be competitive, and SOAP mitigates this by applying Adam in Shampoo's eigenbasis -- at the cost of additional memory overhead from Adam in both methods. Prior analyses have largely relied on the Frobenius norm to motivate these estimation schemes. We instead recast their estimation procedures as covariance estimation under Kullback-Leibler (KL) divergence minimization, revealing a previously overlooked theoretical limitation and motivating principled redesigns. Building on this perspective, we develop $\textbf{KL-Shampoo}$ and $\textbf{KL-SOAP}$, practical schemes that match or exceed the performance of Shampoo and SOAP in NN pre-training while achieving SOAP-level per-iteration runtime. Notably, KL-Shampoo does not rely on Adam to attain competitive performance, eliminating the memory overhead introduced by Adam. Across our experiments, KL-Shampoo consistently outperforms SOAP, Shampoo, and even KL-SOAP, establishing the KL-based approach as a compelling foundation for designing structured methods in NN optimization.

ArcMemo: Abstract Reasoning Composition with Lifelong LLM Memory

arXiv:2509.04439v3 Announce Type: replace-cross Abstract: While inference-time scaling enables LLMs to carry out increasingly long and capable reasoning traces, the patterns and insights uncovered during these traces are immediately discarded once the context window is reset for a new query. External memory is a natural way to persist these discoveries, and recent work has shown clear benefits for reasoning-intensive tasks. We see an opportunity to make such memories more broadly reusable and scalable by moving beyond instance-based memory entries (e.g. exact query/response pairs, or summaries tightly coupled with the original problem context) toward concept-level memory: reusable, modular abstractions distilled from solution traces and stored in natural language. For future queries, relevant concepts are selectively retrieved and integrated into the prompt, enabling test-time continual learning without weight updates. Our design introduces new strategies for abstracting takeaways from rollouts and retrieving entries for new queries, promoting reuse and allowing memory to expand with additional experiences. We evaluate on ARC-AGI, a benchmark that stresses compositional generalization and abstract reasoning, making it a natural fit for concept memory. Our method yields a 7.5% relative gain over a strong no-memory baseline with performance continuing to scale with inference compute. We find abstract concepts to be the most consistent memory design, outscoring the baseline at all tested inference compute scales. Moreover, dynamically updating memory during test-time outperforms fixed settings, supporting the hypothesis that accumulating and abstracting patterns enables further solutions in a form of self-improvement. Code is available at https://github.com/matt-seb-ho/arc_memo.

Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotation

arXiv:2509.08825v2 Announce Type: replace-cross Abstract: Large language models are rapidly transforming social science research by enabling the automation of labor-intensive tasks like data annotation and text analysis. However, LLM outputs vary significantly depending on the implementation choices made by researchers (e.g., model selection or prompting strategy). Such variation can introduce systematic biases and random errors, which propagate to downstream analyses and cause Type I (false positive), Type II (false negative), Type S (wrong sign), or Type M (exaggerated effect) errors. We call this phenomenon where configuration choices lead to incorrect conclusions LLM hacking. We find that intentional LLM hacking is strikingly simple. By replicating 37 data annotation tasks from 21 published social science studies, we show that, with just a handful of prompt paraphrases, virtually anything can be presented as statistically significant. Beyond intentional manipulation, our analysis of 13 million labels from 18 different LLMs across 2361 realistic hypotheses shows that there is also a high risk of accidental LLM hacking, even when following standard research practices. We find incorrect conclusions in approximately 31% of hypotheses for state-of-the-art LLMs, and in half the hypotheses for smaller language models. While higher task performance and stronger general model capabilities reduce LLM hacking risk, even highly accurate models remain susceptible. The risk of LLM hacking decreases as effect sizes increase, indicating the need for more rigorous verification of LLM-based findings near significance thresholds. We analyze 21 mitigation techniques and find that human annotations provide crucial protection against false positives. Common regression estimator correction techniques can restore valid inference but trade off Type I vs. Type II errors. We publish a list of practical recommendations to prevent LLM hacking.

Population-Aligned Persona Generation for LLM-based Social Simulation

arXiv:2509.10127v2 Announce Type: replace-cross Abstract: Recent advances in large language models (LLMs) have enabled human-like social simulations at unprecedented scale and fidelity, offering new opportunities for computational social science. A key challenge, however, is the construction of persona sets that authentically represent the diversity and distribution of real-world populations. Most existing LLM-based social simulation studies focus primarily on designing agentic frameworks and simulation environments, often overlooking the complexities of persona generation and the potential biases introduced by unrepresentative persona sets. In this paper, we propose a systematic framework for synthesizing high-quality, population-aligned persona sets for LLM-driven social simulation. Our approach begins by leveraging LLMs to generate narrative personas from long-term social media data, followed by rigorous quality assessment to filter out low-fidelity profiles. We then apply importance sampling to achieve global alignment with reference psychometric distributions, such as the Big Five personality traits. To address the needs of specific simulation contexts, we further introduce a task-specific module that adapts the globally aligned persona set to targeted subpopulations. Extensive experiments demonstrate that our method significantly reduces population-level bias and enables accurate, flexible social simulation for a wide range of research and policy applications.

CEMTM: Contextual Embedding-based Multimodal Topic Modeling

arXiv:2509.11465v2 Announce Type: replace-cross Abstract: We introduce CEMTM, a context-enhanced multimodal topic model designed to infer coherent and interpretable topic structures from both short and long documents containing text and images. CEMTM builds on fine-tuned large vision language models (LVLMs) to obtain contextualized embeddings, and employs a distributional attention mechanism to weight token-level contributions to topic inference. A reconstruction objective aligns topic-based representations with the document embedding, encouraging semantic consistency across modalities. Unlike existing approaches, CEMTM can process multiple images per document without repeated encoding and maintains interpretability through explicit word-topic and document-topic distributions. Extensive experiments on six multimodal benchmarks show that CEMTM consistently outperforms unimodal and multimodal baselines, achieving a remarkable average LLM score of 2.61. Further analysis shows its effectiveness in downstream few-shot retrieval and its ability to capture visually grounded semantics in complex domains such as scientific articles.

QDFlow: A Python package for physics simulations of quantum dot devices

arXiv:2509.13298v2 Announce Type: replace-cross Abstract: Recent advances in machine learning (ML) have accelerated progress in calibrating and operating quantum dot (QD) devices. However, most ML approaches rely on access to large, representative datasets designed to capture the full spectrum of data quality encountered in practice, with both high- and low-quality data for training, benchmarking, and validation, with labels capturing key features of the device state. Collating such datasets experimentally is challenging due to limited data availability, slow measurement bandwidths, and the labor-intensive nature of labeling. QDFlow is an open-source physics simulator for multi-QD arrays that generates realistic synthetic data with ground-truth labels. QDFlow combines a self-consistent Thomas-Fermi solver, a dynamic capacitance model, and flexible noise modules to simulate charge stability diagrams and ray-based data closely resembling experiments. With an extensive set of parameters that can be varied and customizable noise models, QDFlow supports the creation of large, diverse datasets for ML development, benchmarking, and quantum device research.

FedMentor: Domain-Aware Differential Privacy for Heterogeneous Federated LLMs in Mental Health

arXiv:2509.14275v2 Announce Type: replace-cross Abstract: Privacy-preserving adaptation of Large Language Models (LLMs) in sensitive domains (e.g., mental health) requires balancing strict confidentiality with model utility and safety. We propose FedMentor, a federated fine-tuning framework that integrates Low-Rank Adaptation (LoRA) and domain-aware Differential Privacy (DP) to meet per-domain privacy budgets while maintaining performance. Each client (domain) applies a custom DP noise scale proportional to its data sensitivity, and the server adaptively reduces noise when utility falls below a threshold. In experiments on three mental health datasets, we show that FedMentor improves safety over standard Federated Learning (FL) without privacy, raising safe output rates by up to three points and lowering toxicity, while maintaining utility (BERTScore F1 and ROUGE-L) within 0.5% of the non-private baseline and close to the centralized upper bound. The framework scales to backbones with up to 1.7B parameters on single-GPU clients, requiring < 173 MB of communication per-round. FedMentor demonstrates a practical approach to privately fine-tune LLMs for safer deployments in healthcare and other sensitive fields.

The Narcissus Hypothesis: Descending to the Rung of Illusion

arXiv:2509.17999v3 Announce Type: replace-cross Abstract: Modern foundational models increasingly reflect not just world knowledge, but patterns of human preference embedded in their training data. We hypothesize that recursive alignment-via human feedback and model-generated corpora-induces a social desirability bias, nudging models to favor agreeable or flattering responses over objective reasoning. We refer to it as the Narcissus Hypothesis and test it across 31 models using standardized personality assessments and a novel Social Desirability Bias score. Results reveal a significant drift toward socially conforming traits, with profound implications for corpus integrity and the reliability of downstream inferences. We then offer a novel epistemological interpretation, tracing how recursive bias may collapse higher-order reasoning down Pearl's Ladder of Causality, culminating in what we refer to as the Rung of Illusion.

Fine-Grained AI Model Caching and Downloading With Coordinated Multipoint Broadcasting in Multi-Cell Edge Networks

arXiv:2509.19341v2 Announce Type: replace-cross Abstract: 6G networks are envisioned to support on-demand AI model downloading to accommodate diverse inference requirements of end users. By proactively caching models at edge nodes, users can retrieve the requested models with low latency for on-device AI inference. However, the substantial size of contemporary AI models poses significant challenges for edge caching under limited storage capacity, as well as for the concurrent delivery of heterogeneous models over wireless channels. To address these challenges, we propose a fine-grained AI model caching and downloading system that exploits parameter reusability, stemming from the common practice of fine-tuning task-specific models from a shared pre-trained model with frozen parameters. This system selectively caches model parameter blocks (PBs) at edge nodes, eliminating redundant storage of reusable parameters across different cached models. Additionally, it incorporates coordinated multipoint (CoMP) broadcasting to simultaneously deliver reusable PBs to multiple users, thereby enhancing downlink spectrum utilization. Under this arrangement, we formulate a model downloading delay minimization problem to jointly optimize PB caching, migration (among edge nodes), and broadcasting beamforming. To tackle this intractable problem, we develop a distributed multi-agent learning framework that enables edge nodes to explicitly learn mutual influence among their actions, thereby facilitating cooperation. Furthermore, a data augmentation approach is proposed to adaptively generate synthetic training samples through a predictive model, boosting sample efficiency and accelerating policy learning. Both theoretical analysis and simulation experiments validate the superior convergence performance of the proposed learning framework.

The Syntax and Semantics of einsum

arXiv:2509.20020v2 Announce Type: replace-cross Abstract: In 2011, einsum was introduced to NumPy as a practical and convenient notation for tensor expressions in machine learning, quantum circuit simulation, and other fields. It has since been implemented in additional Python frameworks such as PyTorch and TensorFlow, as well as in other programming languages such as Julia. Despite its practical success, the einsum notation still lacks a solid theoretical basis, and is not unified across the different frameworks, limiting opportunities for formal reasoning and systematic optimization. In this work, we discuss the terminology of tensor expressions and provide a formal definition of the einsum language. Based on this definition, we formalize and prove important equivalence rules for tensor expressions and highlight their relevance in practical applications.

Region-of-Interest Augmentation for Mammography Classification under Patient-Level Cross-Validation

arXiv:2509.20585v2 Announce Type: replace-cross Abstract: Breast cancer screening with mammography remains central to early detection and mortality reduction. Deep learning has shown strong potential for automating mammogram interpretation, yet limited-resolution datasets and small sample sizes continue to restrict performance. We revisit the Mini-DDSM dataset (9,684 images; 2,414 patients) and introduce a lightweight region-of-interest (ROI) augmentation strategy. During training, full images are probabilistically replaced with random ROI crops sampled from a precomputed, label-free bounding-box bank, with optional jitter to increase variability. We evaluate under strict patient-level cross-validation and report ROC-AUC, PR-AUC, and training-time efficiency metrics (throughput and GPU memory). Because ROI augmentation is training-only, inference-time cost remains unchanged. On Mini-DDSM, ROI augmentation (best: p_roi = 0.10, alpha = 0.10) yields modest average ROC-AUC gains, with performance varying across folds; PR-AUC is flat to slightly lower. These results demonstrate that simple, data-centric ROI strategies can enhance mammography classification in constrained settings without requiring additional labels or architectural modifications.

Do Sparse Subnetworks Exhibit Cognitively Aligned Attention? Effects of Pruning on Saliency Map Fidelity, Sparsity, and Concept Coherence

arXiv:2509.21387v3 Announce Type: replace-cross Abstract: Prior works have shown that neural networks can be heavily pruned while preserving performance, but the impact of pruning on model interpretability remains unclear. In this work, we investigate how magnitude-based pruning followed by fine-tuning affects both low-level saliency maps and high-level concept representations. Using a ResNet-18 trained on ImageNette, we compare post-hoc explanations from Vanilla Gradients (VG) and Integrated Gradients (IG) across pruning levels, evaluating sparsity and faithfulness. We further apply CRAFT-based concept extraction to track changes in semantic coherence of learned concepts. Our results show that light-to-moderate pruning improves saliency-map focus and faithfulness while retaining distinct, semantically meaningful concepts. In contrast, aggressive pruning merges heterogeneous features, reducing saliency map sparsity and concept coherence despite maintaining accuracy. These findings suggest that while pruning can shape internal representations toward more human-aligned attention patterns, excessive pruning undermines interpretability.

HEART: Emotionally-driven test-time scaling of Language Models

arXiv:2509.22876v2 Announce Type: replace-cross Abstract: Test-time scaling has shown considerable success in improving the performance of language models on complex reasoning tasks without requiring fine-tuning. However, current strategies such as self-reflection primarily focus on logical or structural refinement. They do not leverage the guiding potential of affective feedback. Inspired by psychological research showing that emotions can modulate cognitive performance, we introduce HEART--a novel framework that uses emotionally-driven prompts for iterative self-correction. HEART provides feedback on a model's incorrect response using a curated set of concise, emotionally charged phrases based on the six universal emotions categorized by Dr. Paul Ekman. By systematically varying the emotional tone of the feedback across iterations, our method guides the model to escape flawed reasoning paths and explore more promising alternatives. We evaluate our framework on challenging reasoning benchmarks including OlympiadBench, Humanity's Last Exam, and SimpleQA. Our results reveal a significant new phenomenon: when guided by an oracle verifier, this affective iteration protocol unlocks significantly deeper reasoning, leading to consistent and substantial increases in accuracy over state-of-the-art baselines with the same verifier. However, we also identify a critical bottleneck for practical deployment. In a verifier-free setting, it struggles to harness these gains consistently, highlighting as a key challenge for future work. Our findings suggest that the next frontier in machine reasoning may lie not just in refining logic, but also in understanding and leveraging the `HEART' of the models.

Extracting the Structure of Press Releases for Predicting Earnings Announcement Returns

arXiv:2509.24254v2 Announce Type: replace-cross Abstract: We examine how textual features in earnings press releases predict stock returns on earnings announcement days. Using over 138,000 press releases from 2005 to 2023, we compare traditional bag-of-words and BERT-based embeddings. We find that press release content (soft information) is as informative as earnings surprise (hard information), with FinBERT yielding the highest predictive power. Combining models enhances explanatory strength and interpretability of the content of press releases. Stock prices fully reflect the content of press releases at market open. If press releases are leaked, it offers predictive advantage. Topic analysis reveals self-serving bias in managerial narratives. Our framework supports real-time return prediction through the integration of online learning, provides interpretability and reveals the nuanced role of language in price formation.

Pushing LLMs to Their Logical Reasoning Bound: The Role of Data Reasoning Intensity

arXiv:2509.24836v3 Announce Type: replace-cross Abstract: Recent advances in large language models (LLMs) highlight the importance of training data structure and quality in shaping reasoning behavior. However, most existing approaches focus on transforming data formats while neglecting the internal reasoning complexity of training samples, leaving the reasoning potential of data under-explored and underutilized. In this work, we posit that LLM logical reasoning performance is jointly constrained by the potential of the training data and the cognitive capacity of the model. To make this relationship measurable, we introduce Data Reasoning Intensity (DRI), a novel metric that quantifies the latent logical reasoning complexity of samples by decomposing and aggregating their logical structures. This allows us to analyze how well current LLMs utilize logical reasoning signals and identify performance gaps relative to data potential. Based on this insight, we introduce a re-cognizing optimization strategy that systematically enhances the logical reasoning intensity of training data. Rather than increasing data volume, our method re-optimizes existing samples to better align with the LLM's logical reasoning boundary. Extensive experiments show that our approach significantly improves performance and generalization over data-centric strategies. We further validate our method under a reinforcement learning framework. Our results indicate that prioritizing reasoning complexity in data rather than sheer scale or superficial form is essential to realizing LLMs' full cognitive potential.

Low Resource Audio Codec Challenge Baseline Systems

arXiv:2510.00264v2 Announce Type: replace-cross Abstract: The Low-Resource Audio Codec (LRAC) Challenge aims to advance neural audio coding for deployment in resource-constrained environments. The first edition focuses on low-resource neural speech codecs that must operate reliably under everyday noise and reverberation, while satisfying strict constraints on computational complexity, latency, and bitrate. Track 1 targets transparency codecs, which aim to preserve the perceptual transparency of input speech under mild noise and reverberation. Track 2 addresses enhancement codecs, which combine coding and compression with denoising and dereverberation. This paper presents the official baseline systems for both tracks in the 2025 LRAC Challenge. The baselines are convolutional neural codec models with Residual Vector Quantization, trained end-to-end using a combination of adversarial and reconstruction objectives. We detail the data filtering and augmentation strategies, model architectures, optimization procedures, and checkpoint selection criteria.

A first-order method for constrained nonconvex--nonconcave minimax problems under a local Kurdyka-{\L}ojasiewicz condition

arXiv:2510.01168v2 Announce Type: replace-cross Abstract: We study a class of constrained nonconvex--nonconcave minimax problems in which the inner maximization involves potentially complex constraints. Under the assumption that the inner problem of a novel lifted minimax problem satisfies a local Kurdyka-{\L}ojasiewicz (KL) condition, we show that the maximal function of the original problem enjoys a local H\"older smoothness property. We also propose a sequential convex programming (SCP) method for solving constrained optimization problems and establish its convergence rate under a local KL condition. Leveraging these results, we develop an inexact proximal gradient method for the original minimax problem, where the inexact gradient of the maximal function is computed via the SCP method applied to a locally KL-structured subproblem. Finally, we establish complexity guarantees for the proposed method in computing an approximate stationary point of the original minimax problem.

Silent Tokens, Loud Effects: Padding in LLMs

arXiv:2510.01238v2 Announce Type: replace-cross Abstract: Padding tokens are widely used in large language models (LLMs) to equalize sequence lengths during batched inference. While they should be fully masked, implementation errors can cause them to influence computation, and the extent of this influence is not well understood. We systematically study this effect across three open-source model families (Llama, Gemma, Qwen), inserting controlled amounts of padding and evaluating outcomes along four axes: activations, generation quality, bias, and safety. Even small amounts of padding shift hidden representations, degrade quality in smaller models, alter bias in unpredictable ways, and weaken safety guardrails. These findings demonstrate that padding is not a harmless detail but a robustness risk that must be carefully handled in deployment.

NGGAN: Noise Generation GAN Based on the Practical Measurement Dataset for Narrowband Powerline Communications

arXiv:2510.01850v2 Announce Type: replace-cross Abstract: To effectively process impulse noise for narrowband powerline communications (NB-PLCs) transceivers, capturing comprehensive statistics of nonperiodic asynchronous impulsive noise (APIN) is a critical task. However, existing mathematical noise generative models only capture part of the characteristics of noise. In this study, we propose a novel generative adversarial network (GAN) called noise generation GAN (NGGAN) that learns the complicated characteristics of practically measured noise samples for data synthesis. To closely match the statistics of complicated noise over the NB-PLC systems, we measured the NB-PLC noise via the analog coupling and bandpass filtering circuits of a commercial NB-PLC modem to build a realistic dataset. To train NGGAN, we adhere to the following principles: 1) we design the length of input signals that the NGGAN model can fit to facilitate cyclostationary noise generation; 2) the Wasserstein distance is used as a loss function to enhance the similarity between the generated noise and training data; and 3) to measure the similarity performances of GAN-based models based on the mathematical and practically measured datasets, we conduct both quantitative and qualitative analyses. The training datasets include: 1) a piecewise spectral cyclostationary Gaussian model (PSCGM); 2) a frequency-shift (FRESH) filter; and 3) practical measurements from NB-PLC systems. Simulation results demonstrate that the generated noise samples from the proposed NGGAN are highly close to the real noise samples. The principal component analysis (PCA) scatter plots and Fr\'echet inception distance (FID) analysis have shown that NGGAN outperforms other GAN-based models by generating noise samples with superior fidelity and higher diversity.

Automated Defect Detection for Mass-Produced Electronic Components Based on YOLO Object Detection Models

arXiv:2510.01914v2 Announce Type: replace-cross Abstract: Since the defect detection of conventional industry components is time-consuming and labor-intensive, it leads to a significant burden on quality inspection personnel and makes it difficult to manage product quality. In this paper, we propose an automated defect detection system for the dual in-line package (DIP) that is widely used in industry, using digital camera optics and a deep learning (DL)-based model. The two most common defect categories of DIP are examined: (1) surface defects, and (2) pin-leg defects. However, the lack of defective component images leads to a challenge for detection tasks. To solve this problem, the ConSinGAN is used to generate a suitable-sized dataset for training and testing. Four varieties of the YOLO model are investigated (v3, v4, v7, and v9), both in isolation and with the ConSinGAN augmentation. The proposed YOLOv7 with ConSinGAN is superior to the other YOLO versions in accuracy of 95.50\%, detection time of 285 ms, and is far superior to threshold-based approaches. In addition, the supervisory control and data acquisition (SCADA) system is developed, and the associated sensor architecture is described. The proposed automated defect detection can be easily established with numerous types of defects or insufficient defect data.

Comparing Contrastive and Triplet Loss: Variance Analysis and Optimization Behavior

arXiv:2510.02161v2 Announce Type: replace-cross Abstract: Contrastive loss and triplet loss are widely used objectives in deep metric learning, yet their effects on representation quality remain insufficiently understood. We present a theoretical and empirical comparison of these losses, focusing on intra- and inter-class variance and optimization behavior (e.g., greedy updates). Through task-specific experiments with consistent settings on synthetic data and real datasets-MNIST, CIFAR-10-it is shown that triplet loss preserves greater variance within and across classes, supporting finer-grained distinctions in the learned representations. In contrast, contrastive loss tends to compact intra-class embeddings, which may obscure subtle semantic differences. To better understand their optimization dynamics, By examining loss-decay rate, active ratio, and gradient norm, we find that contrastive loss drives many small updates early on, while triplet loss produces fewer but stronger updates that sustain learning on hard examples. Finally, across both classification and retrieval tasks on MNIST, CIFAR-10, CUB-200, and CARS196 datasets, our results consistently show that triplet loss yields superior performance, which suggests using triplet loss for detail retention and hard-sample focus, and contrastive loss for smoother, broad-based embedding refinement.

Quantum Fisher information matrices from R\'enyi relative entropies

arXiv:2510.02218v2 Announce Type: replace-cross Abstract: Quantum generalizations of the Fisher information are important in quantum information science, with applications in high energy and condensed matter physics and in quantum estimation theory, machine learning, and optimization. One can derive a quantum generalization of the Fisher information matrix in a natural way as the Hessian matrix arising in a Taylor expansion of a smooth divergence. Such an approach is appealing for quantum information theorists, given the ubiquity of divergences in quantum information theory. In contrast to the classical case, there is not a unique quantum generalization of the Fisher information matrix, similar to how there is not a unique quantum generalization of the relative entropy or the R\'enyi relative entropy. In this paper, I derive information matrices arising from the log-Euclidean, $\alpha$-$z$, and geometric R\'enyi relative entropies, with the main technical tool for doing so being the method of divided differences for calculating matrix derivatives. Interestingly, for all non-negative values of the R\'enyi parameter $\alpha$, the log-Euclidean R\'enyi relative entropy leads to the Kubo-Mori information matrix, and the geometric R\'enyi relative entropy leads to the right-logarithmic derivative Fisher information matrix. Thus, the resulting information matrices obey the data-processing inequality for all non-negative values of the R\'enyi parameter $\alpha$ even though the original quantities do not. Additionally, I derive and establish basic properties of $\alpha$-$z$ information matrices resulting from the $\alpha$-$z$ R\'enyi relative entropies. For parameterized thermal states and time-evolved states, I establish formulas for their $\alpha$-$z$ information matrices and hybrid quantum-classical algorithms for estimating them, with applications in quantum Boltzmann machine learning.

VidGuard-R1: AI-Generated Video Detection and Explanation via Reasoning MLLMs and RL

arXiv:2510.02282v2 Announce Type: replace-cross Abstract: With the rapid advancement of AI-generated videos, there is an urgent need for effective detection tools to mitigate societal risks such as misinformation and reputational harm. In addition to accurate classification, it is essential that detection models provide interpretable explanations to ensure transparency for regulators and end users. To address these challenges, we introduce VidGuard-R1, the first video authenticity detector that fine-tunes a multi-modal large language model (MLLM) using group relative policy optimization (GRPO). Our model delivers both highly accurate judgments and insightful reasoning. We curate a challenging dataset of 140k real and AI-generated videos produced by state-of-the-art generation models, carefully designing the generation process to maximize discrimination difficulty. We then fine-tune Qwen-VL using GRPO with two specialized reward models that target temporal artifacts and generation complexity. Extensive experiments demonstrate that VidGuard-R1 achieves state-of-the-art zero-shot performance on existing benchmarks, with additional training pushing accuracy above 95%. Case studies further show that VidGuard-R1 produces precise and interpretable rationales behind its predictions. The code is publicly available at https://VidGuard-R1.github.io.

Pretraining with hierarchical memories: separating long-tail and common knowledge

arXiv:2510.02375v2 Announce Type: replace-cross Abstract: The impressive performance gains of modern language models currently rely on scaling parameters: larger models store more world knowledge and reason better. Yet compressing all world knowledge into parameters is unnecessary, as only a fraction is used per prompt, and impractical for edge devices with limited inference-time memory and compute. We address this shortcoming by a memory-augmented architecture and a pretraining strategy aligned with existing hardware paradigms. We introduce small language models that access large hierarchical parametric memory banks encoding world knowledge. During pretraining and inference, we fetch a small, context-dependent memory block and add it to the model. Our pretraining learns to store long-tail world knowledge in the memory parameters, while the small language model acts as an anchor capturing common knowledge and general reasoning abilities. Through trillion-token-scale experiments, we show significant gains: a 160M-parameters model augmented with an 18M-parameters memory fetched from a 4.6B memory bank obtains comparable performance to a regular model with more than 2x the parameters. Through extensive experiments, we study the optimal type and size of parametric memories in transformers, scaling them to over 21B parameters. We find that our proposed hierarchical feed-forward memories work robustly across transformer architectures, whether added during pretraining or post-hoc.

Agentic Additive Manufacturing Alloy Discovery

arXiv:2510.02567v2 Announce Type: replace-cross Abstract: Agentic systems enable the intelligent use of research tooling, augmenting a researcher's ability to investigate and propose novel solutions to existing problems. Within Additive Manufacturing (AM), alloy discovery remains a complex challenge, often requiring expertise in the various domains of materials science, thermodynamic simulations, and experimental analysis. Large Language Model (LLM) enabled agents can facilitate this endeavor by utilizing their extensive knowledge base to dispatch tool calls via Model Context Protocol (MCP) to perform actions such as Thermo-Calc property diagram calculations and lack of fusion process map generation. In addition, the multi-agent system developed in this work is able to effectively reason through complex user prompts and provide analysis on the printability of proposed alloys. These agents can dynamically adjust their task trajectory to the outcomes of tool call results, effectively enabling autonomous decision-making in practical environments. This work aims to utilize LLM enabled agents to automate and accelerate the task of alloy discovery within the field of additive manufacturing and showcase the benefits of adopting this multi-agent system.

FLOWR.root: A flow matching based foundation model for joint multi-purpose structure-aware 3D ligand generation and affinity prediction

arXiv:2510.02578v2 Announce Type: replace-cross Abstract: We present FLOWR:root, an equivariant flow-matching model for pocket-aware 3D ligand generation with joint binding affinity prediction and confidence estimation. The model supports de novo generation, pharmacophore-conditional sampling, fragment elaboration, and multi-endpoint affinity prediction (pIC50, pKi, pKd, pEC50). Training combines large-scale ligand libraries with mixed-fidelity protein-ligand complexes, followed by refinement on curated co-crystal datasets and parameter-efficient finetuning for project-specific adaptation. FLOWR:root achieves state-of-the-art performance in unconditional 3D molecule generation and pocket-conditional ligand design, producing geometrically realistic, low-strain structures. The integrated affinity prediction module demonstrates superior accuracy on the SPINDR test set and outperforms recent models on the Schrodinger FEP+/OpenFE benchmark with substantial speed advantages. As a foundation model, FLOWR:root requires finetuning on project-specific datasets to account for unseen structure-activity landscapes, yielding strong correlation with experimental data. Joint generation and affinity prediction enable inference-time scaling through importance sampling, steering molecular design toward higher-affinity compounds. Case studies validate this: selective CK2$\alpha$ ligand generation against CLK3 shows significant correlation between predicted and quantum-mechanical binding energies, while ER$\alpha$ and TYK2 scaffold elaboration demonstrates strong agreement with QM calculations. By integrating structure-aware generation, affinity estimation, and property-guided sampling, FLOWR:root provides a comprehensive foundation for structure-based drug design spanning hit identification through lead optimization.

What Drives Compositional Generalization in Visual Generative Models?

arXiv:2510.03075v2 Announce Type: replace-cross Abstract: Compositional generalization, the ability to generate novel combinations of known concepts, is a key ingredient for visual generative models. Yet, not all mechanisms that enable or inhibit it are fully understood. In this work, we conduct a systematic study of how various design choices influence compositional generalization in image and video generation in a positive or negative way. Through controlled experiments, we identify two key factors: (i) whether the training objective operates on a discrete or continuous distribution, and (ii) to what extent conditioning provides information about the constituent concepts during training. Building on these insights, we show that relaxing the MaskGIT discrete loss with an auxiliary continuous JEPA-based objective can improve compositional performance in discrete models like MaskGIT.

Improving Online-to-Nonconvex Conversion for Smooth Optimization via Double Optimism

arXiv:2510.03167v2 Announce Type: replace-cross Abstract: A recent breakthrough in nonconvex optimization is the online-to-nonconvex conversion framework of [Cutkosky et al., 2023], which reformulates the task of finding an $\varepsilon$-first-order stationary point as an online learning problem. When both the gradient and the Hessian are Lipschitz continuous, instantiating this framework with two different online learners achieves a complexity of $O(\varepsilon^{-1.75}\log(1/\varepsilon))$ in the deterministic case and a complexity of $O(\varepsilon^{-3.5})$ in the stochastic case. However, this approach suffers from several limitations: (i) the deterministic method relies on a complex double-loop scheme that solves a fixed-point equation to construct hint vectors for an optimistic online learner, introducing an extra logarithmic factor; (ii) the stochastic method assumes a bounded second-order moment of the stochastic gradient, which is stronger than standard variance bounds; and (iii) different online learning algorithms are used in the two settings. In this paper, we address these issues by introducing an online optimistic gradient method based on a novel doubly optimistic hint function. Specifically, we use the gradient at an extrapolated point as the hint, motivated by two optimistic assumptions: that the difference between the hint and the target gradient remains near constant, and that consecutive update directions change slowly due to smoothness. Our method eliminates the need for a double loop and removes the logarithmic factor. Furthermore, by simply replacing full gradients with stochastic gradients and under the standard assumption that their variance is bounded by $\sigma^2$, we obtain a unified algorithm with complexity $O(\varepsilon^{-1.75} + \sigma^2 \varepsilon^{-3.5})$, smoothly interpolating between the best-known deterministic rate and the optimal stochastic rate.

The Space Between Intention and Impact

It’s been said to me that at times my responses can come across as a “no” first-unintentionally, of course, but real all the same.

Not long ago, someone on my team shared an idea they were exploring. I began with a caution, meaning to help them avoid a pitfall I’d learned the hard way. I followed up with suggestions and examples from my own experience. From my perspective, the conversation went smoothly. Only later did I learn, through someone else, that it hadn’t landed that way. They felt dismissed.

That realisation surprised me. I had genuinely believed it was a supportive, collaborative discussion. The truth is, there’s a gap in my self-awareness-a blind spot around how my words or tone might shape someone else’s perception. My intention was to guide, but the impact was to discourage.

I’ve been working to bridge that gap. Thinking has become my best tool so far-slowing down just long enough to consider how I begin a response, to pause before speaking, to ask a question instead of offering an answer. Curiosity softens the space where instinct once jumped in.

Lately I’ve also started ending my days with a short “catalytic retro.” Ten to fifteen minutes to quietly revisit the day’s interactions and decisions, asking not just what went well but what landed wrong. Wins tend to speak for themselves. It’s the misses that offer the real insight, if you’re willing to see them.

That small pause between intention and impact-that’s where growth lives.

Next time: When Accountability Feels Like Blame

AWS Cloud Tutorial | Build and Deploy Your First Application

In today’s digital-first world, cloud computing has become the backbone of modern software development. Among the various cloud platforms available, Amazon Web Services (AWS) stands tall as the global leader. In this AWS Tutorial by Tpoint Tech, we’ll walk you through the fundamentals of AWS and show you how to build and deploy your first cloud application using its powerful tools and services.

What is AWS?

Amazon Web Services (AWS) is a comprehensive cloud platform that offers over 200 services ranging from computing and storage to machine learning and analytics. Whether you’re a startup, enterprise, or individual developer, AWS provides the flexibility and scalability needed to innovate faster and reduce costs.

Some of the most popular AWS services include:

• Amazon EC2 (Elastic Compute Cloud) – Virtual servers for running applications.
• Amazon S3 (Simple Storage Service) – Highly scalable object storage.
• AWS Lambda – Serverless compute service for event-driven tasks.
• Amazon RDS (Relational Database Service) – Managed database service.
• AWS CloudFormation – Infrastructure as code for automated deployment.

This AWS tutorial will help you understand how these services work together to help you build and deploy your first application with ease.

Why Learn AWS?

Before jumping into the hands-on section, let’s quickly understand why AWS skills are in such high demand:

1. Industry Standard: AWS is used by top companies like Netflix, Airbnb, and Spotify.
2. Scalability: You can scale your application from one user to millions seamlessly.
3. Cost Efficiency: Pay only for what you use.
4. Career Growth: AWS certifications open doors to high-paying cloud careers.

At Tpoint Tech, we believe that learning AWS is not just about mastering a platform—it’s about future-proofing your tech career.

Step 1: Setting Up Your AWS Account

The first step in this AWS tutorial is to create an AWS account.

• Visit aws.amazon.com and click Create an AWS Account.
• Provide your email, password, and payment information (you can use the free tier).
• Once verified, you’ll have access to the AWS Management Console.

The AWS Free Tier gives you access to limited usage of services like EC2, S3, and Lambda for the first 12 months—perfect for learning and experimentation.

Step 2: Launching Your First EC2 Instance

Your first cloud application needs a server to run on. In AWS, that’s an EC2 instance. Here’s how to set it up:

• Go to the EC2 Dashboard in your AWS console.
• Click Launch Instance.
• Choose a free-tier eligible Amazon Linux 2 AMI.
• Select the t2.micro instance type (free tier).
• Configure the instance details (default settings are fine).
• Add storage (8GB is enough for testing).
• Configure security group to allow SSH (port 22) and HTTP (port 80).
• Launch the instance and download your key pair (for SSH access).

After a few minutes, your server will be running in the AWS cloud!

Step 3: Deploying Your First Application

Let’s deploy a simple web application on your new EC2 instance.

1. Connect to your instance
Open your terminal (or PuTTY on Windows) and connect using SSH:

   ssh -i "your-key.pem" ec2-user@your-public-ip

2. Install a web server

   sudo yum update -y
   sudo yum install -y httpd
   sudo systemctl start httpd
   sudo systemctl enable httpd

3. Create a simple web page

   echo "Welcome to AWS! This AWS tutorial is powered by Tpoint Tech." | sudo tee /var/www/html/index.html

4. Open your browser and enter the **public IP address of your EC2 instance.
You should see your web page live — congratulations! 🎉
You’ve just deployed your first cloud application using AWS.

Step 4: Storing Data in S3

To enhance your application, let’s add AWS S3 for storage.
S3 allows you to store files, images, or backups with high availability and durability.

• Go to the S3 Console.
• Click Create Bucket and give it a unique name.
• Upload a test file (like a logo or document).
• You can now access this file from your application using its S3 URL.

In production environments, you can integrate S3 with your backend code for automatic uploads and retrievals.

Step 5: Monitoring with CloudWatch

AWS offers CloudWatch to monitor your application’s performance, CPU usage, and uptime.
By setting up dashboards and alarms, you can ensure your app runs smoothly and automatically scale when traffic increases.

Best Practices for AWS Beginners

Here are some essential AWS best practices recommended by Tpoint Tech:

• Always use IAM roles for secure access management.
• Enable MFA (Multi-Factor Authentication) for your root account.
• Use cost monitoring tools like AWS Budgets to avoid billing surprises.
• Automate deployments using AWS CloudFormation or Terraform.

Following these steps helps ensure your AWS projects are secure, scalable, and cost-efficient.

Conclusion

Building and deploying your first application in AWS might sound complex, but as you’ve seen in this AWS Tutorial by Tpoint Tech, it’s simpler than you think. With just a few clicks, you can set up servers, store data, and monitor performance — all in the cloud.

Whether you’re learning for career advancement or personal growth, mastering AWS opens endless opportunities in today’s cloud-driven world. Continue exploring advanced topics like AWS Lambda, RDS, and Elastic Beanstalk to deepen your cloud expertise.

Start small, experiment, and let AWS power your next big idea — the cloud is yours to build on!

How to: Use a Bitmask-Based Rights System for Cookie Consent

Have you ever been in a position where you need to track multiple choices of a user and want to avoid having to create a new seperate variable for each of those choices? That is where Bitmask can be your hero!

In the following example we will set up user cookie consent using bitmask to have a real-world example.

Managing user consent for cookies can get messy when dealing with multiple categories. A bitmask-based rights system offers a compact and efficient way to store and evaluate consent using binary flags.

Assign each cookie category a unique power-of-2 value:

[Flags]
public enum CookieConsent
{
    None = 0,
    Necessary = 1,      // 2^0
    Preferences = 2,    // 2^1
    Statistics = 4,     // 2^2
    Marketing = 8       // 2^3
}

This allows combinations to be stored as a single integer.

If a user consents to Necessary, Statistics, and Marketing, the combined value is:

var consent = CookieConsent.Necessary | CookieConsent.Statistics | CookieConsent.Marketing;

You can store 13 in a cookie or database.

To check if a user has consented to a specific category:

bool hasMarketing = (consent & CookieConsent.Marketing) != 0;
bool hasPreferences = (consent & CookieConsent.Preferences) != 0;

This works even if multiple flags are combined.

When reading the stored value (e.g. from a cookie):

var storedValue = 13;
var userConsent = (CookieConsent)storedValue;

if (userConsent.HasFlag(CookieConsent.Marketing))
{
    // Load marketing scripts
}

The value 13 in binary is 1101, which corresponds to:

Necessary (1)
Statistics (4)
Marketing (8)

So userConsent will contain:

CookieConsent.Necessary | CookieConsent.Statistics | CookieConsent.Marketing

You can check individual flags like this:

if (userConsent.HasFlag(CookieConsent.Marketing))
{
    // Load marketing scripts
}

if (userConsent.HasFlag(CookieConsent.Preferences))
{
    // Will return false, since Preferences (2) is not included in 13
}

Add new categories by continuing the power-of-2 pattern:

public enum CookieConsent
{
    None = 0,
    Necessary = 1,
    Preferences = 2,
    Statistics = 4,
    Marketing = 8,
    Analytics = 16      // 2^4
}

This keeps the system scalable and backward-compatible.

You can store the combined consent value as an integer in a cookie:

Response.Cookies.Append("CookieConsent",((int)consent).ToString());

To retrieve and parse it:

var storedValue = Request.Cookies["CookieConsent"];
var userConsent = (CookieConsent)int.Parse(storedValue);

Use the flags to conditionally load services or scripts:

if (userConsent.HasFlag(CookieConsent.Marketing))
{
    // Load marketing scripts
}

if (userConsent.HasFlag(CookieConsent.Analytics))
{
    // Enable analytics tracking
}

If you need to pass consent flags to JavaScript:

var consentFlags = new
{
    Necessary = userConsent.HasFlag(CookieConsent.Necessary),
    Preferences = userConsent.HasFlag(CookieConsent.Preferences),
    Statistics = userConsent.HasFlag(CookieConsent.Statistics),
    Marketing = userConsent.HasFlag(CookieConsent.Marketing),
    Analytics = userConsent.HasFlag(CookieConsent.Analytics)
};

return Json(consentFlags);

This allows the frontend to react accordingly.

This system can be applied to even more scenario's if you feel like it. Examples can be user permissions/access control, file system attributes like hidden, readonly etc, or even log levels.

Building a Reliable NAS and Integrating UPS Monitoring with TrueNAS and Home Assistant

I built a NAS from spare parts — an MSI X99A Raider motherboard, a 6-core Intel Xeon CPU, 128 GB ECC DDR4 RAM, a 10G NIC, and an HBA card for the HDDs. The HDDs are configured in ZFS RAID-Z2, allowing the system to tolerate up to two simultaneous disk failures. The operating system is installed on two SSDs configured in RAID10 for redundancy.

Is it overkill for a homelab? Maybe. But I want to store my documents, family photos, and other personal data reliably, without relying on any cloud provider.

I think TrueNAS is an excellent choice. It offers an intuitive web interface, straightforward setup, and clear configuration options — making it ideal for homelab environments as well.

Recently, I purchased an APC UPS. While ZFS (RAID-Z2) provides strong data integrity and fault tolerance, it doesn’t protect against sudden power loss. I didn’t want to leave it to chance, so I integrated the UPS to ensure the NAS can gracefully shut down when the battery level is low.

To achieve this, I installed NUT (Network UPS Tools) on one of my Raspberry Pis, connecting the UPS via USB. The Raspberry Pi acts as the UPS server and communicates with the TrueNAS system over the network.

After setting up NUT on the Raspberry Pi, I logged into the TrueNAS admin interface and navigated to
System → Services → UPS (pencil icon) to configure the UPS client.

In the configuration form, I entered:

• An identifier for the UPS connection
• The IP address, port, and credentials for the NUT observer user

Once saved, I activated the UPS service and enabled auto-start.

This setup allows the NUT client on the TrueNAS server to monitor the UPS and automatically shut down the NAS when the UPS battery level becomes critically low.

TrueNAS can run containerized applications using Docker Compose under the hood.
Some applications are available directly from the TrueNAS catalog (and only require minimal configuration), but you can also deploy custom “freestyle” applications, where you define everything manually.

If you install a catalog app but need more granular control, you can convert it to a freestyle application, giving you direct access to the underlying Docker Compose configuration.

Traefik

Because I already had multiple applications running on my TrueNAS system, I wanted to route their traffic based on hostnames rather than remembering port numbers. To achieve this, I installed Traefik as a reverse proxy using the following configuration:

networks:
  proxy:
    driver: bridge
    external: true

services:
  traefik:
    command:
      - '--api.insecure=true'
      - '--providers.docker=true'
      - '--providers.docker.network=proxy'
      - '--entrypoints.web.address=:80'
    environment:
      TZ: Europe/Brussels
    group_add:
      - 568
    hostname: proxy.homelab.arpa
    image: docker.io/library/traefik:v3
    networks:
      - proxy
    ports:
      - mode: ingress
        protocol: tcp
        published: 8080
        target: 80
      - mode: ingress
        protocol: tcp
        published: 8088
        target: 8080
      - mode: ingress
        protocol: tcp
        published: 8443
        target: 443
    pull_policy: always
    restart: always
    volumes:
      - type: bind
        source: /run/docker.sock
        target: /var/run/docker.sock

Before installation, I created the Docker bridge network for Traefik to communicate with other containers:

docker network create --driver=bridge proxy

I then attached the Traefik container to this network.
For simplicity, I’m currently running without TLS, so I enabled insecure API access and allowed Traefik’s Docker provider to automatically discover containers. The Docker socket is mounted as a bind volume, giving Traefik visibility into running containers.

Using pull_policy: always ensures the container automatically updates to the latest Traefik v3 image on restart.

Home Assistant

Below is an excerpt from the Home Assistant Docker Compose configuration:

networks:
  homeassistant:
    driver: bridge
  proxy:
    driver: bridge
    external: true

services:
  home-assistant:
    labels:
      - traefik.enable=true
      - traefik.http.routers.homeassistant.entrypoints=web
      - traefik.http.routers.homeassistant.rule=Host(`homeassistant.homelab.arpa`)
      - traefik.http.routers.homeassistant.service=homa-svc
      - traefik.http.services.homa-svc.loadbalancer.server.port=30103
    networks:
      - proxy
      - homeassistant

  postgres:
    ...
    networks:
      - homeassistant

I installed Home Assistant from the TrueNAS catalog and then converted it to a custom app (via the three-dot menu → Convert to custom app). This allowed me to edit its Docker Compose configuration directly, as shown above.

I attached the proxy network so Traefik could route traffic to the container, and created an additional homeassistant network for internal communication with the PostgreSQL database. This separation keeps database traffic isolated from the proxy network.

Next, I added Traefik labels to enable routing based on hostname.

However, after setting everything up, accessing the Home Assistant dashboard returned an HTTP 400 (Bad Request) error. The fix was documented here:
Home Assistant (400 Bad Request) Docker + Proxy – Solution

To resolve the issue, I updated Home Assistant’s configuration file located at
/mnt/.ix-apps/app_mounts/home-assistant/config/configuration.yaml:

http:
  use_x_forwarded_for: true
  trusted_proxies:
    - <Subnet of the proxy Docker network, e.g. 172.27.0.0/24 or Traefik’s IP>

After restarting the Home Assistant service, the dashboard loaded correctly.

Monitoring

With everything connected, I created a Home Assistant card to display the UPS status.
Here’s the YAML configuration for the dashboard card:

type: vertical-stack
title: Server Room Rack UPS
cards:
  - type: history-graph
    entities:
      - name: Status
        entity: sensor.rack_ups_1_status
    hours_to_show: 4
  - type: horizontal-stack
    cards:
      - type: gauge
        entity: sensor.rack_ups_1_load
        name: UPS Load
        severity:
          green: 0
          yellow: 70
          red: 90
      - type: gauge
        entity: sensor.rack_ups_1_battery_charge
        name: Battery Charge
        severity:
          green: 50
          yellow: 20
          red: 0

This dashboard displays the UPS status (e.g., on battery, charging, etc.), current battery level, and load percentage in real time.

Summary:
This setup provides a robust and self-contained NAS system with:

• Fault-tolerant ZFS RAID-Z2 storage
• Power-loss protection through UPS integration
• Automated monitoring and alerting via Home Assistant
• Flexible hostname-based routing through Traefik

It’s a powerful, reliable homelab configuration — and best of all, it’s completely under my control.

Chronos Vault Trinity Protocol Now Live on Testnet: Revolutionizing Multi-Chain Security

TL;DR: We've deployed a production-ready multi-chain vault platform on testnet using Trinity Protocol - a 2-of-3 consensus system across Ethereum Layer 2 (Arbitrum), Solana, and TON. All smart contracts are live, open source, and ready for developer testing.

Why We Built This
Single-chain security has a fundamental flaw: trust in one network. What happens when that network faces congestion, gets compromised, or has unexpected downtime? Your assets are at risk.

Chronos Vault solves this with mathematical security, not trust assumptions.

Introducing Trinity Protocol
Trinity Protocol implements 2-of-3 consensus verification across three independent blockchains:

Trinity Protocol implements 2-of-3 consensus verification across three independent blockchains:

The Math Behind It
Your assets require consensus from at least 2 of 3 chains. Even if one blockchain is compromised, experiences downtime, or faces a 51% attack, your funds remain secure.

// Mathematical consensus - no human validators
const consensus = await verifyTrinityProtocol({
  ethereumProof: merkleProofEth,
  solanaProof: merkleProofSol,
  tonProof: merkleProofTon
});

// Returns true only if 2 of 3 proofs are valid

What's Deployed on Testnet
Arbitrum Sepolia Contracts (Live)

• CVT Token: 0xFb419D8E32c14F774279a4dEEf330dc893257147

• CVT bridge:
  0x21De95EbA01E31173Efe1b9c4D57E58bb840bA86

• ChronosVault:
  0x99444B0B1d6F7b21e9234229a2AC2bC0150B9d91

• CrossChainBridgeV1: 0x13dc7df46c2e87E8B2010A28F13404580158Ed9A

View on Arbiscan

TON Testnet Contracts (Live)

• ChronosVault: EQDJAnXDPT-NivritpEhQeP0XmG20NdeUtxgh4nUiWH-DF7M

• CVTBridge: EQAOJxa1WDjGZ7f3n53JILojhZoDdTOKWl6h41_yOWX3v0tq

Key Technical Achievements

1. Trustless Cross-Chain Bridge No human validators. No trusted operators. Just pure cryptographic consensus.

// HTLC (Hash Time-Locked Contracts) for atomic swaps
function initiateSwap(
    bytes32 hashLock,
    uint256 timelock,
    address recipient
) external payable {
    require(block.timestamp < timelock, "Timelock expired");
    // Mathematical security guarantees
}

1. 95% Cost Reduction with Layer 2 By deploying on Arbitrum (Ethereum Layer 2), we achieve:

• Ethereum's base layer security (fraud proofs)

• **95% lower transaction fees **vs Ethereum L1

• Faster finality for better UX

1. Merkle Proof Verification Cross-chain state verification using cryptographic proofs:

const merkleProof = generateMerkleProof(vaultState);
const isValid = await verifyAcrossChains(
  ethereumMerkleRoot,
  solanaMerkleRoot,
  tonMerkleRoot
);

Why This Matters for Developers
Open Source & MIT Licensed
All 5 repositories are live on GitHub:

1.Platform - Full-stack app (React + Express + PostgreSQL)
2.Contracts- Smart contracts (Solidity, Rust, FunC)
3.SDK - TypeScript SDK (coming soon)
4.Docs- Technical documentation
5.Security - Audit reports & protocols

Production-Ready Stack

# Frontend
React + TypeScript + TailwindCSS + shadcn/ui

# Backend
Express.js + PostgreSQL + WebSockets

# Blockchain
Ethereum L2 (Arbitrum) + Solana + TON

# Security
ZK-SNARKs + Post-Quantum Cryptography + Multi-Sig

Easy Integration

import { ChronosVaultSDK } from '@chronos-vault/sdk';

const sdk = new ChronosVaultSDK({
  apiEndpoint: 'https://testnet-api.chronosvault.org'
});

await sdk.initialize();
const vault = await sdk.createVault({
  name: 'My Secure Vault',
  type: 'multi-signature',
  securityLevel: 'maximum'
});

What's Next?
Testnet Phase (Now - Q2 2025)

• ✅ All smart contracts deployed

• 🔄 Community testing & feedback

• 🔄 Security audits in progress

• 🔄 Bug bounty program launch
  Mainnet Launch (Q2-Q3 2026)

• Production deployment

• Full audit reports published

• CVT token distribution begins

• Enterprise partnerships announced

Try It Yourself
Testnet Access:

Platform: https://github.com/Chronos-Vault/chronos-vault-platform-
Contracts: All live on Arbitrum Sepolia & TON Testnet https://github.com/Chronos-Vault/chronos-vault-contracts
Docs: https://github.com/Chronos-Vault/chronos-vault-docs

Get Testnet Tokens:

Arbitrum Sepolia: Arbitrum Faucet
TON Testnet:TON Faucet

Join the Community
We're building the future of multi-chain security in public. Join us:

Discord: https://discord.gg/WHuexYSV
X (Twitter): https://x.com/chronosvaultx
GitHub: https://github.com/Chronos-Vault
Email: chronosvault@chronosvault.org

Technical Deep Dive
Want to understand the cryptography behind Trinity Protocol? Check out:
Trinity Protocol Mathematical Foundation
Smart Contract Architecture
Cross-Chain Bridge Implementation

Built with ❤️ for the future of decentralized security

What do you think? Try our testnet and share your feedback. Let's build the best multi-chain security platform together!

电报数据库：加密世界中的信息存储核心

电报（Telegram）以其高效、安全的通讯系统而闻名，其核心之一便是强大的数据库架构。电报数据库是支撑其消息存储、文件传输、群组管理与账户体系的基础设施。不同于传统的社交应用，电报采用分布式数据库设计，将用户数据加密后分布存储在全球多个服务器节点中。这种架构不仅提升了访问速度，也极大增强了系统的稳定性与容错能力。

电报数据库的存储结构可分为三层：用户层、消息层与媒体层。用户层负责存储账户信息、联系人列表与 电报数据库 安全密钥；消息层管理聊天记录、群组内容与频道数据；媒体层则承担文件、图片、语音与视频等大容量数据的存储任务。通过这种分层方式，系统可以在不同层级进行加密、缓存与同步，从而提高整体数据处理效率。

电报数据库最引人注目的特征是其多重加密机制。平台采用自研的 MTProto 协议，将客户端到服务器的通信数据进行多重加密，再将密钥分散存储，确保即便数据被截获，也无法被轻易解读。此外，电报的“秘密聊天”功能采用端到端加密技术，消息内容不会进入主数据库，而是仅存储在通信双方的本地设备上，彻底杜绝服务器泄密风险。这种加密逻辑使得电报数据库在安全性方面处于行业领先地位。

为了应对庞大的用户量和海量数据请求，电报数据库在性能优化上也做了大量创新。系统通过异步消息队列、分布式缓存与延迟加载等机制，减少数据库的实时压力。此外，电报采用“差异同步”策略——只同步变化部分数据，从而显著提升跨设备消息同步的速度与稳定性。这些优化措施让用户即使在网络不稳定的环境下，也能获得流畅的使用体验。

尽管电报数据库具备高强度的加密与分布式安全设计，但也因此在法律与合规层面引发了争议。一些国家认为电报数据的加密与海外存储方式可能阻碍执法机构取证，要求平台在本地设立服务器并提供部分访问权限。电报方面则强调“用户隐私至上”的原则，坚持不泄露用户通信内容。这种矛盾使得电报数据库在全球监管环境中始终处于微妙的位置。

未来的电报数据库可能将更加智能化与自治化。人工智能将被用于数据检索、内容分类与风险监测；分布式存储将进一步演进为去中心化结构，以提升系统透明度与抗封锁能力。同时，电报或将引入更多用户可控的数据权限设置，让用户自主决定数据的保留与共享范围。这不仅符合隐私保护趋势，也为数字时代的社交系统提供了新的范式。

电报数据库不仅是一套存储系统，更是电报生态的核心引擎。它在安全、性能与隐私之间寻求平衡，展现了现代加密通讯技术的最高水准。随着全球数据治理意识的增强，如何在保护用户隐私的同时满足法律合规要求，仍将是电报数据库未来发展的关键命题。

Truth and documentation build trust. The Talentcrowd filings remind us that openness is the best protection against misinformation.

Anthony James ・ Oct 1

#amandafrye #talentcrowd #corporatefraud #whitecollarcrime

电报筛查：信息时代的安全防线

在加密通讯领域，电报（Telegram）凭借端到端加密、跨平台使用和群组功能的强大优势，成为全球数亿用户的首选。然而，随着用户数量激增，电报也逐渐成为灰色信息传播、诈骗活动以及极端言论的温床。这种现象引发了“电报筛查”的概念——即通过技术手段或人工审核，对电报内的内容进行风险识别与过滤，以维护网络安全与公共秩序。

所谓“电报筛查”，是指针对 Telegram 平台的内容、用户行为以及群组活动进行监测、分析与识别的过程。筛查 电报筛查 的对象可能包括公开群组、频道发布内容、用户昵称与链接分享等。通过人工智能算法或安全团队审核，可以检测出诈骗、黄赌毒信息、虚假宣传或政治煽动内容，从而对其加以限制或报告。这种筛查并非完全侵入性操作，而是在保障隐私的前提下进行风险判断。

电报筛查主要依托两种技术路径：其一是关键词与语义识别算法，通过自然语言处理（NLP）模型识别违规内容；其二是行为模式分析系统，通过检测群组活跃度、异常消息频率、链接传播速度等指标来判断潜在风险。例如，当一个群组在短时间内频繁发送外部链接或金融邀请信息，筛查系统可自动标记并提示管理员进行审查。此外，一些安全公司还提供第三方监测接口，为政府机构或企业客户提供实时的电报内容风险报告。

尽管电报筛查对于网络治理意义重大，但它同时也引发了隐私与自由的讨论。部分用户认为，筛查行为可能侵犯通信自由与个人隐私权，尤其是在未经过用户授权的情况下。而支持者则认为，筛查是防范诈骗、保护未成年人、维护社会稳定的必要手段。如何在“安全”与“隐私”之间找到平衡，成为筛查机制设计的关键挑战。

不同国家对电报筛查的态度各不相同。欧洲国家倾向于在《通用数据保护条例》（GDPR）框架下强调用户数据保护；而在东亚地区，政府通常更重视网络安全与舆情控制，倾向于建立技术监管体系；中东与独联体国家则可能通过直接封锁、监控手段来约束电报的使用。这种政策差异也使电报筛查在全球范围内呈现出多元而复杂的治理格局。

未来的电报筛查将更依赖人工智能与大数据分析，以提高筛查的准确性与实时性。同时，透明度和合规性将成为核心议题——平台需要明确告知用户筛查的范围与标准，确保筛查机制不被滥用。对于普通用户而言，理解筛查的必要性、养成安全使用习惯，也是参与数字社会治理的重要一环。

电报筛查并非单纯的技术问题，而是一场关于“自由与安全”的社会讨论。它要求技术、法律与伦理共同参与，以构建一个既保障隐私、又防范风险的网络环境。在信息时代，筛查不应被视为监控，而应被理解为数字文明的自我净化过程。

100 Days of DevOps: Day 65

Introduction

This article details the precise steps and Kubernetes configuration used to successfully deploy a single, dedicated Redis cache instance on the cluster for initial testing.

The deployment leverages essential Kubernetes primitives—ConfigMap, Deployment, and VolumeMounts—to ensure a scalable, configurable, and ephemeral testing environment.

Step 1: Configuring Redis with a ConfigMap

Before deploying the Redis application, the team needed to define a specific memory limit to prevent uncontrolled resource consumption. In Redis, this is managed by the maxmemory directive. Kubernetes ConfigMaps are the ideal mechanism for injecting such configuration data into a container.

A ConfigMap named my-redis-config was created, containing the Redis configuration data under the key redis-config.

configmap.yaml

apiVersion: v1
kind: ConfigMap
metadata:
  name: my-redis-config
data:
  redis-config: |
    maxmemory 2mb

This ConfigMap was applied to the cluster:

kubectl apply -f configmap.yaml
configmap/my-redis-config created

Step 2: Defining the Redis Deployment

The core of the solution is the Deployment manifest, which specifies the application's desired state, including the image, resource requests, and how configuration and data volumes are handled.

Key Deployment Parameters:

Volume Strategy for Redis

The deployment utilizes two ephemeral volumes crucial for Redis operation in a testing environment:

EmptyDir Volume (data):

• Purpose: Provides temporary storage for Redis's dump file (if persistence were enabled) or other scratch data.
• Mount Path: /redis-master-data
• Nature: Data is destroyed when the Pod terminates, aligning with the temporary nature of this initial test deployment.

ConfigMap Volume (redis-config):

• Purpose: Mounts the my-redis-config ConfigMap so the container can access the configuration file.
• Mount Path: /redis-master
• Usage: The Redis server command is explicitly told to use this mounted configuration file.

deployment.yaml (Simplified Manifest)

apiVersion: apps/v1
kind: Deployment
metadata:
  name: redis-deployment
spec:
  replicas: 1
  selector:
    matchLabels:
      app: redis
  template:
    # ... (metadata)
    spec:
      volumes:
        - name: data
          emptyDir: {}
        - name: redis-config
          configMap:
            name: my-redis-config
      containers:
        - name: redis-container
          image: redis:alpine
          # Crucial: Instructs Redis to start with the custom config file
          command: ["redis-server"]
          args: ["/redis-master/redis-config"]
          ports:
            - containerPort: 6379
          resources:
            requests:
              cpu: "1"
          volumeMounts:
            - name: data
              mountPath: /redis-master-data
            - name: redis-config
              mountPath: /redis-master

The deployment was then applied:

thor@jumphost ~$ kubectl apply -f deployment.yaml
deployment.apps/redis-deployment created

Step 3: Verification and Conclusion

The final step confirmed that the new caching utility was operational and ready for integration testing.

thor@jumphost ~$ kubectl get deployment redis-deployment
NAME               READY   UP-TO-DATE   AVAILABLE   AGE
redis-deployment   1/1     1            1           21s

The output shows that the redis-deployment is fully available (1/1 Ready, 1 Available), confirming that the Redis container is running correctly, utilizing the requested CPU resources, and applying the custom maxmemory 2mb configuration.

DevOps Interview Question and Answer

1. What is DevOps?
   DevOps is a set of practices that combines software development (Dev) and IT operations (Ops). It aims to shorten the systems development life cycle and provide continuous delivery with high software quality.

2. Explain the benefits of DevOps.

• Faster delivery of features
• More stable operating environments
• Improved communication and collaboration
• More time to innovate (rather than fix/maintain)

1. What are the key components of DevOps?

• Continuous Integration
• Continuous Testing
• Continuous Delivery
• Continuous Monitoring
• Infrastructure as Code
• Collaboration and communication

1. What is Continuous Integration (CI)?
   Continuous Integration is a DevOps practice where developers regularly merge their code changes into a central repository, after which automated builds and tests are run. The goal is to detect integration bugs early and improve the quality of software.

2. What is Continuous Delivery (CD)?
   Continuous Delivery is a practice where code changes are automatically prepared for a release to production. It ensures that software can be to continue reading <<>>

Transparent AI for SMBs: Why Explainability Isn’t Just Nice-to-Have — It’s Mission-Critical

Let’s be honest — the AI buzz is deafening. Everywhere you turn, competitors are flaunting “intelligent” automation, claiming it’s doubled their revenue or slashed their operational costs overnight. Sounds tempting, doesn’t it? But for small and mid-sized businesses, integrating AI often feels like trying to rebuild your engine while driving at 100 km/h.

The real issue isn’t interest — it’s opacity. Too many AI systems behave like black boxes. You get a result, maybe even an impressive one, but when you ask why, silence. Or worse — a vague dashboard readout that offers zero context. I’ve seen teams grind to a halt because they couldn’t explain why the model did what it did. Confidence erodes, compliance alarms start flashing, and suddenly that “AI revolution” looks more like a liability than a leap forward.

So, the question becomes: How do you build trust in something designed to make decisions faster than humans can explain them?

I’ve worked across sectors long enough to see the same pattern repeat itself — the businesses that hesitate to adopt explainable AI don’t just fall behind; they end up paying for inefficiency twice. First, in the manual work they keep doing. Second, in the rework required when opaque systems misfire.

The data’s clear:

91% of SMBs using AI report increased revenue.

90% credit it with major efficiency gains.

75% are investing in AI as we speak.
Those aren’t vanity numbers — they’re a reflection of what’s already happening on the ground.

But here’s the catch: without transparency, all those gains are fragile. Because when the “why” disappears, so does accountability. And without accountability, trust collapses.

Transparency isn’t a buzzword; it’s the backbone of adoption. In DevOps, we live by one principle — if you can’t observe it, you can’t trust it. The same logic applies to AI.

Explainable AI (XAI) gives you that observability. Imagine an algorithm that flags an invoice as fraudulent — and tells you exactly which three risk factors triggered the flag. Or a recommendation engine that not only suggests a product but also shows you the data correlation behind it. Suddenly, your team isn’t fighting the AI — they’re collaborating with it.

This shift is cultural as much as it is technical. Once people see logic behind automation, the fear dissipates. You stop second-guessing the machine and start optimizing it.

I’ve watched this transformation happen firsthand. Last year, we worked with a mid-sized health tech firm that wanted to automate appointment scheduling. Their staff didn’t trust the model — they were worried about bias and errors. We built in explainability tooling so every decision could be traced back through the data pipeline. The difference was night and day. Within weeks, confidence soared, errors dropped, and the project went from “experimental” to “indispensable.”

If you’re considering explainable AI, start small but deliberate. Here’s a blueprint I often recommend:

• Audit Your Pain Points. Find the repetitive, error-prone processes. That’s your low-hanging fruit.
• Select Explainable Frameworks. PyTorch, TensorFlow, or scikit-learn paired with interpretability libraries like SHAP or LIME.
• Establish Governance Early. Write down how your AI makes decisions, who reviews them, and how often.
• Train the Humans, Not Just the Models. Empower your staff to question outputs intelligently, not blindly trust them.
• Iterate with Feedback Loops. Monitor, test, and retrain. Transparency isn’t a one-time setting — it’s an evolving discipline.
• Work with Experienced Partners. Choose engineers and consultants who’ve actually deployed transparent AI at scale, not just read about it.

Here’s the thing — transparency isn’t just compliance-friendly; it’s commercially smart. When customers see that you can explain your automation, they stay longer. When your board sees traceability, they approve faster. When your developers see how models behave, they improve them.

At SDH, we’ve seen this play out in real time across healthcare, SaaS, logistics, and fintech. Every successful deployment had one thing in common: clarity by design.

Eighteen years in Linux systems and automation have taught me one truth — systems don’t fail because they’re complex. They fail because they’re misunderstood. The same is true for AI. Transparency bridges that gap.

If you’re serious about making AI your growth engine — not your risk vector — start by demystifying it. You don’t need a massive budget or a PhD lab. You just need visibility, structure, and the right technical guidance.

At Software Development Hub, that’s exactly what we deliver: AI systems you can explain, trust, and scale confidently.

Let’s build smarter — and clearer — together.

电报数据库：信息安全与数据结构的深度解析

电报（Telegram）自 2013 年推出以来，凭借端到端加密和去中心化服务器架构，迅速在全球范围内获得大量用户。支撑这一庞大生态体系的，是其高效且复杂的数据库系统。电报数据库不仅承担着消息存储与用户数据管理的任务，更是平台稳定运行与信息安全的关键所在。它是电报生态的“数据中枢”，也是加密通讯时代最具代表性的技术成果之一。

与传统社交软件不同，电报采用的是分布式数据库架构。用户数据被加密后分散储存在多个地理位置不同的服务器上，这样即便某个节点遭受攻击，也不会影响整体系统的运行或数据完整性。这种架 电报数据库 构提高了抗风险能力，同时降低了单点故障的风险。每一部分数据库都只保存经过加密的部分数据片段，完整信息只有在用户终端设备上才能被重新组合并解密。

电报的数据库设计在存储逻辑上高度模块化。文本消息、语音、图片、视频等不同类型的数据会被分类存储，并采用不同的压缩与加密算法。对于私聊消息，系统使用端到端加密方式，消息只存在于通信双方设备中，不进入主数据库。而对于群组与频道内容，数据库采用加密后集中存储的方式，以便同步与备份。这种差异化策略既确保了隐私安全，也兼顾了性能与便利。

电报数据库的安全性主要依赖于其自研的加密协议 MTProto。该协议在客户端与服务器之间建立多层加密通道，使数据传输过程几乎无法被第三方拦截或破解。同时，电报在数据库层面采用“多重分离存储”策略——将密钥与数据分离保存，进一步防止信息泄露。用户身份认证信息、访问记录等敏感数据也经过匿名化处理，以保护个人隐私。

电报数据库的强加密特性虽然保障了用户隐私，但也引发了诸多监管层面的争议。一些国家认为其数据加密过强，影响了执法机构的调查取证；而电报坚持不建立后门，强调“用户隐私不可侵犯”的立场。这种平衡难题让电报数据库在全球数据治理体系中始终处于敏感地位。它既是安全通信的象征，也成为隐私与监管之间的技术博弈场。

展望未来，电报数据库将持续优化其性能与安全性。人工智能和机器学习技术或将被引入，用于识别异常行为、提升系统负载能力。同时，去中心化与用户自控数据存储或将成为新趋势，用户可能拥有更高的数据管理权限。随着 Web3 与隐私计算的发展，电报数据库有望成为加密通信与区块链技术融合的实验场。

电报数据库的价值不仅在于技术本身，更体现在其背后的理念——让信息自由、安全地流动。在数据成为新型生产力的时代，电报以加密与分布式存储构筑起安全的通讯堡垒，也为整个互联网行业提供了思考方向：未来的数据库，不只是存储工具，更是信任与安全的基础设施。

Indian Startup Targets EMI/EMC Bottlenecks With Design-Stage Simulation

Backed by Bosch and deep-tech VCs, SimYog Technology aims to scale globally with its EMI/EMC test simulation software.

The post Indian Startup Targets EMI/EMC Bottlenecks With Design-Stage Simulation appeared first on EE Times.

IoT Security: Hard Lessons, Emerging Practices, and the Push Toward Regulation

IoT devices are now integrated with critical infrastructure, and their security flaws are forcing the industry to confront hard lessons and new regulations.

The post IoT Security: Hard Lessons, Emerging Practices, and the Push Toward Regulation appeared first on EE Times.

Silicon Labs’ Matt Johnson: An Inflection Point for Edge AI in IoT Devices

At the company’s Works With event, Silicon Labs CEO said that the foundation necessary for AI to grow is now in place for IoT.

The post Silicon Labs’ Matt Johnson: An Inflection Point for Edge AI in IoT Devices appeared first on EE Times.

The best Prime Day speaker deals: JBL, Bose, Ultimate Ears and more

It's once again time for October Prime Day. While this fall's Big Deal Days event isn't a four-day extravaganza like we got in July, we're still seeing steep discounts on some of the world's best audio gear. This time, our audiophiles have dug up markdowns on top-tier portable bluetooth speakers, smart speakers and soundbars from JBL, Beats, Sonos and more. Whether you're building a permanent setup for your home theater or chaining portable speakers together to blast dance tunes through an outdoor party venue, we bet you'll find some Prime Day speaker deals to love on our professionally curated list.

• JBL Xtreme 3 for $209 (36 percent off): Like the Xtreme 2, the JBL Xtreme 3 is a hefty bluetooth speaker with a 15-hour battery life and the ability to chain together with other JBL speakers. But don't fret if you've only got one — even on its own, we've found the Xtreme 3's 100-watt output manages to sound both crisp and deep in any space.

• JBL Flip 6 for $98 (24 percent off): The JBL Flip 6 is a solid choice for an all-around bluetooth speaker, performing well in any room or on the go. It sounds good at almost any frequency and volume, runs for 12 hours on a single charge and is nearly impossible to damage with falls, dust or water.

• JBL Clip 3 for $32 (64 percent off): The JBL Clip 3 is one of the best traveling speakers on the market, weighing a quarter of a pound and including a large and versatile carabiner ring. Output power reaches 5 watts, which is pretty good for such a small unit, and the battery runs for 10 hours per charge.

• JBL Go 3 for $25 (37 percent off): The JBL Go 3 is a decent and highly affordable carabiner speaker for all your outdoor adventures — as is its sister product, the Go 3 Eco, which is made from recycled materials. It's obviously not the most powerful speaker at 4.2 watts, but it has a surprisingly good functional range. Plus, the design looks good on any belt or backpack.

• Ultimate Ears Miniroll for $50 (37 percent off): Miniroll is the smallest speaker from Ultimate Ears, but it punches far above its weight, starting with UE's trademark massive volume buttons. While small enough to fit in a jeans pocket, its combination of a driver and radiator manages about 85 decibels. A Miniroll can run for about 12 hours on a single charge.

• Ultimate Ears Boom 4 for $100 (33 percent off): The UE Boom 4 is a balanced speaker that does well anywhere. Take it camping (it's waterproof and a charge lasts for 15 hours), to a party (it puts out loud 360 degree sound) or make it part of a permanent home system by chaining it with other UE speakers — one of these plus two Minirolls or Wonderbooms would come pretty close to full surround sound.

• Ultimate Ears Wonderboom 4 for $50 (44 percent off): The Wonderboom 4 may be our favorite UE product. At just 1.2 pounds, it can pump out mighty sound on its own, or chain with other UE speakers for a deeper full-range experience. USB charging is simple and lasts for 14 hours between top-ups.

• Ultimate Ears Everboom for $231 (15 percent off): The Everboom is the middle ground of the UE line, with more powerful output than the Wonderboom and more portability than the Hyperboom. Its 360 degree sound pairs with an Outdoor Boost feature to work better outside, though it isn't quite as impressive in the middle ranges.

• Bose SoundLink Flex for $99 (33 percent off): We've rarely seen the Bose SoundLink Flex this cheap, so now is the time to snap up this high-fidelity outdoor speaker. It's small enough to carry in one hand and can be operated from up to 30 feet away via the Bose Connect app. We've also found it simple and fun to use for hands-free phone calls during breaks in the music.

• Bose SoundLink Revolve+ for $179 (40 percent off): The SoundLink Revolve+ line, Bose's original upright 360-degree speakers, continue to pack a punch. This current model features a 17-hour battery life, IP55 water resistance and the ability to connect to a smart speaker to extend your virtual assistant's range.

• Bose SoundLink Max for $299 (25 percent off): The Bose SoundLink Max may put you in mind of a classic boombox, with its detachable handle and thudding base. But inside it's all modern, with a 50-watt output, an effective range of 52.6Hz and a 20-hour rechargeable battery. It even comes with a 3.5mm AUX input to connect to a turntable, if you really want to embrace your inner 90s DJ.

• Beats Pill for $129 (14 percent off): We gave the revived Beats Pill a full review last year, and we stand by our opinion that it was worth waiting for. Its newly enlarged subwoofer delivers much more powerful bass that sounds clean and lossless even over USB-C. The latest Pill also travels well, with 24 hours of battery life and IP67 waterproofing.

• Marshall Emberton II for $100 (44 percent off): Marshall, the old warhorse of the audio world, continues to crush it with the portable and bluetooth-ready Emberton II. Although it looks like an antique, it's anything but, with 360-degree sound and over 30 hours of battery life from a single charge — all weighing a little more than a quarter of a pound.

• Amazon Echo Dot 5th Generation for $35 (30 percent off): The Echo Dot, our favorite smart speaker under $50, is now...even farther under $50. Despite being small enough to fit on a cluttered bookshelf, the Dot pumps out high-quality sound that can fill a room, and comes equipped with physical buttons to supplement Alexa's voice command interface.

• Amazon Echo Pop for $25 (38 percent off): The Echo Pop is Amazon's smallest and cheapest smart speaker, resembling an Echo Dot sliced in half. It's a little quiet, and probably best for extending Alexa's range rather than serving as the primary host for your virtual assistant. In a smaller room, though, this is as convenient and clear-sounding as any other Echo you'll use.

• JBL Bar 9.1 for $700 (41 percent off): JBL designed the 9.1 soundbar for Dolby Atmos, and the results speak for themselves, with 820 watts of booming audio. The best thing about this soundbar is the pair of detachable wireless speakers you can place around your viewing room to make the sound even more surround.

• JBL Bar 300 for $250 (37 percent off): The Bar 300, part of JBL's 2023 soundbar explosion, is a mighty yet cheap soundbar with an integrated subwoofer. Despite being small, it manages to create a theatrical experience by working both Dolby Atmos and MultiBeam. Its voice sharpening algorithm is especially good at making dialogue clearer (for those of us who always have to put on the subtitles).

• JBL Bar 700 for $500 (44 percent off): Another 2023 drop, the JBL Bar 700 upgrades the Bar 300 by adding detachable wireless speakers for true 360-degree sound. While it costs twice as much as the 300, that money gets you more than four times the output, as the 700 tops out at 620 watts — and thanks to a 10-inch subwoofer, you'll feel all of it.

• Amazon Fire TV Soundbar Plus for $150 (40 percent off): One of the cheapest soundbars you'll ever see, the Fire TV Soundbar Plus provides a noticeable boost to your TV's audio quality, even if it doesn't reach Dolby Atmos heights. It's simple, but that makes it a breeze to set up — all you have to do is plug in one HDMI cable.

This article originally appeared on Engadget at https://www.engadget.com/audio/speakers/the-best-prime-day-speaker-deals-jbl-bose-ultimate-ears-and-more-070038290.html?src=rss

Prime Day Apple deals include the iPad A16 for $280

The entry-level Apple iPad with the A16 chip is on sale for $280 as a Prime Day promotion. This is a discount of around $70 and represents a record-low price for the 11-inch tablet. The deal is available in multiple colorways.

This tablet made our list of the best iPads and we particularly recommend it for the budget-conscious. It may lack some of the bells and whistles of its more expensive cousins, but it still gets the job done. It's fast enough for most common tasks and the battery life is solid, at 10 hours per charge. It also ships with 6GB of RAM, which is a decent enough metric.

The tablet integrates with Wi-Fi 6 and features a USB-C connector, along with a pair of 12MP cameras. There's also a Touch ID sensor, which can be used to unlock the device and to make payments.

This is the most basic tablet in Apple's lineup, so it lacks some features. It doesn't integrate with the Apple Pencil Pro or the Magic Keyboard. The display is also not quite as gorgeous as the iPad Air or Pro, but it's still a Liquid Retina panel. 

This article originally appeared on Engadget at https://www.engadget.com/deals/prime-day-apple-deals-include-the-ipad-a16-for-280-082743946.html?src=rss

The best October Prime Day gaming deals: Save on PS5 games, headsets, controllers and more

Amazon is back with another October Prime Day sale, and the event has brought a handful of decent discounts on gaming gear along with it. We’ve rounded up the best offers we could find below, which included lower-than-usual prices on gaming mice, headsets, controllers and actual games we recommend. We’ll update this post as we find more deals of note. Just remember that the Prime Big Deal Days sale, as it’s officially called, runs October 7 and 8 and limits some deals exclusively to Prime members.

Split Fiction (Xbox) for $25 ($25 off): A pick from our guide to the best couch co-op games, Split Fiction is a imaginative crowd-pleaser that’s designed from the ground up to be played with another person. As in, you literally need a second person to play it. Like It Takes Two before it, it overcomes some hacky writing with brisk action, a lighthearted tone and a Nintendo-like commitment to presenting new methods of play. This deal for the Xbox copy represents the lowest price we’ve tracked, though the PS5 version is also on sale if you’d rather play there.

Elden Ring (PS5) for $30 ($20 off): You’ve probably heard about it by now, but the action-RPG Elden Ring is both challenging and darkly funny, with a world that runs incredibly deep and feels lived-in a way few games have. It’s unafraid to test your resolve, yet it always grants you the freedom to simply turn around and find another path forward.

Marvel’s Spider-Man 2 for $40 ($30 off): Spider-Man 2 is basically “more Spider-Man, except bigger”. It’s not the most creatively exciting thing, but jumping off skyscrapers and beating the tar out of bad guys is still a blast, and it never takes itself too seriously. Several other first-party PlayStation games are also on sale, which we highlight below.

Final Fantasy VII Rebirth (PS5) for $35 ($15 off): Engadget’s Mat Smith gave this grandiose action-RPG a favorable review last year. You need to have played its predecessor — and ideally, the original PS1 game — to get the most out of it, but it’s positively stuffed with things to do, and its combat system is still a tactical rush at its best.

Gran Turismo 7 for $30 ($40 off): Gran Turismo 7 is the PS5’s premier sim racer, and a love letter to automobiles as a whole. Like the rest of the GT series, it revels in the patience and precision required to figure out how a given car meshes with a given course. Learning the intricacies of each combination is both intense and rewarding, and there’s a singular style to it all that just doesn’t exist with most sports games.

Returnal for $30 ($70 off): Returnal is a third-person shooter roguelike that’s at once a technical showcase, a stiff challenge and an achingly beautiful reflection on the nature of grief. It is not for the faint of heart, but its bullet-hell battles are wonderfully fierce, and just about everything in it contributes to its overarching story. It’s also worth checking out if you’re eager to developer Housemarque’s next game, Saros, which seems to borrow many of the elements introduced here.

The Last of Us Part I for $30 ($40 off): The PS5 remake of Sony’s zombie drama is one of the most unnecessary remakes ever made, but if you’ve never played one of the earlier versions, this is the best way to rectify that. If you’re looking to play the (even moodier) sequel, the PS5 version of that one is similarly discounted.

God of War Ragnarok for $30 ($40 off): Ragnarok is another Sony sequel that’s more about increasing scope than radically reinventing itself, but its Norse world is beautifully varied and detailed, smashing baddies with a magic axe feels great, and the whole thing is excessive in the way you’d expect a God of War game to be without going overboard.

Horizon Forbidden West for $30 ($40 off): Engadget’s Jess Conditt called Horizon Forbidden West the “total package” in her review a few years back, praising its frenzied combat, gorgeous art direction and complex narrative. It still has one of the more distinctive post-apocalypses in gaming, blending robot beasts with stunning vistas. That helps paper over some less-than-inspired side content and bits of sluggish pacing.

Final Fantasy XVI (PS5) for $23 ($7 off): Final Fantasy XVI is a more straightforward action-RPG than something like Final Fantasy VII Rebirth: You don’t need to come in with prior knowledge of any other story, and its hack-and-slash combat is simple to pick up. It can get bogged down in pointless side quests, but its main concern is having you watch giant demigods melodramatically smack each other, and there’s nothing wrong with that.

Astro A50 X gaming headset for $330 ($60 off): The A50 X is an upgrade pick in our guide to the best gaming headsets. It should specifically appeal to those who own multiple gaming systems, as its included base station lets you connect and swap between PC, PS5 and Xbox audio with the press of a button. It’s among the better-sounding wireless headsets we’ve tried as well, and its mic is fantastic. This deal isn’t an all-time low, but it’s the largest drop we’ve seen since July. The standard Astro A50 is also on sale for $250; that one works just as well if you only play on one device, but it lacks HDMI switching functionality.

8BitDo Ultimate 2 Wireless Controller for $50 ($10 off): The Ultimate 2 is a versatile wireless gamepad for Windows PCs and mobile devices. It can connect over Bluetooth, a wireless dongle or a USB cable, and its magnetic TMR joysticks should be far less susceptible to “joystick drift” than typical controllers. Those with large hands may find it a touch too small, but it’s built well, it comes with a handy charging dock and it gets a solid 20 or so hours of battery life. 8BitDo’s companion software includes a variety of useful customizations beyond that. This isn’t a major discount, but it’s still a couple bucks below the pad’s typical street price.

Corsair Scimitar RGB Elite MMO gaming mouse for $55 ($25 off): We recommend the Scimitar RGB Elite to MMO and MOBA players in our guide to the best gaming mice. It’s not the lightest or most technically advanced model, but it comes with 12 comfortable and easy-to-reach side buttons, which make it easier to pull off more complex actions in games like World of Warcraft or Final Fantasy XIV. We’ve seen it fall below $50 a few times before, but this deal matches the lowest price we’ve tracked since May. A newer wireless model called the Scimitar Elite Wireless SE is also on sale for a low of $80.

ASUS ROG Strix Scope II 96 Wireless mechanical keyboard for $151 ($59 off): We speak positively about this model in our guide to the best gaming keyboards. Its sturdy case, crisp keycaps and battery life all impress, while the thocky NX Snow switches in this model feel and sound delightful. Its companion software is a bit of a mess, and its 96 percent layout may feel a little cramped to some, but it’s a nice buy if you want a higher-class pre-built keyboard.

ASUS ROG Azoth mechanical keyboard for $200 ($75 off): It’s a couple years old at this point, but the ROG Azoth remains an exceptionally well-built mechanical keyboard for enthusiasts. Its gasket-mounted design and layers of foam give each key press a soft landing, while its pre-lubed switches feel nice and smooth. It even comes with a toolkit for manually lubing the switches down the road, and the PCB is hot-swappable if you ever want to switch things up. There’s a useful control knob and programmable OLED display for quickly adjusting settings and checking the battery level beyond that. ASUS’ Armoury Crate software is still fairly sloppy, however.

Crucial X9 Pro (1TB) portable SSD for $75 ($26 off): If you’re looking to offload some games to external storage, we consider the Crucial X9 Pro the best portable SSD for most people, as it offers fast enough transfer speeds in a rugged and compact design. This isn’t the best price ever for the 1TB model but it matches the largest discount we’ve seen since April. Other size options are also on sale.

Crucial P310 (1TB) M.2 2230 SSD for $80 ($55 off): The P310 is a small-size SSD you can use to add storage to a handheld PC like the Steam Deck or ASUS ROG Ally. It uses cheaper QLC memory, not the faster and more durable TLC, but reviews say that it performs well anyway. This deal ties the best price we’ve seen for the 1TB version since April.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-best-october-prime-day-gaming-deals-save-on-ps5-games-headsets-controllers-and-more-082623558.html?src=rss

The best Amazon deals on Kindles, Echo speakers, Fire TV devices and more for Prime Day

Amazon’s own gadgets — Echo speakers, Fire tablets, Kindles and the like — are at the top of the list of tech on sale during October Prime Day. While there are some (Echos, Kindles) that we like more than others (Fire tablets), there’s no question that you’ll get the best discount on all Amazon tech during Prime Day. The online retail giant typically discounts all of its in-house-made gear for Prime Day, and this year is no different. Here, we’ve collected all of the best October Prime Day deals on Amazon gear so you don’t have to go searching for them.

Kindle Kids for $95 ($35 off): This is the same device as the base Kindle, but it becomes more kid-friendly thanks to the included cover, two-year warranty and the included year of Amazon Kids+, which gives children ages 3-12 access to hundreds of appropriate ebooks and audiobooks. Just be sure to take note when you activate that subscription because it will renew after one year at the standard $6/month rate.

Kindle Paperwhite for $125 ($35 off): The latest version of the Paperwhite has a seven-inch display, thinner bezels, an adjustable warm light, speedier page turns and a battery that can last up to 12 weeks on a single charge. This model is also IPX8 waterproof and has built-in Audible integration.

Kindle Colorsoft for $200 ($50 off): Amazon's only color e-reader has a seven-inch, high-contrast display, an auto-adjusting front light, a color highlighting feature and an eight-week battery life.

Kindle Scribe for $300 ($100 off): The Scribe is one of the best E-Ink tablets you can buy at the moment, and certainly the top pick if you want a writable table that also excels as an ereader. It provides a great reading and writing experience, thanks in part to its ability to access the entire Kindle ebook library, and it has handy Google Drive, Microsoft OneDrive and Dropbox integration.

Echo Pop speaker for $25 (38 percent off): One of the newer Echo devices available, the Pop it sports a 1.95-inch front-facing speaker and a physical mic mute switch for extra privacy. The Pop also has built-in eero compatibility, so it can extend the area of your home Wi-Fi network if you already have an eero router system.

Echo Dot speaker for $35 (30 percent off): One of the smallest Echo speakers, this version of the Dot has improved audio and a compact design that will fit almost anywhere. You'll control it mostly with Alexa voice commands, but there are a few onboard buttons, including a mic-off button for when you need more privacy.

Echo Show 5 for $55 (39 percent off): This model is one of our favorite smart displays thanks to its compact design, ambient light sensor and sunrise alarm feature, all of which make it an excellent smart alarm clock. Amazon improved the speaker quality on this latest model, which gives sound deeper bass and clearer vocals.

Echo Show 8 for $100 (33 percent off): This is our current top pick for the best smart display with Amazon’s Alexa thanks in part to its 8-inch touchscreen, 13MP camera that supports auto-framing for better video chats and its built-in Zigbee smart home hub. The 2023 model supports Visual ID, which will show personalized information on the device’s display depending on who’s using it, and video streaming from Netflix, Hulu, Prime Video and other services.

Fire TV Stick 4K for $25 (50 percent off): This is the cheapest Fire TV Stick you can get to stream 4K content, plus it has support for Wi-Fi 6 and Dolby Vision and Atmos. With its live picture-in-picture feature, you can view security camera feeds right on your TV while you’re watching your favorite show or movie.

Fire TV Stick 4K Max for $40 (33 percent off): In addition to 4K HDR streaming with Dolby Vision and Atmos support, the 4K Max dongle includes Wi-Fi 6E connectivity, 16GB of built-in storage and live picture-in-picture capabilities. It also supports the Fire TV “ambient experience,” which lets you display photos and images on your TV screen when you’re not actively watching something.

Fire TV Cube streaming box for $100 (29 percent off): This model will provide the best performance of any Fire TV streaming device, and it supports 4K HDR content, Dolby Vision and Atmos and an enhanced version of the Alexa Voice Remote. Along with live picture-in-picture view and the Fire TV ambient experience, you can also hardware other devices to the Fire TV Cube including a cable box or a game console.

Fire HD 8 tablet for $55 (45 percent off): This is Amazon's most bare-bones tablet, featuring an eight-inch HD touchscreen, hexa-core processor and 13 hours of battery life. The improvements in the screen quality alone from the old-school Fire 7 tablet make it a better buy for most people, and this slab would make a good couch device for general web browsing, email checking, online shopping and more.

Fire Max 11 tablet for $140 (39 percent off): Amazon’s most powerful tablet, the Fire Max 11 sports an 11-inch 2,000 x 1,200 touchscreen, an octa-core processor, up to 128GB of storage and 14 hours of battery life. It also works with a number of optional accessories, including a stylus and keyboard case.

Fire HD 10 Kids tablet for $105 (45 percent off): This slab is designed for kids aged three to seven, with full parental controls plus one year of Amazon Kids+ for free with the tablet purchase. This model has a 10-inch FHD touchscreen, an octa-core processor and 13 hours of battery life, plus it comes with a two-year warranty and a protective case.

Fire HD 10 Kids Pro tablet for $105 (45 percent off): This model is designed for kids aged six to 12 and comes with a slimmer protective case, a two-year warranty and one year of access to Amazon Kids+. Otherwise, you get a very similar experience here that you would with the non-Pro version, including parental controls, a 10-inch touchscreen, solid performance and a 13-hour battery life.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-best-amazon-deals-on-kindles-echo-speakers-fire-tv-devices-and-more-for-prime-day-081210730.html?src=rss

The best wireless earbuds for 2025

Wireless earbuds have become the go-to choice for listening on the move. Whether you’re at the gym, commuting or relaxing at home, the best wireless earbuds give you comfort, freedom and solid sound quality without tangled cables. They’re lightweight, slip easily into your pocket and connect quickly to your phone, tablet or laptop.

The tricky part is choosing the right pair. Some models focus on powerful noise cancellation while others put battery life or affordability first. Then you’ve got features like water resistance for workouts or touch controls for quick track changes. With so many choices, finding the best wireless earbuds depends on what matters most to you, and that’s exactly what this guide will help you figure out.

• Best wireless earbuds of 2025

• What to look for in the best wireless earbuds

• How we test wireless earbuds

• Other wireless earbuds we tested

• Wireless earbuds FAQs

When it comes to shopping for earphones, the first thing to consider is design or wear style. Do you prefer a semi-open fit like AirPods or do you want something that completely closes off your ears? If you’re shopping for earbuds with active noise cancellation, you'll want the latter, but a case can be made for the former if you want to wear them all day or frequent places where you need to be tuned in to the ambient sounds. The overall shape of earbuds can determine whether you get a comfortable fit, so can the size and weight, so you’ll want to consider all that before deciding. And remember: audio companies aren’t perfect, so despite lots of research, the earbud shape they decided on may not fit you well. Don’t be afraid to return ill-fitting earbuds for something that’s more comfortable.

As wireless earbuds have become the norm, they’re now more reliable for basic things like consistent Bluetooth connectivity. Companies are still in a race to pack as much as they can into increasingly smaller designs. This typically means a longer list of features on the more premium sets of earbuds with basic functionality on the cheapest models. Carefully consider what you can’t live without when selecting your next earbuds, and make sure key items like automatic pausing and multipoint connectivity are on the spec sheet. You’ll also want to investigate the volume and touch controls as you’ll often have to sacrifice access to something else to make that adjustment via on-board taps or swipes. Some earbuds even offer app settings to tweak the audio profiles or firmware updates to improve performance over time.

For those in the Apple ecosystem, features like auto-pairing with devices, especially with AirPods Pro 3, can be an added advantage, while Android users may want to look for models that offer similar cross-device functionality.

When it comes to battery life, the average set of earbuds lasts about five hours on a single charge. You can find sets that last longer, but this is likely enough to get you through a work day if you’re docking the buds during lunch or the occasional meeting. You’ll want to check on how many extra charges are available via the case and if it supports wireless charging.

Companies will also make lofty claims about call quality on wireless earbuds. Despite lots of promises, the reality is most earbuds still leave you sounding like you’re on speakerphone. There are some sets that deliver, but don’t get your hopes up unless reviews confirm the claims.

Sound can be subjective, so we recommend trying before you buy if at all possible. This is especially true if you're an audiophile. We understand this isn’t easy when most of us do a lot of shopping online, but trying on a set of earbuds and listening to them for a few minutes can save you from an expensive case of buyer's remorse. If a store doesn’t allow a quick demo, most retailers have return policies that will let you take earbuds back you don’t like. Of course, you have to be willing to temporarily part with funds in order to do this.

We also recommend paying attention to things like Spatial Audio, Dolby Atmos, 360 Reality Audio and other immersive formats. Not all earbuds support them, so you’ll want to make sure a perspective pair does if that sort of thing excites you, especially if you plan to use them for playback of high-quality audio.

The primary way we test earbuds is to wear them as much as possible. We prefer to do this over a one- to two-week period, but sometimes embargoes don’t allow it. During this time, we listen to a mix of music and podcasts, while also using the earbuds to take both voice and video calls. Since battery life for earbuds is typically less than a full day, we drain the battery with looping music and the volume set at a comfortable level (usually around 75 percent).

To judge audio quality, we listen to a range of genres, noting any differences in the sound profile across the styles. We also test at both low and high volumes to check for consistency in the tuning. To assess call quality, we’ll record audio samples with the earbuds’ microphones as well as have third parties call us.

When it comes to features, we do a thorough review of companion apps, testing each feature as we work through the software. Any holdovers from previous models are double checked for improvements or regression. If the earbuds we’re testing are an updated version of a previous model, we’ll spend time getting reacquainted with the older buds. Ditto for the closest competition for each new set of earbuds that we review.

Sony WF-C710N

The WF-C710N is a set of compact and comfy earbuds that offer several of Sony’s best features. While the ANC performance is above average for this price ($120), sound quality isn’t as good as the company’s slightly more expensive options. Battery life fell below stated figures and call performance isn’t good enough to use these buds for work.

Beats Powerbeats Pro 2

The newest version of the Powerbeats Pro have an improved, comfortable design, balanced bass and new H2 chips and a heart rate sensor inside. But heart rate support is currently limited on iOS.

Samsung Galaxy Buds 3

The Galaxy Buds 3 combine ANC with an open-type design, which renders the noise-blocking abilities of the earbuds mostly useless. Still, there’s great low-end tone with ample bass when a track demands it. There are also lots of handy features, most of which require a Samsung phone. But at this price, there are better options from Google, Beats and Sony

Sennheiser Momentum Sport

I really like the overall shape of the Momentum Sport earbuds. They’re more comfortable than the Momentum True Wireless 4 and fit in my ears better. What’s more, the body temperature and heart rate sensors work well, sending those stats to a variety of apps. However, that sport-tracking feature works best with Polar’s app and devices, so there’s that consideration. Also, the audio quality and ANC performance isn’t as good as the MTW4, and these earbuds are pricey.

Beats Solo Buds

There’s a lot to like about the Solo Buds for $80. For me, the primary perk is they’re very comfortable to wear for long periods of time thanks to some thoughtful design considerations. You only get the basics here in terms of features and, as expected, the overall sound quality isn’t as good as the pricier models in the Beats lineup. You will get 18 hours of battery life though, since the company nixed the battery in the case and beefed up the listening time in the buds themselves.

Bose Ultra Open Earbuds

Bose created something very unique for this set of earbuds that allows you to stay in-tune with the world while listening to audio content. The clip-on design is very comfortable, but sound quality suffers due to the open-type fit, especially when it comes to bass and spatial audio.

Audio-Technica ATH-TWX7

These stick buds have a compact design that’s comfortable to wear and the warm sound profile is great at times. However, overall audio performance is inconsistent and there’s no automatic pausing.

Master & Dynamic MW09

Retooled audio, better ambient sound mode and reliable multipoint Bluetooth are the best things the MW09 has to offer. They’re expensive though, and you can find better ANC performance elsewhere.

What is considered good battery life for true wireless earbuds?

Most wireless earbuds will last five hours on a single charge, at the least. You can find some pairs that have even better battery life, lasting between six and eight hours before they need more juice. All of the best wireless earbuds come with a charging case, which will provide additional hours of battery life — but you'll have to return each bud to the case in order to charge them up.

Is sound quality better on headphones or earbuds?

Comparing sound quality on earbuds and headphones is a bit like comparing apples and oranges. There are a lot of variables to consider and the differences in components make a direct comparison difficult. Personally, I prefer the audio quality from over-ear headphones, but I can tell you the sound from earbuds like Sennheiser’s Momentum True Wireless 3 is also outstanding.

Which wireless earbuds have the longest battery life?

With new models coming out all the time, tracking the hours of battery life for each this can be difficult to keep tabs on. The longest-lasting earbuds we’ve reviewed are Audio-Technica’s ATH-CKS5TW. The company states they last 15 hours, but the app was still showing 40 percent at that mark during our tests. The only downside is these earbuds debuted in 2019 and both technology and features have improved since. In terms of current models, Master & Dynamic’s MW08 offers 12 hours of use on a charge with ANC off (10 with ANC on) and JBL has multiple options with 10-hour batteries.

What wireless earbuds are waterproof?

There are plenty of options these days when it comes to increased water resistance. To determine the level of protection, you’ll want to look for an IP (ingress protection) rating. The first number indicates intrusion protection from things like dust. The second number is the level of moisture protection and you’ll want to make sure that figure is 7 or higher. At this water-resistance rating, earbuds can withstand full immersion for up to 30 minutes in depths up to one meter (3.28 feet). If either of the IP numbers is an X, that means it doesn’t have any special protection. For example, a pair of wireless earbuds that are IPX7 wouldn’t be built to avoid dust intrusion, but they would be ok if you dropped them in shallow water.

Which earbuds stay in ears the best?

A secure fit can vary wildly from person to person. All of our ears are different, so audio companies are designing their products to fit the most people they can with a single shape. This is why AirPods will easily fall out for some but stay put for others. Design touches like wing tips or fins typically come on fitness models and those elements can help keep things in place. You’ll likely just have to try earbuds on, and if they don’t fit well return them.

What wireless earbuds work with PS5?

PlayStation 5 doesn’t support Bluetooth audio without an adapter or dongle. Even Sony’s own gaming headsets come with a transmitter that connects to the console. There are universal options that allow you to use any headphones, headset or earbuds with a PS5. Once you have one, plug it into a USB port on the console and pair your earbuds with it.

September 2025: Updated to add AirPods Pro 3 to our top picks.

May 2025: Updated to ensure top picks and buying advice remain accurate.

March 2025: Updated the top pick for the best sounding wireless earbuds - runner up.

January 2025: Updated the top pick for best sounding wireless earbuds.

July 2024: Updated our list to include the Samsung Galaxy Buds 3 Pro.

This article originally appeared on Engadget at https://www.engadget.com/audio/headphones/best-wireless-earbuds-120058222.html?src=rss

DJI's tiny but powerful Neo drone drops to just $159 for Prime Day

Amazon's October Prime Day gives you the rare occasion to buy a DJI drone at a discount. The company's popular Neo drone is now available for $159, making an already affordable product even more manageable with a $40 (20 percent) discount. That sum will get you a drone that's very friendly for beginners thanks to features like palm takeoff, but also powerful with subject tracking and even support for FPV flying. You can also grab it with two extra batteries and a charger for $229 (21 percent off) — a deal I'd recommend if you plan to fly a lot.

The Neo is tiny at just 135 grams so it's exempt from any drone license requirements and small enough to slide into a jacket pocket. At the same time, it's safe around people thanks to the protective propeller guards. Beginners can launch it directly from their palm either by pressing a button or giving a voice command, and it will fly one of six preprogrammed smart shots like Follow, Dronie or Rocket. 

The Neo can also be controlled using a smartphone via Wi-Fi or DJI controller, including the RC Motion 3 and even the DJI Goggles 3 for acrobatic FPV flying. It has a very respectable 6.2-mile line-of-sight range and can fly at up to 18 mph, or 36 mph with a special controller. Video quality is very good for such a cheap drone, with up to 4K 30 fps video via the 1/2-inch 12-megapixel sensor and software-based Rocksteady stabilization. 

The main drawbacks are the lack of obstacle detection, relatively short 17-minute battery life and banshee-like noise. However, this is a great first drone for creators and other users, particularly at the Prime Day sale price of $159. As mentioned, you can also get iti with three batteries and a charger for $229

This article originally appeared on Engadget at https://www.engadget.com/cameras/the-dji-neo-drone-drops-to-only-159-for-prime-day-075559142.html?src=rss

Amazon's Echo Spot alarm clock drops to only $45 in this Prime Day deal

If you’re looking to replace your old alarm clock with an Alexa-powered modern smart alternative then, well, you’re more overwhelmed with options than ever. But a $35 saving on last year’s updated Echo Spot might make your decision a bit easier — you can thank Prime Day deals for that.

The latest incarnation of the diminutive Spot was introduced in July 2024, and this matches its record-low price. For that you get a comfortably bedside-sized device with a sharper display than its predecessor, as well as superior sound. The front face is divided into two halves, with a speaker positioned below the hemispherical display.

What screen you do have is more than enough to display the time and weather information, plus it can show you the song or album title and accompanying artwork when you’re listening to music on those improved speakers. It can naturally be used to boss around your other connected smart devices, too.

Alexa might be baked in, but the Echo Spot is intended to be a fairly bare-bones smart alarm clock, so don’t expect as many features as you’ll find on something like the Echo Show 5. But a lot of people just want a modern alarm clock, and arguably the biggest selling point for the Echo Spot is its total lack of a camera. While that means it can do less than the original 2017 Echo Spot, which Amazon did put a camera in, the decision to remove it from a device that lives right next to your bed was probably for the best.

Amazon’s Prime Day sale returns on October 7, so you can expect a range of deals on its various Echo devices. For our guide to all of the best early deals, head here.

This article originally appeared on Engadget at https://www.engadget.com/deals/amazons-echo-spot-alarm-clock-drops-to-only-45-in-this-prime-day-deal-142550522.html?src=rss

The best Prime Day camera and drone deals: Get up to 41 percent off gear from DJI, Canon, Sony and others

Camera buyers usually mark Amazon's October Prime Day on their calendars as it's a great way to get a nice discount on a major purchase — especially since recent tariffs have caused price increases with some brands. This year's sale looks like a good one, with genuine value available on camera equipment, drone and accessory purchases.

Some key deals include discounts on DJI drones, action cams and wireless mics, along with sales on cameras from Sony, Canon, Insta360, GoPro and Nikon. Many of those cameras are ranked highly in our most recent camera guide, like Nikon's Z6 III and the Panasonic S5 IIx — so feel free to check that out as you add items to your cart. Without further delay, here are the best Prime Day camera deals you can grab this week.

DJI Neo drone for $159 (20 percent off): The DJI Neo may be an inexpensive, beginner-friendly drone, but it has powerful features like subject tracking and quick shots. It can shoot video at up to 4K 30 fps and is quick and maneuverable, though it's also fairly loud. You can also grab one with three batteries and a charger for $229 (21 percent off).

DJI Osmo Action 4 for $229 (23 percent off): DJI's Osmo Action 4 is a solid deal at this price, as you're not giving up much in comparison to the most recent model, the Action 5. It has the same big 1/1.3-inch sensor that delivers excellent video quality, especially in low light. It also features a D-LogM profile to boost dynamic range, DJI's excellent clip-on mount and high-quality 4K 120p video.

DJI Mini 4K drone for $239 (20 percent off): The Mini 4K is a beginner-friendly drone, but it's still got plenty of nifty features. As the name suggests, it can capture 4K video at up to 30 FPS. It can also record 2.7K footage at 60 FPS. There's 2x digital zoom, a 3-axis gimbal and electrical image stabilization. The drone can transmit 1080p video from up to 10km away.

DJI Osmo Action 5 Pro for $269 (21 percent off): DJI's Osmo Action 5 Pro is a good alternative to GoPro's Hero 13 Black. I has a 1/1.3-inch sensor that makes it better than its rival in low light and rivals it in battery life as well with over two hours while shooting in 4K 60p. Another nice feature is 47GB of internal storage, something its main rivals lack. And the Osmo Action 5 Pro works directly with DJI’s noise-cancelling Mic 2 and Mic Mini wireless microphones, providing an easy way to record high-quality sound.

GoPro Hero 13 Black for $359 (17 percent off): GoPro’s Hero 13 Black is now available with a new family of modular lenses called the HB series (ultra-wide, anamorphic and macro), making it far more versatile for action creators. It also has a slightly bigger battery that allows longer capture times, up to 90 minutes for 4K 30fps video. Other specs remain the same, including up to 5.3K at 60fps in 10-bit color, along with the excellent Hypersmooth stabilization.

DJI Osmo 360 action cam for $413 (25 percent off): It took DJI awhile to get a panoramic action cam into the market, but the Osmo 360 was worth the wait. It stacks up well against its main rival, Insta360’s X5, by offering better video quality in low light or high-contrast situations. It’s also easy to use, offers good battery life and trumps its competition with a large amount of built-in storage.

Insta360 X5 action camera for $500 (9 percent off): The X5 has a larger 1/1.28-inch sensor compared to the X4's 1/2-inch sensor, ensuring more detail and a notable boost to low-light performance. Another key update is for PureVideo, allowing it work single lens mode rather than just in 360 mode as before.

Sony ZV-1F for $498 (17 percent off): Sony's ZV-1F is designed purely for vlogging thanks to the fixed 20mm lens that works perfectly for self-filming. It uses a 1-inch sensor like other compacts in this series and supports 4K shooting at up to 30 fps, the same as the ZV-1, and 1080p at up to an impressive 120fps. It offers vlog-ready features like background defocus and product showcase. 

Canon EOS R100 with RT-S18-45mm lens for $529 (15 percent off): Canon's EOS R100 is the only modern mirrorless camera you can get with a kit lens for under $600. It comes with a 24-megapixel APS-C sensor that delivers great picture quality with Canon's pleasing skin tones.

Canon EOS R50 with RF-S18-45mm lens $799 (9 percent off): Canon's 24-megapixel APS-C EOS R50 is great for travel, parties, street photography and more thanks to the portable size and built-in flash. It offers nice image quality with warm, human friendly photos and is capable for creators as well with 4K 30fps supersampled video, with 10-bit and HDR capability. It comes with a kit lens that's solid for vlogging or photography with a 24-70mm full-frame equivalent zoom range. 

Sony A6100 with 16-50mm and 55-210mm lenses $800 (33 percent off): If you want to hit the ground running with a camera and lenses on the cheap, Sony's A6100 is an outstanding value right now. While five years old, its autofocus system is still among the best thanks to its intelligent face- and eye-tracking, along with 4K 30 fps video. The color science and low-light capabilities are excellent, so photos are sharp and color accurate, even in dimly-lit environments. The drawbacks are bad rolling shutter and a low-resolution EVF. Still, you won't find much comparable in this price range with two lenses to boot.

Panasonic Lumix S9 for $1,198 (20 percent off): The S9 may be Panasonic's smallest full-frame camera but it comes with the same 24MP sensor and similar video features to the popular S5 II. To that end, it offers up to 6.2K video, in-body stabilization and reliable autofocus. The key feature though is a special LUT button that lets you easily choose a custom cinematic look designed by professional creators.

Panasonic S5 II for $1,498 (25 percent off): It’s the company’s first camera with hybrid phase-detect AF designed to make focus "wobble" and other issues a thing of the past. You can shoot sharp 4K 30p video downsampled from the full sensor width, or 4K 60p from an APS-C cropped size, all in 10-bit color. It even offers 5.9K 30p video. You also get a flip-out screen for vlogging and updated five-axis in-body stabilization that’s the best in the industry. Photo quality is also good thanks to the dual-gain 24-megapixel sensor. The main drawback is the slowish burst speeds.

Sony A7 IV for $2,198 (19 percent off): Resolution is up considerably from the 24-megapixel A7 III to 33 megapixels, with image quality much improved overall. Video is now up to par with rivals with 4K at up to 60p with 10 bit 4:2:2 quality. Autofocus is incredible for both video and stills, and the in-body stabilization does a good job. The biggest drawback is rolling shutter that limits the use of the electronic shutter.

Panasonic S5 IIx with two lenses for $2,298 (22 percent off): The S5 IIx is identical to the S5 II except that it offers high bitrate Apple ProRes capture directly to an external SSD. This deal is ideal for creators who want to hit the ground running as it includes both a 50mm f/1.8 portrait lens and a 22-60mm zoom. 

Nikon Z6 III for $2,397 (11 percent off): Nikon's Z6 III is the best hybrid mirrorless camera for the money, thanks to the incredible performance of the 24-megapixel partially stacked sensor. You can shoot RAW bursts at up to 20 fps in electronic shutter mode while nailing sharp shots thanks to the much-improved autofocus. On the video side, it supports 6K RAW at up to 60 fps, or 4K 120p. All of that makes it ideal for creators and photographers alike, with the only drawback being the reduced resolution.

Sony A7R V for $3,798 (10 percent off): With a 61-megapixel sensor, the A7R V shoots sharp and beautiful images at a very respectable speed for such a high-resolution model (10 fps). It has equally fast and reliable autofocus, the sharpest viewfinder on the market and in-body stabilization that’s much improved over the A7R IV. Video has even improved, with 8K and 10-bit options now on tap, albeit with significant rolling shutter

Osmo Mobile 7 for $68 (24 percent off): One of the best things to elevate your smartphone footage is a gimbal like the Osmo Mobile 7 from DJI. The company makes some of the best drones you can buy, so it knows a thing or two about camera stabilization, so it will make your smarphone videos smooth as silk. It only weighs 300 grams, so it’ll be a lightweight addition to almost any travel kit.

DJI Mic Mini $99 (41 percent off): DJI's Mic Mini bundle includes two mic transmitters and a receiver that can be connected to a camera or phone, along with a charging case. The mics weigh just a third of an ounce and attach easily to subjects via magnets or clips. Everything pairs automatically and they offer high quality 24-bit 20Hz to 20K Hz recording. They also connect easily to smartphones over Bluetooth and other DJI devices like the Action 5 and Osmo Pocket 3 cams using DJI's OsmoAudio direct connection. 

SanDisk 512GB Extreme Pro CFexpress Card Type B for $120 (24 percent off): If you need tons of storage for a newer camera with a CFexpress Type B slot, SanDisk's 512GB is a great deal. It can handle reading and writing at 1,700MB/s and 1,400MB/s respectively, enough speed for RAW video capture. 

Lowepro Flipside BP 300 AW III for $182 (24 percent off): The Flipside BP300 AW III is a rugged, lightweight bag that securely holds and protects your camera gear in the worst weather. It has rear access for protection and a special compartment to hold lenses. 

DJI Mic 2 for $269 (23 percent off): DJI's Mic 2 can not only be connected to a mirrorless camera like other mics, but also supports Bluetooth audio so can easily be connected to a smartphone as well. It picks up audio from your subject at a range of up to 820 feet (less with a Bluetooth connection) and automatically reduces background noise like wind, city sounds and even drone noise. This deal includes two mic transmitters, a receiver for your camera, a charging case and wind muffs. 

DJI RS 4 Mini $309 (16 percent off): The RS 4 Mini is one of the best handheld camera stabilizers on the market thanks to features like its 4.4 pound payload and automatic axis lock. On top of that, the company introduced automatic tracking via its RS Intelligent Tracking Module to keep human subjects in frame no matter where they go. It also offers smoother gimbal tracking, longer battery life and faster charging. 

Lexar 128GB Professional 2000x SD Card $90 (31 percent off): If you need very fast SD UHS II memory, Lexar currently has an excellent deal on its 128GB 2000x memory card. It offers 300MB/s read and 260Mb/s write speeds, quick enough to capture high data rate video or bursts of high-resolution photos. Amazon is also offering a solid discount on 256GB memory in the same lineup at $170 or 23 percent off. 

Manfrotto Befree GT PRO 4-section travel tripod $299 (36 percent off): Belying its miniscule 4.23 pound weight, this model offers ultra-stable performance for timelapse and other shots, even in bad weather. It's also great for video thanks to the integrated 3-way pan/tilt head, featuring fluid cartridges for precise control and seamless movements.

Panasonic Lumix G 25mm f/1.7 lens $248 (17 percent off): Panasonic and Micro Four Thirds users on a budget in need of a fast 50mm equivalent prime should check this one out. It offers beautiful bokeh and fast speeds in low light, but costs nearly half the price of f/1.4 models. 

Nikon Nikkor Z 50mm f/1.4 Lens $497 (17 percent off): If you want a fast 50mm portrait lens but don't have a ton to spend, Nikon's Nikkor Z 50mm f/1.4 lens is on sale for just $497, making it one of the cheapest full-frame options available. At the same time, it offers exceptional sharpness and excellent color rendition, with beautiful bokeh that will flatter your subjects. 

Panasonic Lumix S 85mm f/1.8 lens $448 (31 percent off): I own this lens and it's fast, sharp and versatile for portraits while offering considerable savings compared to an f/1.4 portrait. It's also great for working pros due to the dust-, splash- and freeze-resistant design and very light weight next to other brands. 

Nikon Nikkor Z 35mm f/1.4 Lens $577 (15 percent off): If you're looking for a fast full-frame Nikon Z prime in the 35-50mm category for portraits, architecture or street photos, Nikon has a stellar deal. The Nikkor Z 35mm f/1.4 lens is fast and sharp with painterly bokeh thanks to the 9-blade iris design. It's great for video as well, with the silent STM mechanism offering silent autofocus and suppressed focus breathing.

Canon RF24-70mm F2.8 L lens $2,399 (8 percent off): If you've been eyeballing a fast Canon RF mount zoom, now's the time to act. The RF24-70mm f/2.8 L lens rarely goes on sale, but is now available with a $200 (8 percent) discount. It offers an f/2.8 aperture across the entire zoom range so you get awesome bokeh for portraits and high versatility in low light situations. 

This article originally appeared on Engadget at https://www.engadget.com/cameras/the-best-prime-day-camera-and-drone-deals-get-up-to-41-percent-off-gear-from-dji-canon-sony-and-others-073852143.html?src=rss

Prime Day deals include this battery-powered Ring video doorbell for $80

The Ring Battery Doorbell Plus is on sale for almost half off and is at the lowest price we've ever seen for this model. Normally retailing for $150, the smart doorbell is on sale for $80, a discount of 47 percent. This aggressive sale is part of another Prime Day that runs October 7-8.

The Battery Doorbell Plus offers a 150-by-150-degree "head to toe" field of vision and 1536p high-resolution video. This makes it a lot easier to see boxes dropped off at your front door since it doesn't cut off the bottom of the image like a lot of video doorbells.

This model features motion detection, privacy zones, color night vision and Live View with two-way talk, among other features. Installation is a breeze since you don't have to hardwire it to your existing doorbell wiring. Most users report that the battery lasts between several weeks and several months depending on how users set up the video doorbell, with power-heavy features like motion detection consuming more battery life.

With most video doorbells today, you need a subscription to get the most out of them, and Ring is no exception. Features like package alerts require a Ring Home plan, with tiers ranging from Basic for $5 per month to Premium for $20 per month. You'll also need a plan to store your video event history.

Ring was acquired by Amazon in 2018, and now offers a full suite of home security products including outdoor cameras, home alarm systems and more. This deal is part of a larger sale on Ring and Blink devices leading up to Prime Day.

This article originally appeared on Engadget at https://www.engadget.com/deals/prime-day-deals-include-this-battery-powered-ring-video-doorbell-for-80-154508903.html?src=rss

The best October Prime Day deals on Apple devices: Save on AirPods, iPads, AirTags and more

Amazon’s latest October Prime Day sale is upon us, and it’s brought a few good discounts on Apple devices along with it. If you’ve been waiting to pick up a new iPad, MacBook or set of AirPods, we’ve quickly rounded up the best Apple Prime Day deals we can find below. We’ll update this post as we find more price drops. Just keep in mind that Black Friday is around the corner, so there’s always a chance we see more deals over the next month or so. This latest “Prime Big Deal Days” sale officially runs October 7 and 8, and many (but not all) of the discounts exclusive to Amazon Prime subscribers.

Apple AirPods 4 for $90 ($39 off): The entry-level AirPods get you tons of Apple-friendly features — from fast pairing with your iCloud devices to spatial audio to hands-free Siri — and their lightweight design should fit most ears more comfortably than previous models. They still aren’t the richest-sounding or longest-lasting pair, and they don't support wireless charging, Find My tracking or built-in volume controls. They won’t block much outside noise either. But if you’re an iPhone owner who truly hates the feeling of earphones that jut into your ear canal, they’re a worthy value with this discount, which comes within a dollar of the lowest price we’ve seen. Engadget’s Billy Steele gave the AirPods 4 a review score of 88 last year.

Apple AirPods Max (USB-C) for $430 ($119 off): The AirPods Max are bulkier and older than Sony’s WH-1000XM6 — the top pick in our guide to the best wireless headphones — plus they rely on a weird, flimsy case to preserve power, so we only recommend them to hardcore Apple fans. That said, they offer most of the conveniences of the in-ear AirPods, while their ANC and warm sound profile still rank among the better options on the market. This model is virtually the same as the pair we reviewed way back in 2020, only it has a USB-C port and supports lossless audio with a cable. This discount is $30 off the pair’s all-time low, but it’s the best price we’ve seen since July.

Apple iPad Air (11-inch, M3) for $449 ($150 off): We call the Air the best Apple tablet for most people in our guide to the best iPads. While it’s a good bit more expensive than the base model, it has a handful of subtle improvements that add up in day-to-day use. Its speakers are better. Its display is more color-rich, more glare-resistant and laminated, which means there’s no “gap” between the display and the actual glass. There’s a faster chip and more RAM for gaming or more involved tasks. It also works with the Apple Pencil Pro and comfier Magic Keyboard. This $150 discount matches the lowest price we’ve seen. Other sizes and storage configurations are also similarly discounted.

Apple iPad Air (13-inch, M3) for $649 ($150 off): The 13-inch iPad Air is exactly the same as the 11-inch model, only it has a larger and slightly brighter display. It should be nice if you’re willing to trade a little portability for more space for working or taking in movies. We gave this model a score of 89 in our review back in March.

Apple iPad mini (A17 Pro) for $399 ($100 off): The iPad mini is exactly what it sounds like: the smaller iPad. This latest iteration has an improved A17 Pro chip — which is enough to support Apple Intelligence — along with 128GB of storage in the base model and Apple Pencil Pro support. As always, you’d buy it if you want a display you can more easily hold with one hand. This is another deal we've seen several times, but it's only $20 more than the lowest price we've tracked. 

Apple AirTags (4-pack) for $65 ($34 off): With its enormous finding network and ultra-wideband tech for accurately locating nearby items, Apple’s AirTag is the top pick for iPhone users in our guide to the best Bluetooth trackers. This deal comes within a dollar of the best price we've seen for a four-pack. Single AirTags are also on sale for $19. If you’re not in desperate need for these, though, note that we may see an updated model later this year.

Apple Pencil Pro for $99 ($30 off): The Pencil Pro is Apple’s premium stylus, offering pressure sensitivity, wireless charging, haptic feedback and unique gesture controls compared to the standard Pencil. It’s a dependably tool for sketching and note-taking; just note that it’s not compatible with the base iPad and most older models. This is another discount we’ve seen regularly over the past year, but it’s $30 off the pen’s list price.

Apple MacBook Air (15-inch, M4) for $999 ($200 off): The 15-inch MacBook Air is essentially the same as the 13-inch version, only it has a roomier display, a more spacious trackpad and better speakers. This deal is slightly higher than the laptop’s all-time low, and it’s another discount we’ve seen plenty of times before, but it’s a decent drop from Apple’s usual list price. Again, other versions with more memory and storage are also on sale.

Apple Watch SE 3 (40mm, GPS) for $240 ($9 off): Apple only released its latest crop of Apple Watches a couple weeks ago, but Amazon is selling each with a modest discount. The Watch SE 3, for instance, is available for $240: That’s just $9 off its list price, but it’s something if you’re buying soon after launch anyway. As for the device itself, this third-gen version of the SE adds an always-on display, which makes it so you no longer have to wake the watch to check the time or your notifications. It now runs on the same S10 chip as its higher-end siblings, too, and it still covers the essential health and fitness features. If you’re in the market for your first smartwatch, or if you only want a watch for step counting, sleep tracking and notifications, it should be a strong value.

Apple Watch Series 11 (42mm, GPS) for $390 ($9 off): The Series 11 is the top overall pick in our guide to the best smartwatches. It’s not as compelling as the SE 3 from a value perspective, but it has a bigger, brighter and more scratch-resistant display, slightly longer battery life and support for more advanced health features like hypertension alerts and an ECG app. We gave it a score of 90 in our review. Just note that there’s no pressing need to upgrade if you’re happy with your current Apple Watch. This is a slight $9 discount, but we note it since the Series 11 has been on sale for less than a month, so any bit of savings are welcome.

Apple Watch Series 10 (42mm, GPS + Cellular) for $329 ($170 off): As a general rule, most people should buy the newest version of whatever Apple product they want. That said, the Series 11 is only a minor upgrade over the prior Series 10 — the glass is more scratch-resistant, the cellular model has 5G support, the battery is marginally bigger and that’s about it. If you’re mostly worried about saving cash, want LTE support and still need the more advanced health features the Apple Watch SE lacks, this is a worthwhile discount for a version of last year’s model with cellular support.

Apple Watch Ultra 3 (49mm) for $780 ($19 off): The Apple Watch Ultra 3 is a much more niche device than the Series 11 or SE 3, aimed at serious athletes and adventurers more than the typical gymgoer. It is the biggest and most rugged Apple Watch, with the brightest display (up to 3,000 nits), longest battery life (up to 42 hours) and most advanced components. It also supports satellite communications. The watch normally retails for $799, so this isn’t a major discount, but since this is a brand-new device, any drop at all is worth noting for early adopters.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-best-october-prime-day-deals-on-apple-devices-save-on-airpods-ipads-airtags-and-more-125644485.html?src=rss

The best Amazon Prime Day deals: Get up to 58 percent off Apple, Samsung, Lego, Dyson, Shark and others during the October sale

October Prime Day 2025 is here, so Prime members can finally take advantage of the long-awaited discounts across Amazon’s site. The online retailer has hosted this kickstart to the holiday shopping season for the past few years, and it typically brings decent sale prices to all things homeware, clothing, shoes, and yes, technology.

This year is no different — a number of our top picks for the best tech you can buy right now have been discounted for Prime Day. Most Prime Day deals are exclusively available to Prime members, but there are always a few stragglers that anyone with an Amazon account can get. We’ve gathered the best October Prime Day deals on tech that we could find here. We’ll be updating this post regularly throughout the duration of Prime Day, so check back to see the latest discounts.

Apple AirPods 4 (no ANC) for $90 (30 percent off): Apple's entry-level wireless earbuds are the best AirPods for folks on a budget. They have an open-ear design that some will prefer over eartips that stick into your ears full, plus they have much improved audio quality and the H2 chip that enables features like Voice Isolation, Personalized Spatial Audio with dynamic head tracking and Adaptive EQ.

Apple MacBook Air (13-inch, M4) for $799 (20 percent off): Our top pick for the best laptop for most people, the latest MacBook Air is impressively thin and light without skimping on performance. The M4 chipset is powerful enough to handle everyday tasks without breaking a sweat, plus some gaming and labor-intensive work. It has a comfortable keyboard, luxe-feeling trackpad and an excellent battery life.

Amazon Smart Plug for $13 ($12 off): We named this the best smart plug for Alexa users because it hooks up painlessly and stays connected reliably. Use it to control lamps or your holiday lights using programs and schedules in the Alexa app, or just your voice by talking to your Echo Dot or other Alexa-enabled listener.

Leebein 2025 electric spin scrubber for $40 (43 percent off, Prime exclusive): This is an updated version of my beloved Leebein electric scrubber, which has made cleaning my shower easier than ever before. It comes with seven brush heads so you can use it to clean all kinds of surfaces, and its adjustable arm length makes it easier to clean hard-to-reach spots. It's IPX7 waterproof and recharges via USB-C.

Jisulife Life7 handheld fan for $25 (14 percent off, Prime exclusive): This handy little fan is a must-have if you life in a warm climate or have a tropical vacation planned anytime soon. It can be used as a table or handheld fan and even be worn around the neck so you don't have to hold it at all. Its 5,000 mAh battery allows it to last hours on a single charge, and the small display in the middle of the fan's blades show its remaining battery level.

Dyson V15 Detect Plus for $570 (33 percent off): Our top pick for the best cordless vacuum on the market right now, the V15 Detect has some of the strongest suction power you'll find in a stick vacuum, plus a lightweight design and a dustbin that can hold more dirt and debris than it might seem. This model comes with five cleaning attachments, including the Fluffy Optic cleaner head that has LEDs to illuminate the floor as you're vacuuming so you can better see where all the dust bunnies are.

Apple iPad Air (11-inch, M3) for $449 ($150 off): The only major difference between the latest iPad Air and the previous generation is the addition of the faster M3 chip. We awarded the new slab an 89 in our review, appreciating the fact that the M3 chip was about 16 percent faster in benchmark tests than the M2. This is the iPad to get if you want a reasonable amount of productivity out of an iPad that's more affordable than the Pro models.

Samsung EVO Select microSD card (128GB) for $13 (24 percent off): This Samsung card has been one of our recommended models for a long time. It's a no-frills microSD card that, while not the fastest, will be perfectly capable in most devices where you're just looking for simple, expanded storage.

Lego Star Wars Advent Calendar 2025 75418 for $31 (30 percent off): 'Tis the season for advent calendars. They've flooded the internet, as they usually do this time of year, and there are plenty to choose from (and stock up on) before we get to December. Lego has a bunch, and this Star Wars one will be fun for everyone, not just kids, to open up each day at the end of the year.

Anker 622 5K magnetic foldable power bank with stand for $34 (29 percent off, Prime exclusive): This 0.5-inch thick power bank attaches magnetically to iPhones and won't get in your way when you're using your phone. It also has a built-in stand so you can watch videos, make FaceTime calls and more hands-free while your phone is powering up.

Blink Mini 2 security cameras (two-pack) for $35 (50 percent off): Blink makes some of our favorite security cameras, and the Mini 2 is a great option for indoor monitoring. It can be placed outside with the right weatherproof adapter, but since it needs to be plugged in, we like it for keeping an eye on your pets while you're away and watching over entry ways from the inside.

Amazon Fire TV Stick 4K Max for $40 (33 percent off): Amazon's most powerful streaming dongle supports 4K HDR content, Dolby Vision and Atmos and Wi-Fi 6E. It also has double the storage of cheaper Fire TV sticks.

Anker Soundcore Space A40 for $45 (44 percent off): Our top pick for the best budget wireless earbuds, the Space A40 have surprisingly good ANC, good sound quality, a comfortable fit and multi-device connectivity.

Anker MagGo 10K power bank (Qi2, 15W) for $63 (22 percent off, Prime exclusive): A 10K power bank like this is ideal if you want to be able to recharge your phone at least once fully and have extra power to spare. This one is also Qi2 compatible, providing up to 15W of power to supported phones.

Levoit Core 200S smart air purifier for $70 ($20 off, Prime exclusive): This compact air purifier cleans the air in rooms up to 140 square feet and uses a 3-in-1 filter that removes microscopic dust, pollen and airborne particles. It has a mobile app that you can use to set runtime schedules, and it works with Alexa and Google Assistant voice commands.

iRobot Roomba 104 Vac for $150 (40 percent off): This entry-level robot vacuum is an upgraded version of one of our favorite budget robot vacuums. iRobot makes robovacs that are easy to use, even for those who have never had a machine like this, and they do a good job of cleaning up all kinds of messes, including pet hair.

Levoit LVAC-300 cordless vacuum for $250 ($100 off, Prime exclusive): One of our favorite cordless vacuums, this Levoit machine has great handling, strong suction power for its price and a premium-feeling design. Its bin isn't too small, it has HEPA filtration and its battery life should be more than enough for you to clean your whole home many times over before it needs a recharge.

Shark Robot Vacuum and Mop Combo for $300 (57 percent off, Prime exclusive): If you're looking for an autonomous dirt-sucker that can also mop, this is a good option. It has a mopping pad and water reservoir built in, and it supports home mapping as well. Its self-emptying base can hold up to 60 days worth of debris, too.

Amazon Kindle Scribe for $300 (25 percent off): The latest Kindle Scribe is one of the best E Ink tablets you can get, particularly if you want a slab that can handle both note-taking and e-reading well. Its edge is providing users full access to the Kindle ebook store and the ability to take notes to accompany reading material. This model has 16GB of storage and includes the Premium Pen.

XReal One Pro AR glasses for $649 (16 percent off): The latest from XReal, these smart glasses can let you use almost any device, including your smartphone, with a large virtual display. Their 1080p Micro-OLED screens are bright and sharp, plus they're pretty comfortable to wear.

Nintendo Switch 2 for $449: While not technically a discount, it's worth mentioning that the Switch 2 and the Mario Kart Switch 2 bundle are both available at Amazon now, no invitation required. Amazon only listed the new console for the first time in July after being left out of the initial pre-order/availability window in April. Once it became available, Amazon customers looking to buy the Switch 2 had to sign up to receive an invitation to do so. Now, that extra step has been removed and anyone can purchase the Switch 2 on Amazon.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-best-amazon-prime-day-deals-get-up-to-58-percent-off-apple-samsung-lego-dyson-shark-and-others-during-the-october-sale-050801203.html?src=rss

The Amazon Smart Plug drops to only $13 for Prime Day

There are few things as simple yet exceedingly annoying as having to get up and turn off a light. Whether you're already comfortable in bed or live with mobility limitations, smart plugs can be a great option — especially when they're on sale.

Right now, you can pick up an Amazon Smart Plug for a record-low price of $13, down from $25. The 48 percent discount comes as part of Amazon Prime Day sales. You can also pick up a two-pack of Amazon's Smart Plugs for $24, down from $50 — a 52 percent discount.

The Amazon Smart Plug is our pick for best smart plug if you have an Alexa-enabled home. You can tell Alexa to turn off the lights or control it with the Alexa app. It's compatible with most plugged in devices, from lamps and fans to even kitchen appliances. You can also set it to turn on lights or devices at a certain time each day.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-amazon-smart-plug-drops-to-only-13-for-prime-day-130446520.html?src=rss

The best Prime Day SSD deals: Save on gear from Samsung, SanDisk, Crucial and others

October Prime Day is almost here again, and as always, we've put together a list of the best, most-discounted extra storage options for consoles and gaming PCs. Our guide to the top storage deals is split into three categories — tiny microSD cards, portable external SSDs and generally faster (but more labor-intensive) internal SSDs. Whether you're building a gaming rig or just want to stop your PS5 from lagging, there's almost certainly a deal for you on our curated list. Take a look now to see what kind of extra storage might best fit your needs.

Best October Prime Day SSD deals

Crucial X10 4TB Portable SSD for $240 (39 percent off): The Crucial X9 is our current favorite portable SSD, but the X10 is a step beyond. Based on USB 3.2, it's compatible with a wide range of devices, and reaches reading speeds of 2,100MB/s — all in an extremely portable pocket-sized case.

Crucial BX500 1TB 3D NAND SATA 2.5-inch Internal SSD for $64 (33 percent off): The BX500 is Crucial's budget-conscious SSD option, but that doesn't mean it disappoints. This internal solid state drive cuts down on battery consumption and improves processing with read speeds reaching 540MB/s. It comes with a three-year warranty, too.

Crucial P310 2TB for $122 (49 percent off): The 2TB level of the Crucial P310 is available at the lowest price we've seen all year. It's a compact drive that works great in small laptops or Steam Decks — especially the latter, given Valve's warning against cramming in large SSDs. For this small size and great price, you get read speeds of over 7,000MB/s.

Samsung 990 Evo Plus 1TB for $70 (33 percent off): This is an incredible deal on an internal SSD from a reliable brand. All capacities of the Samsung 990 Evo Plus are currently on sale, but even the 1TB option can visibly boost your performance with top read speeds of 7,250MB/s. There's no heatsink, but nickel plating on the controller reduces both heat output and energy use.

Samsung 990 PRO 2 TB for $170 (19 percent off): Samsung's 990 Pro series represents a massive leap forward for the brand, cutting energy costs by around half while boosting speeds up to at least 7,000MB/s write and 5,000MB/s read. This version comes with 2TB of storage and its own built-in heatsink.

Samsung Fit Plus 256GB for $23 (30 percent off): The Samsung Fit Plus isn't just the best thumb drive on the market right now — it's one of the best SSDs, period. This deal gets you 256GB of storage and read speeds of 400MB/s for almost unfathomably cheap. It's also built to resist water, extreme temperatures, magnets and even radiation.

Kingston SX1000 1TB High Performance for $89 (23 percent off): If you can't swing a Crucial X9 or X10 right now, Kingston offers a much more affordable alternative. The SX1000 can handle both PC backups and gaming storage with ease, nearly matching the X9 in our speed tests. It's also easy to carry around, though keep a tight grip as it's not officially rated for any drop height.

SanDisk 2TB Extreme Portable SSD for $150 (29 percent off): The SanDisk Extreme line of portable SSDs provides a great middle ground in both price and performance. Read speeds of 1,050MB/s are enough for transferring most files you'll encounter day-to-day. Its external design holds up too, with a large carabiner loop, IP65 waterproofing and dustproofing and drop protection as high as three meters.

Seagate Storage Expansion Card 2TB for $249 (31 percent off): This SeaGate SSD is specifically for expanding storage on the Xbox Series X and S, and it's designed to meet Xbox specs exactly. You'll get exactly the same performance booting a game from this card as you would from internal Xbox storage, which makes data management infinitely easier.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-best-prime-day-ssd-deals-save-on-gear-from-samsung-sandisk-crucial-and-others-170039457.html?src=rss

Get a bundle of two Blink Mini 2 cameras for only $35 for Prime Day

Amazon's Prime Big Deal Days is here, and you can save on all sorts of tech this time around. A mainstay of Prime Day sales, a pair of Blink Mini 2 cameras is on sale for only $35. That's 50 percent off, a record low and less than what you'd usually pay for one. It's also Engadget's pick for the best budget security camera.

This budget-friend, wired camera is well-suited for nighttime video: It has a built-in LED spotlight, color night vision and a low-light sensor. Day or night, it records in sharp 1080p resolution. It also has a wider field of view than its predecessor.

The Blink Mini 2 is primarily designed for indoor use. But you can use it outdoors, too. You'll just need to fork over $10 for a weather-resistant adapter. Wherever you use the camera, it works with Alexa and supports two-way audio. ("Hello, doggy, I'll be home soon.")

It also supports person detection. (That's a neat feature that differentiates between people and other types of movement.) However, the feature requires a Blink Subscription Plan. They start at $3 per month or $30 per year for one device.

The camera is available in black or white. Both colors are available for the $35 Prime Day deal, but they can't be mixed unless you buy each separately. It's worth noting that this deal is open to anyone — no Prime subscription necessary. Amazon recently announced an upgraded version of this camera that supports 2K video, but it won't be available until later in October.

You can also save on a bunch of other Blink (and Ring) security gear for Prime Day. The Blink Outdoor 4 cameras are some of our favorites, and most configurations are on sale for Prime Day, including bundles like this three-camera system that's 61 percent off.

This article originally appeared on Engadget at https://www.engadget.com/deals/get-a-bundle-of-two-blink-mini-2-cameras-for-only-35-for-prime-day-201049013.html?src=rss

The Roku Streaming Stick Plus is cheaper than ever for Prime Day

The Roku Streaming Stick Plus is on sale for just $24. That's a discount of 40 percent and the lowest we've ever seen it.

Roku has held the top spot in the TV OS market for years thanks to its user-friendly interface, an affordable range of streaming devices and its own lineup of TVs. We picked the Streaming Stick Plus as the best streaming device for free and live content, thanks in large part to The Roku Channel app that accompanies it. The Roku Channel features over 500 free TV channels with live news, sports coverage and a rotating lineup of TV shows and movies.

In our hands-on review of the Roku Streaming Stick Plus, we thought it was perfect for travel thanks to its small size and the fact that it can be powered by your TV's USB port, nixing the need for a wall adapter. Menu navigation and opening or closing apps won't happen at quite the same speeds as more expensive streamers, but it's quick enough for what is ultimately a pretty low-cost option. The Wi-Fi range on this one is also weaker than Roku's pricier devices, but unless you are placing it exceedingly far from your router, it shouldn't be an issue.

The Roku Streaming Stick Plus supports both HD and 4K TVs, as well as HDR10+ content. It doesn't support Dolby Vision, however; for that you'll need to upgrade to Roku's Streaming Stick 4K or Roku Ultra. It comes with Roku's rechargeable voice remote with push-to-talk voice controls. Roku's remote can also turn on your TV and adjust the volume while you're watching.

If you've been thinking about getting a Roku device, or you already love the platform and want a compact and convenient way to take it with you when you travel, then this sale provides a great opportunity. If you'd prefer a more powerful and slightly more stationary device, the Roku Ultra is on sale for $69 right now, too.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-roku-streaming-stick-plus-is-cheaper-than-ever-for-prime-day-134656043.html?src=rss

The best Prime Day kitchen deals include up to 50 percent off our favorite air fryers and more

New gear can make cooking easier (and sometimes more fun) but the best kitchen tech isn’t cheap. That’s why it’s a good idea to wait for a sale like Amazon's Prime Day. At Engadget, we’ve tested plenty of kitchen gear (and have opinions to accompany said testing) so we searched for discounts on what we actually recommend. We found deals on some of our preferred brands like air fryers from Breville, Instant Pot cookers, sous vide machines and even a Ninja ice cream maker. We’ve linked to our testing and reviews so you can get the information you need before you buy. Here are the best Prime Day deals on kitchen tech we could find.

Ninja Slushi for $300 ($50 off): We haven’t officially tested this one just yet, but given our love of both the Creami and the Creami Swirl, it’s a fair bet that this one will do good things with cold stuff too. This is the lowest price of the year so far.

Vitamix 2-Speed Immersion Blender for $90 ($40): For quick sauces and soups, we like this nifty two-speed hand blender from Vitamix, and said so in our guide to cheap kitchen gadgets. It even comes with a whisk for making your own whipped cream.

Vitamix 5-Speed Immersion Blender for $200 ($100 off): If you want a little extra oomph from your hand blender, grab this one. We named it the best counter top replacement model in our guide to the best immersion blenders. The 625 watt motor is more powerful than most and the blender head is intelligently designed to reduce suction and prevent scratches plus it fits inside a wide mouth mason jar.

Ninja Creami ice cream maker with extra pints $200 ($55 off with Prime): Here's a bundle that includes two additional pint jars for the ice cream maker. It's one of our favorite pieces of kitchen tech and we called it a frozen fantasy-maker in our review. Note that the unit itself dipped down to $160 last Black Friday and the extra pints have gone as low as $16. 

Hamilton Beach Digital Rice Cooker for $37 ($13 off): Our top pick for a budget rice cooker is great for small kitchens (but probably not for big families). We thought it outperformed other cookers that are four times the price, though it’s not the best for all-day warming of rice as the bottom bigs got a little overdone.

Instant Pot Vortex Plus with Clear Cook for $80 ($20 off with Prime): This Vortex air fryer model is similar to our top airfryer pick, but is missing the Odor Ease feature. It still has the Clear Cook window that lets you keep an eye on your food as it crisps and the Vortex cooking tech heats up remarkably fast, with almost no pre-heating time.

Instant Pot Vortex Plus Air Fryer (4QT) for $65 ($65 off): Here’s a smaller version of our best overall air fryer. This one has a four-quart capacity, which is perfect for one person and small kitchens. And, like its larger sibling, pre-heats quickly thanks to a 1600-watt output. 

Instant Pot Vortex 2-QT Mini for $38 ($22 off with Prime): The budget model from our air fryer guide may not be large but its two-quart basket is enough to reheat leftovers for two or cook up a batch of frozen appetizers. And, because of its small size, it doesn’t take up a ton of space on your countertops — ideal for a small kitchen.

Breville InFizz Fusion beverage maker for $200 ($50 off): We called this fizz-maker the bubble master in our review. True, it’s pricier than rivals in the same space, but the upscale design — that actually looks good on a countertop — somewhat makes up for the price premium. We also appreciate the Fusion Cap that helps prevent messy eruptions when you’re bubbling up your drinks.

Cosori Air Fryer Pro Compact for $85 ($15 off with Prime): One of our concerns with the Cosori 9-in-1 was that it was a little on the wide side, taking up extra countertop space. The Pro Compact has a more space-saving design while still packing a five-quart basket.

Instant Pot 4QT Vortex mini air fryer for $55 ($35 off with Prime): This model’s four-quart capacity falls between our top Instant Pot air fryer pick and the budget model. It has the Clear Cook window feature, six presets and it comes in pink in addition to the standard white and black.

Instant Pot Duo Plus for $90 ($50 off): We named this the best multicooker in our guide to the best kitchen tech we’ve tested. It can cook a dizzying array of foods from basic beans and rice to homemade yogurt. We like this one because it’s simple to use, and has quick-cooking modes for soup, eggs and grains. There’s even a sous vide cooking function.

Breville Bambino Plus for $400 ($100 off): In our gift guide for coffee lovers, this espresso machine earned our respect for its compact size and the fact that it doesn’t cost a grand, like some machines do. Plus the controls are easy for beginners to learn but makes silky milkfoam for pro-level latte artists.

Breville Juice Fountain Plus for $130 ($20 off): This went as low as $110 back in January, but it’s still a decent discount on a high-powered juice extractor. We were won over by its impressive juicing abilities and despite how it looks, it's surprisingly easy to clean — as long as you do it right away.

Ninja Dual Foodie Zone Air Fryer (DZ302) for $180 ($50 off): If you want to air fry two different things at the same time, this is the one to get. This is the same in specs, capacity and wattage as our top pick for a dual-zone air fryer, just with a different model number. It even has a feature that makes sure the two different foods are ready at the same time. Note the price was $20 lower for Prime members back in July.

Ninja 5.5-quart Air Fryer XL (AF150AMZ) for $130 ($50 off): This one earned an honorable mention in our guide to air fryers. It’s double the capacity of our budget pick and has a dehydrate preset. While we found the round basket a little cramped, we liked how the fryer’s vertical design saved counter space. Just note that this went as low as $90 in July.

Vitamix Explorian E310 Blender for $330 ($50 off): This isn’t the best deal we’ve seen this year — this Vitamix dropped to $300 as recently as May. But if you’re in the market for a blender that can turn the most recalcitrant nut into the creamiest butter, we think this won’t let you down.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-best-prime-day-kitchen-deals-include-up-to-50-percent-off-our-favorite-air-fryers-and-more-193009268.html?src=rss

The second-gen Kindle Scribe is $100 off in this Prime Day deal

Amazon Prime Day is here and you can save on all manner of tech, including Amazon's devices, from Kindle tablets to Echo speakers.

Take the Amazon Kindle Scribe, which has dropped to $300 from $400. The 25 percent discount brings the second-generation device back down to its all-time low price. This deal is available on the 16GB model in Tungsten (dark grey) and doesn't include Kindle Unlimited. If you want more storage, the 32GB and 64GB versions are also 25 and 24 percent off, respectively.

The Kindle Scribe is our pick for best e-reader E Ink tablet. It scored an 86 in our review, thanks to the smooth reading and writing experience it provides. We were also fans of its sleek and slender design. It has a 10.2-inch screen and offers front lights that auto-adjust to your needs. Our biggest gripe was the cost which, thanks to this deal, is less of an issue.

It's worth noting that Amazon announced a few new Kindle Scribe models recently, but none of them are available just yet. The Kindle Scribe 3 has a thinner and lighter design, plus faster page-turning and writing experiences. The Kindle Scribe Colorsoft is Amazon's first full-color E Ink tablet. If you want the latest and greatest, it's worth it to wait — but Amazon hasn't yet named specific release dates for either new model. During its latest hardware event, it only said the new Scribes were coming "later this year."

This article originally appeared on Engadget at https://www.engadget.com/deals/the-second-gen-kindle-scribe-is-100-off-in-this-prime-day-deal-134701405.html?src=rss

Prime Day Lego deals include up to 56 percent off Star Wars and Super Mario sets

October Prime Day is here, and Amazon is awash with discounts to consider. Amazon uses October Prime Day to kickstart the holiday shopping season — even if you haven't even thought about the holidays yet, you should give the latest Prime Day Lego deals a look. A number of Lego sets from the Super Mario and Star Wars collections are already on sale for up to 38 percent off.

When shopping for Lego sets on Amazon, we highly recommend checking a price tracker like Keepa or CamelCamelCamel before buying. It's not difficult to find Lego sets "on sale" at Amazon, but often those discounted prices have been around for a long time. We've clocked "deals" in which the sale price has been available for months already, going back as far as late spring and early summer 2025. Here, we're mostly highlighting discounts on new Lego sets, recent price drops and record-low (and close to them) prices on popular Lego sets from franchises like Star Wars, Super Mario and others.

• Lego City Advent Calendar 2025 60475 for $21 (40 percent off)

• Lego Star Wars Advent Calendar 2025 75418 for $31 (30 percent off)

• Lego Minecraft Advent Calendar 2025 21280 for $38 (16 percent off)

• Lego Harry Potter Advent Calendar 2025 76456 for $38 (16 percent off)

• Lego Star Wars: The Mandalorian Paz Vizsla and Moff Gideon Battle set 75386 for $17 (56 percent off)

• Lego Star Wars Brick-Built Star Wars Logo set 75407 for $50 (17 percent off)

• Lego Star Wars R2-D2 set 75379 for $80 (20 percent off)

• Lego Super Mario: Mario Kart Donkey Kong & DK Jumbo set 72033 for $28 (19 percent off)

• Lego Super Mario World: Mario & Yoshi Building set 71438 for $104 (20 percent off)

• Lego Classic Medium Creative Brick Box 10696 for $18 (49 percent off)

• Lego Botanicals Bouquet of Roses 10328 for $48 (20 percent off)

• Lego City F1 Garage & Mercedes-AMG & Alpine Cars set 60444 for $64 (20 percent off)

• Lego Brick Tile backpack for $50 (29 percent off)

This article originally appeared on Engadget at https://www.engadget.com/deals/prime-day-lego-deals-include-up-to-56-percent-off-star-wars-and-super-mario-sets-121513577.html?src=rss

Amazon Prime Day deals include up to 58 percent off Shark robot vacuums

Fall Prime Day has arrived and there are a good number of deals on tech we love. Case in point: Shark robot vacuums. Shark makes some of our favorite robovacs and a few of them are already discounted for Prime members ahead of the sale. The Shark AV2501S AI Ultra robot vacuum is one of them, with a whopping 58-percent discount that brings it down to $230. This discount marks a record low for this model.

Shark offers several variations of its AI Ultra robot vacuums. There are small variations between them, and a different model is our pick for the best robot vacuum for most people. In general, you can expect solid cleaning performance from these devices, along with accurate home mapping and an easy-to-use app.

The model that's on sale here is said to run for up to 120 minutes on a single charge, which should be enough to clean an entire floor in a typical home. The self-emptying, bagless vacuum can store up to 30 days worth of dirt and debris in its base. Shark says it can capture 99.97 percent of dust and allergens with the help of HEPA filtration.

If you'd rather plump for a model that's able to mop your floors too, you're in luck: a Shark Matrix Plus 2-in-1 vacuum is on sale as well. At $300 for Prime members, this vacuum is available for $400 (or 57 percent) off the list price. Its mopping function can scrub hard floors 100 times per minute. You can also trigger the Matrix Mop function in the app for a deeper clean. This delivers 50 percent better stain cleaning in targeted zones, according to Shark.

This article originally appeared on Engadget at https://www.engadget.com/deals/amazon-prime-day-deals-include-up-to-58-percent-off-shark-robot-vacuums-171836162.html?src=rss

Get three months of Audible for only $3 with this Prime Day deal

Audiobook fans can get three months of Audible for $3 for a limited time ahead of the fall edition of Amazon Prime Day. Users will be charged $0.99 per month for the first three months, after which it will auto-renew at $14.95 per month.

Audible features thousands of titles in its catalog, including podcasts and Audible Originals. Subscribers will also get to choose one audiobook each month to keep in their collection for free, including best-sellers or new releases. Amazon Prime members will receive two credits the first month of their trial.

Amazon has been bringing Prime Day back in the fall for a few years now, and this year it returns October 7 and 8. Great deals tend to start rolling out in the days ahead of the event and this year is no exception. There are already sales on Apple devices, smart doorbells and most importantly, Lego sets.

If you're a book lover but don't have the time to sit down and read a hard copy, or you just prefer listening to the latest novel while on the go, then take advantage of this sale. It's a limited-time offer and will only be available through December 16.

This article originally appeared on Engadget at https://www.engadget.com/deals/get-three-months-of-audible-for-only-3-with-this-prime-day-deal-090052863.html?src=rss

Mina the Hollower, from the makers of Shovel Knight, has been delayed

There's good and bad news coming from Yacht Club Games, the studio behind Shovel Knight. The bad news is that the company's upcoming game, Mina the Hollower, has been delayed from its planned October 31 release date. But the good news is that it doesn't seem like this will be a lengthy wait. 

There's no new launch date given in the post on the developer's website, but Yacht Club says that it won't be much longer before the game is ready. "The entire crew has been working tirelessly day and night to finish development for release on Oct 31st, but we’re just not quite there yet," the post states. "This isn’t a major delay, it’s just a stretch of time to apply some final polish and balancing to make the game truly shine." Mina the Hollower is coming to life thanks to a successful Kickstarter by the team during 2022. It's an action-adventure game with nods to the Game Boy Color's aesthetics.

This article originally appeared on Engadget at https://www.engadget.com/mina-the-hollower-from-the-makers-of-shovel-knight-has-been-delayed-222817866.html?src=rss

Bungie will conduct technical testing on Marathon later this month

After several silent months, Bungie has opened applications for players interested in participating in a closed technical test of Marathon. It's an invitation-only limited-time test, lasting just October 22-28. Those who are chosen to participate can only do so under a non-disclosure agreement, so don't expect to hear any juicy details or insights about the game's status later this month once the event wraps. 

While it's not a new public look at Marathon, news of a fresh round of testing marks a small step forward for this team extraction shooter. The game was originally slated for a September 2025 release, but in June, Bungie said it was indefinitely delaying the project. The pause was meant to give the studio time to address player feedback from the closed alpha test it conducted earlier in the spring. Considering the language used for this testing phase, we'll still have a while to wait before getting a new launch window.

Whenever it does arrive, Marathon will be Bungie's first new title since the company introduced Destiny 2 back in 2017. Before the delay, the studio had also drawn raised eyebrows with Marathon after the alpha build used stolen artwork, which has been a repeat issue for Bungie.

This article originally appeared on Engadget at https://www.engadget.com/bungie-will-conduct-technical-testing-on-marathon-later-this-month-213422857.html?src=rss

Spotify, Canva and other apps can now connect to ChatGPT

You’ll soon be able to interact with some of your favorite apps, including Spotify and Canva, right inside of ChatGPT. OpenAI announced the integration, which is enabled by the company’s new Apps SDK, during its DevDay presentation. As of today, ChatGPT can connect to a handful of apps, with more to come over time and OpenAI working on submission guidelines that will allow developers to monetize their work.

As a ChatGPT user, you can any access available third-party app by referencing it in your conversations with the chatbot. In the case of Spotify, for example, you can write “Spotify, make a playlist for my party this Friday.” The first time you mention an app in this way, you’ll be prompted to connect your account to ChatGPT. When working with Spotify, ChatGPT can make recommendations based on a mood, theme or a topic. The interface will eventually lead you to Spotify itself, where you can listen to what ChatGPT has created.

“It’s early days, so while we might not be able to deliver on every request just yet, we’ll continue to build, refine, and improve the experience over the coming weeks and months,” Spotify says of the integration.

OpenAI showed off other apps working inside of ChatGPT. For instance, an employee demoed Canva creating a few posters for a dog-walking business that they had talked to ChatGPT about starting. With today’s announcement, ChatGPT can connect to Canva, Coursera, Figma, Spotify and Zillow. In the coming weeks, DoorDash, OpenTable, Target and Uber will also work with the chatbot. And later this year, OpenAI says it will begin accepting app submissions for review and publication.

This article originally appeared on Engadget at https://www.engadget.com/spotify-canva-and-other-apps-can-now-connect-to-chatgpt-194830116.html?src=rss

Apple is bringing back Slide Over in iPadOS 26.1

The reaction to iPadOS 26 has been mostly positive, but some users have bemoaned the absence of the decade-old Slide Over multitasking feature. Well, Apple just announced that it's coming back as part of iPadOS 26.1. It's actually already available, though just as part of the beta build.

For the uninitiated, Slide Over is a multitasking tool that lets users quickly reveal and hide an app that slides in and out from the side of the screen. This is great for quickly revealing and hiding an app without permanently rearranging your workspace.

Apple also got rid of Split View with iPadOS 26, but the new windowing system can easily do something similar. There has been no way with the new OS to mimic Slide Over, which has become a source of frustration for tablet fans.

Slide Over has been redesigned to work with that new windowing system and there's a major upgrade. Users can quickly resize the windows and change the aspect ratios, in addition to swiping right to completely hide them from view.

There is also one major downgrade, according to reporting by 9to5Mac. Users can only have one active app in Slide Over at a time. In iPadOS 18 (and earlier), folks could switch between multiple apps in a Slide Over window, but it looks like that option has been removed. 

Given how iPadOS 26 radically reworked the way the system handles multiple apps, it's a safe bet that this old Slide Over behavior was incompatible with the latest OS. Engadget has reached out to Apple for clarification and will update this post when we hear back.

This article originally appeared on Engadget at https://www.engadget.com/mobile/tablets/apple-is-bringing-back-slide-over-in-ipados-261-190036025.html?src=rss

Evernote's biggest update since 2020 goes all in on AI

If you're like me, it's probably been a hot minute since you've thought about Evernote. For years, the note-taking app, once a darling of the App Store, faced declining popularity and profitability. The last time it grabbed headlines was in 2022 when it was sold to Milan-based developer Bending Spoons. Approximately nine months later, Bending Spoons laid off nearly all of Evernote's US workforce, and relocated what remained of the company's operations to Europe. Now, Evernote wants to be in the conversation again. 

Since taking over the app, Bending Spoons has released more than 200 improvements, culminating today with the release of Evernote v11. The company is touting the update, which is the first major release for Evernote since its poorly received v10 release back in 2020. The new version brings a handful of new AI features to Evernote, including one that Bending Spoons collaborated with OpenAI to develop.      

"My wish for v11 is that people give it a try and see how fast and reliable it is, how complete it is in what it offers," Federico Simionato, Evenote's product lead, said over a video chat. "I think people have an outdated opinion of Evernote that used to match reality, when Evernote was slow and unreliable."    

The first of the new features Bending Spoons is introducing today is called AI Assistant. Simionato describes it as having ChatGPT inside of Evernote. No surprise since the company collaborated with OpenAI to make it. As you can probably guess, you can use AI Assistant to search through your notes and notebooks, but the tool's utility goes beyond that. It allows you to interact with your documents and ask follow-up questions. AI Assistant can also generate new content, and you can decide how what it creates is integrated within your existing notes and notebooks. In my demo, Simionato used AI Assistant to add a flight ticket number to a note he had dedicated to an upcoming trip to Spain. Over time, Bending Spoons plans to make AI Assistant more powerful, with the company working on making it possible for the chatbot to assign tags to notes and perform bulk actions across multiple documents, among other capabilities. 

By default, Bending Spoons will enable AI Assistant for all paid users. A set of granular controls will allow people to decide what content AI Assistant can access. The agreement Bending Spoons has with OpenAI means the company can't train its models on data from Evernote users. Additionally, any data sent to the cloud for processing will only be held as long it takes to complete the user's request.  

Part of the reason Bending Spoons is taking a more forward approach with AI Assistant is due to user feedback. In 2023, the company released a tool called AI Search. Simionato says his team was "extremely careful" to add friction to the experience to ensure people felt Evernote was respecting their privacy. To that end, you had to dig into the app's settings to enable AI Search before it would work. According to Simionato, users found that setup was "too cumbersome," and with most people becoming more comfortable with AI tools, the company decided to make AI Assistant visible from the jump. 

The other new feature Bending Spoons is introducing today is one some Evernote users have been waiting for a decade: Semantic Search. Instead of trying to find a note using an exact word or phrase you used inside of it, Semantic Search allows Evernote to filter through your documents using context. Simionato demonstrates the feature in action by typing "Barcelona trip" into the new interface, and Evernote surfaces a note he created about a trip to Girona. If you're not familiar, Girona is a city about 62 miles northeast of Barcelona and part of Catalonia. The document Evernote pulled up had no mention of Barcelona in it, but that didn't prevent the tool from finding it. According to Simionato, this is the most requested feature of longtime Evernote users and one Bending Spoons is happy it can finally deliver on. 

Lastly, the company has expanded the app's built-in recording and transcription capabilities. Building on AI Transcribe, which Bending Spoons first introduced in 2024, Evernote can now record audio from both in-person and online meetings and automatically transcribe and summarize what's said. This feature should work in every language that Evernote currently supports. 

As mentioned, Evernote's new AI features are available to all paid users. The features will roll out to desktop and web first starting this month, with mobile to follow. Simionato said Bending Spoons is working on making a trail available to free users sometime in the next few months. For now, the price of Evernote's paid plans isn't going up, but Simionato noted the company plans to adjust pricing in the coming months, mostly to account for the fact there aren't many notable differences between the Personal and Professional tiers.

This article originally appeared on Engadget at https://www.engadget.com/evernotes-biggest-update-since-2020-goes-all-in-on-ai-180000340.html?src=rss

The best Prime Day vacuum deals: Save on machines from iRobot, Shark, Dyson and others

Robot vacuums are more than just a gimmick. Sure, we all enjoy anthropomorphizing a Roomba from time to time, but they're not gadgets for the sake of gadgetry — keeping your living space clean is an incredibly rewarding task to automate. The only problem is that robot vacuums, and frankly all decent vacuums, are expensive. Fortunately, the steep discounts offered on Amazon this October Prime Day make the purchase a lot more justifiable, whether you want a robot buddy or a traditional vacuum cleaner. In this article, we're sharing a constantly updated list of Prime Day vacuum deals worth taking advantage of.

Dyson V15 Detect Plus for $570 (33 percent off): Our top pick for the best cordless vacuum on the market right now, the V15 Detect has some of the strongest suction power you'll find in a stick vacuum, plus a lightweight design and a dustbin that can hold more dirt and debris than it might seem. This model comes with five cleaning attachments, including the Fluffy Optic cleaner head that has LEDs to illuminate the floor as you're vacuuming so you can better see where all the dust bunnies are.

Dyson Ball Animal Total Clean Upright Vacuum for $410 (38 percent off): Dyson is still the king of reinventing vacuums, and the bagless, hyper-maneuverable Ball Animal is a blast to use. The Ball design is based on ease of steering, but the hidden MVP is the sealing — from the head to the canister, not a hair is getting out of this one once it's in.

Shark AV2501AE AI Robot Vacuum for $459 (29 percent off): If you like the look of the AV2501S but have even more space to clean, the AV2501AE is also on sale. Its self-empty base can go a full 60 days before you have to dump it out, so it's ideal for large spaces, complex homes or areas that see heavy use. It's got the same features otherwise, including LiDAR mapping and two hours of autonomous work.

Shark PowerDetect NeverTouch for $450 (55 percent off): This robotic mop/vacuum combo is engineering so you'll almost never need to revisit it after you set it up. It can clean its own mop, refill its own water tank and empty its own dustbin for up to 30 days at a time. It's also equipped with air jets that blast dirt out of corners the vacuum can't fit into.

Shark Matrix Plus 2-in-1 for $300 (57 percent off, Prime exclusive): The Shark Matrix Plus takes the robot vacuum concept even further by working a mop into the design for hands-off wet cleaning. This model is self-cleaning, self-emptying, self-charging and capable of tackling ground-in stains on hard floors.

Shark Navigator Lift-Away Deluxe for $160 (27 percent off): Moving into manual vacuums, let's start with one of the best. The Shark Navigator Lift-Away is a champion at getting deeply ingrained crud out of carpets, but it's also capable of squaring away bare floors. You can switch between the two settings quickly, and the lift-away canister makes it easy to empty.

iRobot Roomba 104 Vac for $150 (40 percent off, Prime exclusive): This entry-level Roomba is a good pick for anyone who's new to owning a robot vacuum. It features a multi-surface brush and an edge-sweeping brush to clean all types of flooring, and it uses LiDAR navigation to avoid obstacles as it goes. The iRobot mobile app lets you control the robot, set cleaning schedules and more.

iRobot Roomba Plus 504 for $380 (36 percent off): For those looking to upgrade to a more advanced robot vacuum, the Roomba Plus 504 is a great next step. It can clean almost anything that might land on a home floor, and if it can't clean it, it can steer around it. Two brushes and strong suction get at tougher stains, and it even includes an app you can use to set cleaning zones and change suction force remotely.

Roborock Q10 S5+ Robot Vacuum and Mop for $300 (25 percent off): The Roborock Q10 S5+ is another great option for a vacuum-mop combo. It's skilled at knowing what type of floor it's on and deploying or retracting the mop appropriately, and comes with precision LiDAR mapping. You can change its settings remotely from a paired app.

Levoit LVAC-300 cordless vacuum for $250 (29 percent off, Prime exclusive): One of our favorite cordless vacuums, this Levoit machine has great handling, strong suction power for its price and a premium-feeling design. Its bin isn't too small, it has HEPA filtration and its battery life should be more than enough for you to clean your whole home many times over before it needs a recharge.

Shark CarpetXpert HairPro for $280 (24 percent off): Not all vacuums need to reinvent the wheel. The Shark CarpetXpert HairPro is the perfect midrange option, with a large brushroll for getting pet hair and other allergens out of your carpets. It's designed internally to suction and store hair without clogging, making it an even better fit for homes with pets.

Amazon Basics Upright Bagless Vacuum Cleaner for $55 (21 percent off): All right, nobody goes to Amazon Basics to be impressed, but we have to admit this vacuum exceeds expectations. It's light, it has a big dust reservoir and it comes with all the attachments you'll need for a reasonably sized apartment. The filter is also simple to remove and clean.

Black+Decker QuickClean Cordless Handheld Vacuum for $27 (33 percent off): Rounding out the list, we've got this small-but-mighty hand vacuum, perfect for crevices, shelves or cleaning out your car. It weighs about 1.4 pounds and hoovers up small messes in the blink of an eye. The lithium-ion battery stays charged for up to 10 hours.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-best-prime-day-vacuum-deals-save-on-machines-from-irobot-shark-dyson-and-others-151504087.html?src=rss

The best Amazon Prime Day deals under $50 include sales from Anker, Ring, Lego, Roku and others

You can snap up a lot of useful tech for less than $50 during Amazon’s October Prime Day sale. We’re seeing discounts on streaming devices, mini speakers, earbuds, power banks, wall chargers, and more. But this isn’t just random drek. The deals highlighted here are pulled from our own guides and reviews —these are gadgets and accessories we’ve tried ourselves and currently recommend. If you’re looking for small electronics and accessories (or if you just want to fill your Amazon cart without spending too much) this list of the best Prime Day deals under $50 is a great place to start.

UGREEN Nexode Pro 65W USB-C Flat Charger for $45 ($11 off with Prime): One of our picks from our guide to fast chargers, this Ugreen fella has three ports (one of which is USB-A) and a 65-watt max power output. The folding prongs make it even more compact. Just note that it’s a bit wide and may block adjacent outlets in a crowded power strip.

Instant Pot Vortex 2-QT Mini for $38 ($22 off with Prime): The budget model from our air fryer guide may not be large but its two-quart basket is enough to reheat leftovers for two or cook up a batch of frozen appetizers. And, because of its small size, it doesn’t take up a ton of space on your countertops — ideal for a small kitchen.

Amazon Echo Pop Kids for $33 ($17 off with Prime): Among the announcements for the new Echo devices, Amazon did not include new Echo Pop Kids models, so this is still the most current model for now. It’s good for smaller spaces, not necessarily audio fidelity, and it comes with six months of free access to Amazon Kids as well as early access to Alexa+.

Amazon Fire TV Stick 4K Max for $40 ($20 off): Amazon's most powerful streaming dongle supports 4K HDR content, Dolby Vision and Atmos and Wi-Fi 6E with double the storage of cheaper Fire TV sticks. It earned an honorable mention in our guide to streaming devices and also happens to make a good retro gaming emulator.

Blink Video Doorbell for $35 ($35 off): True, Amazon just announced new Blink devices, but those won’t hit the market until after Prime Day is over. If you want a video doorbell right now at an impressively low price, this should serve. We’ve tested Blink security devices before and have been impressed by what you get for such a small price.

Ring Battery Doorbell for $50 ($50 off): At $49.99 this juuust qualifies as an under $50 tech deal. If you don’t have doorbell wires at your front entrance, you can still have a camera to capture all the package deliveries and neighborhood animal sightings with the Ring Battery Doorbell. It records video in HD with more vertical coverage than the last model, so you can see people from head to toe. Just note that newer Ring devices are on the way. 

Blink Mini 2 security cameras (two-pack) for $35 ($35 off): This is currently the top budget pick in our guide to the best security cameras. The Mini 2 is a great option for indoor monitoring or you can put it outside with a weatherproof adapter, but since it needs to be plugged in, we like it for keeping an eye on your pets while you're away and watching over entry ways from the inside.

Blink Mini 2 security cameras (three-pack) for $50 ($50 off): A three pack is also on sale and squeezes in just under the $50 mark.

Lego Star Wars: The Mandalorian Paz Vizsla and Moff Gideon Battle for $17 (56 percent off): You can have a tiny version of the dramatic season-three battle from The Mandalorian to admire whenever you want. This 289-piece set is rated for kids seven and up, but no one will judge if you keep it for your own shelves.

Check out more Lego sets in our Prime Day guide to Lego deals.

Anker 622 5K magnetic power bank with stand for $34 ($14 off with Prime): This 0.5-inch thick power bank attaches magnetically to iPhones and won't get in your way when you're using your phone. It also has a built-in stand so you can watch videos, make FaceTime calls and more hands-free while your phone is powering up.

Ring Indoor Cam for $25 ($25 off): While we thought the Blink Mini 2 was a better overall indoor camera in our guide, we do like the Ring app, which is ideal for beginners. Plus you get access to the Ring Neighbors app which is a fascinating glimpse into your neighborhood’s Ring-captured events.

Amazon Smart Plug for $13 ($12 off): We named this the best smart plug for Alexa users because it hooks up painlessly and stays connected reliably. Use it to control lamps or your holiday lights using programs and schedules in the Alexa app, or just your voice by talking to your Echo Dot or other Alexa-enabled listener.

Levoit Mini Core-P air purifier for $40 ($10 off with Prime): This is the mini version of the top pick in our guide to air purifiers. It has a three-stage filter (pre, activated carbon and particle filters) though that particle filter is not a true HEPA filter. But it’s rated at 250 square feet and can help clear the air in your office or other small room.

8Bitdo Ultimate Bluetooth controller (Switch 2) for $45 ($25 off): If you’ve got a Switch 2 and are looking for a good controller to go with it, we think this is a good one. Recommended in our guide to the best Switch 2 accessories, it’s got an Xbox style configuration with the left stick placed higher on the game pad, which some prefer. The lowest price we tracked before this is $50.

Echo Pop smart speaker for $25 ($15 off): The half sphere Pop is the most affordable Echo speaker in Amazon’s lineup. The sound won’t be as full as its larger siblings, but will do a fine job of bringing Alexa’s help to smaller rooms. Just note that it went as low as $18 for Black Friday and October Prime Day last year.

Roku Streaming Stick Plus 2025 for $24 (40 percent off): This is our top pick for the best streaming device for accessing free and live content. The dongle supports 4K video and HDR and doesn’t need to be plugged into the wall for power. It’s a great way to access any streaming service you could ask for: Netflix, Prime Video, Disney+, HBO Max and many more.

Roku Streaming Stick HD for $18 ($12 off): If you don’t care about 4K (or your screen resolution isn’t that high anyway) you can still get the same simple-to-use Roku OS with this device. The best thing about Roku streaming sticks is the access to all the free content — so this is an affordable way to get it.

Leebein 2025 electric spin scrubber for $40 ($30 off with Prime): This is an updated version of the electric scrubber we love that makes shower cleaning easier than ever before. It comes with seven brush heads so you can use it to clean all kinds of surfaces, and its adjustable arm length makes it easier to clean hard-to-reach spots. It's IPX7 waterproof and recharges via USB-C.

Jisulife Life7 handheld fan for $25 ($4 off with Prime): This handy little fan is a must-have if you live in a warm climate or have a tropical vacation planned anytime soon. It can be used as a table or handheld fan and even be worn around the neck so you don't have to hold it at all. Its 5,000 mAh battery allows it to last hours on a single charge, and the small display in the middle of the fan's blades shows its remaining battery level.

Anker Soundcore Select 4 Go speaker for $26 ($9 off with Prime): This is one of our top picks for Bluetooth speaker. It gets pretty loud for its size and has decent sound quality. You can pair two together for stereo sound as well, and its IP67-rated design will keep it protected against water and dust.

Amazon Echo Spot for $45 ($35 off): Amazon brought the Echo Spot smart alarm clock back from the dead last year with a new design and improved speakers. In addition to being able to control smart home devices and respond to voice commands, the Echo Spot can also act as a Wi-Fi extender for those that have Eero systems. It went as low as $45 for Black Friday last year.

Samsung EVO Select microSD card (256GB) for $23 ($4 off): This Samsung card has been one of our recommended models for a long time. It's a no-frills microSD card that, while not the fastest, will be perfectly capable in most devices where you're just looking for simple, expanded storage.

Ring Pan-Tilt Indoor Cam $40 ($20 off): If you like the idea of being able to move the camera around to follow the action in your home, you may want to get a pan-and-tilt option like this one. We will again note that new Ring devices are on the way, but if you don’t have to have the latest thing and just want to see what your dog gets up to while you’re gone, you may want to snag this 33 percent discount.

Anker Soundcore Space A40 for $45 ($35 off): Our top pick for the best budget wireless earbuds, the Space A40 have surprisingly good ANC, good sound quality, a comfortable fit and multi-device connectivity.

JLab Go Air Pop+ for $17.49 ($12 off): JLab earbuds pop up in a few of our guides including the best running headphones and best budget buds. The Pop+ earbuds are smaller and lighter than the previous model, and the app’s preset EQ modes let you customize your sound. Total battery life with the case comes in at more than 35 hours.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-best-amazon-prime-day-deals-under-50-include-sales-from-anker-ring-lego-roku-and-others-120531317.html?src=rss

The best Prime Day TV deals: Save on sets from LG, Samsung, TCL and others

Black Friday remains the best time to grab a new TV at a discount, but Amazon’s latest October Prime Day sale looks like a decent opportunity to take the plunge if you need to upgrade right away. While the two-day event doesn't officially start until tomorrow, a number of well-regarded TVs from LG, Samsung, TCL and other manufacturers are at or near their best prices to date right now. You can find our full list of the best Prime Day TV deals below, along with a handful of discounts on streaming devices we like. We'll update this roundup as prices change and new offers arise over the course of the event.

TCL QM7K 55-inch Mini-LED TV for $598 ($100 off): If you’re willing to stretch your budget a little further, the TCL QM7K is a fairly comprehensive step up, with noticeably improved contrast, brightness and color volume. It has a 144Hz refresh rate with 288Hz support at 1080p, too, though it’s still limited to two HDMI 2.1 ports. (As with the QM6K, however, neither of those is an eARC port, so hooking up a soundbar won’t block a game console if you have multiple systems.) This is only $1 more than the all-time low for the 55-inch model. Other sizes are similarly discounted.

Hisense U8QG 65-inch Mini-LED TV for $998 ($500 off): Several reviews suggest that the Hisense U8QG ticks most of the requisite boxes for a LCD TV in 2025: robust local dimming and mini-LED backlighting, exceptionally high brightness, vibrant quantum-dot colors, a fast refresh rate (165Hz in this case), support for the major HDR formats and so on. It’s a higher-end option than something like the TCL QM6K or QM7K with superior brightness and contrast, though it still falls short of a good OLED TV when it comes to the latter. Like most LCD panels, it’ll also look a bit washed out if you view it from an angle. It has three HDMI 2.1 ports, which is one fewer than many other TVs in this price range, though it uniquely includes a USB-C video input if you want to hook up a gaming laptop or Nintendo Switch. (Just note that you won’t get VRR or HDR when using that.) You’d mainly get it over an OLED TV if you’re willing to trade some picture quality for something that’s better-suited in a bright room. This deal on the 65-inch model isn’t an all-time low, but it matches the best price we’ve tracked since July.

Samsung S90F 55-inch QD-OLED TV for $1,298 ($300 off): The Samsung S90F is an upper-tier model with a QD-OLED panel, which blends the usual perks of a quality OLED set — near-perfect contrast, wide viewing angles, clear motion, low input lag — with a layer of quantum dots. This helps it produce a wider gamut of more vivid colors compared to traditional WOLED TVs. It also comes with four HDMI 2.1 ports and has a fast refresh rate of 144Hz. It doesn’t support Dolby Vision HDR, however, and reviews we trust say that the LG C5, a competing WOLED model, retains darker black levels in a bright room. (The S90F has a more colorful image, though.) This deal marks a new low for the 55-inch model, while the 65-inch and 77-inch versions are also at all-time lows of $1,598 and $2,298, respectively. Just make sure you only buy the 55-, 65- or 77-inch model, as every other size in the US uses a lesser WOLED panel. Shady, we know.

LG C5 55-inch OLED TV for $1,297 ($503 off): The LG C5 can’t produce the same bold colors as a QD-OLED display like the Samsung S90F, but reviews almost universally agree that it’s an exceptional OLED TV otherwise. It should get brighter with non-HDR content, and as noted above it should produce deeper blacks in well-lit environments. It also has just about all the essential gaming features, plus it supports the popular Dolby Vision HDR format (but not HDR10+). If you need that, or if you want an OLED set in this price range for a bright-ish room, it’s well worth a look. This ties the lowest price we’ve found for the 55-inch model; other sizes are also on sale, but note that the 42- and 48-inch models can’t get as bright as the larger versions.

LG B5 55-inch OLED TV for $997 ($100 off): The B5 is LG’s entry-level OLED TV for 2025, and as such it’s a level below the C5 in terms of brightness and color performance. It’s technically limited to a 120Hz refresh rate instead of 144Hz as well, though that isn’t a huge deal right now unless you plan on hooking up a gaming PC. If anything, last year’s LG C4 — which isn’t seriously discounted as of this writing — is a better value for most on the whole. But if you just want to save cash, the B5 still provides most of the core benefits of an OLED display at a lower price. This is a new all-time low for the 55-inch variant.

Sony Bravia 8 II 65-inch QD-OLED TV for $2,998 ($502 off): It’s certainly not cheap, but the Sony Bravia 8 II has earned plaudits for its excellent image processing, upscaling and overall accuracy alongside the expected color, contrast and motion benefits of its QD-OLED display. This should help it make lots of movies and shows look closer to their original intent. It also uses the handy Google TV interface. Outside of an extremely brief dip in June, this deal matches the best price to date for the 65-inch version. That said, if you can’t stomach the high price, other reviews note that the older Sony A95L offers similar performance a bit less, while more recent competitors like the LG G5 and Samsung S95F can get noticeably brighter (even if they’re not always as accurate). Those two should be better for gaming as well, as the Bravia 8 II only has two HDMI 2.1 ports — one of which is an eARC port for soundbars — and its input lag is slightly higher.

Roku Streaming Stick 4K for $30 ($20 off): The Streaming Stick 4K is worth considering if you prefer a stick-style streamer that plugs directly into your TV, or if you’re partial to Roku’s app-centric interface, which many find simpler to navigate than the content-heavy UIs pushed by Google and Amazon. It supports Apple AirPlay and the major HDR formats, and its performance remains quick enough for most. You’ll have to deal with some ads, though, and no Roku player supports the Twitch app. This deal is $5 more than the device’s all-time low but ties the best price we’ve seen in several years.

Roku Streaming Stick Plus for $24 ($16 off): The Streaming Stick Plus is cheaper alternative to the Streaming Stick 4K with no Dolby Vision support and no long-range Wi-Fi extender built in. The latter means it may be less reliable if you don’t get a consistent connection in your TV’s room. (Both sticks are limited to Wi-Fi 5, though.) That said, this model can typically be powered straight from a TV’s USB port instead of requiring a separate power supply. If you can live with the sacrifices, it may not be worth paying extra. This deal represents a new all-time low.

Amazon Fire TV Stick 4K for $25 ($25 off): The standard Amazon Fire TV Stick 4K offers the same core experience as the pricier Fire TV Stick 4K Max, only it comes with a slightly slower processor, half the storage (8GB) and Wi-Fi 6 instead of Wi-Fi 6E. For most people just looking for a casual streamer on the cheap, those shouldn’t be huge losses. This model is also more powerful than the just-announced Fire TV 4K Select, though its Fire OS interface can still be messy and ad-heavy, with special emphasis on Amazon’s own services. This deal is $3 more than the stick’s all-time low, though it matches the best price we’ve seen since Black Friday last year.

Amazon Fire TV Stick HD for $18 ($17 off): The Fire TV Stick HD is the budget pick in our guide to the best streaming devices. It can only stream up to 1080p, and it can run a bit choppier than the 4K models since it has a slower chipset and half the RAM (1GB). The usual issues with the Fire TV interface still apply here too. But if you just want to add streaming apps to an aging TV or basic monitor for as little cash as possible, it should get the job done. This discount ties the device's lowest price to date.

This article originally appeared on Engadget at https://www.engadget.com/deals/the-best-prime-day-tv-deals-save-on-sets-from-lg-samsung-tcl-and-others-180051219.html?src=rss

Instagram will start handing out awards to select creators

Instagram is launching its own awards program, according to The Hollywood Reporter. This is sort of like the Oscars, but for short-form social media content. There isn't an awards ceremony, a comedian host or anything like that. Instead, the 25 winners will get physical rings.

These rings will be designed by fashionista Grace Wales Bonner. Maybe they'll be worth a few bucks, as the recognition doesn't include a financial payout of any kind. However, winners will get a digital replica of a ring to put on their Instagram profile and stories.

Instagram

Honorees also receive the ability to customize the background color of their profiles, creating a unique gradient. Eva Chen, the head of fashion partnerships at Instagram, says this kind of customization is "something that we haven’t really ever done before." Myspace let people do this more than 20 years ago and Friendster before that.

Instagram has three billion monthly active users and there are only 25 recipients of the aforementioned ring. Judges will whittle down the list down to thousands and then hundreds before settling on the winners. These judges include platform head honcho Adam Mosseri, along with folks like Spike Lee and actress Yari Shahidi.

There are no official categories like with other awards, but the winners will be chosen from across different topics and interest groups. “The criteria was really tough,” Chen says. “But really, we were looking for people who take creative chances and who are always pushing the envelope and thinking about new ways to talk to their audience.”

Instagram hopes this will become a regular thing, just like those other awards. "We’d love to see how it’s received and how the winners enjoy the literal Instagram spotlight,” Chen says. Winners will be announced on October 16. 

This article originally appeared on Engadget at https://www.engadget.com/social-media/instagram-will-start-handing-out-awards-to-select-creators-170531135.html?src=rss

Apple's AirPods 4 drop to $90 for Prime Day

Those on a tight budget and those who prefer open-ear AirPods will want to check out the latest discount on the AirPods 4. Apple's entry-level wireless earbuds are down to $90 for October Prime Day, or 30 percent off their normal rate. That's also the lowest we've seen them.

When Apple updated its standard AirPods in 2024, it released two models: one with active noise cancellation (ANC) and one without. We consider the non-ANC models to be the best budget AirPods you can get, so they're an even better buy at this sale price. The ANC versions are also discounted at the moment, so you can pick them up for $140 instead of the usual $180.

The AirPods' H2 chip brings a bunch of Pro-adjacent features to the standard model. That list includes Adaptive Audio, Conversation Awareness, Transparency, Personalized Spatial Audio and wireless charging. For calls, it supports Voice Isolation, which helps filter out background noise and make it easier for you to be heard. And if that call is via FaceTime, it supports Apple's 48kHz "cinema-quality" audio.

Sound is improved over the AirPods 3. "There's more low-end tone than the previous model right out of the gate, and the punchier treble opens the sound stage so it's wider and more immersive," Engadget's Billy Steele wrote. "The improved frequency response is on full display with Dolby Atmos content (movies, music, and TV), but the AirPods 4 are also more sonically adept with non-spatial tunes and video."

You can expect around five hours of battery life on a single charge. With the ANC model, that estimate drops to around four hours with the marquee feature activated. With either version, you can drop them in the charging case to extend their total time to around 30 hours. And hey, it's good to give your ears a rest now and then anyway.

This article originally appeared on Engadget at https://www.engadget.com/deals/apples-airpods-4-drop-to-90-for-prime-day-130033169.html?src=rss

Assassin's Creed Mirage's free DLC Valley of Memory will be released next month

Remember the free DLC that was teased for Assassin's Creed Mirage? Ubisoft has officially revealed it, Valley of Memory, is coming next month. The free expansion is set in 9th-century Arabia. Specifically, […]

Thank you for being a Ghacks reader. The post Assassin's Creed Mirage's free DLC Valley of Memory will be released next month appeared first on gHacks Technology News.

Microsoft is killing another workaround to set up Windows 11 without a Microsoft account

Microsoft is making another change to the Windows Setup experience, also called Out-of-the-box experience (OOBE). It is removing a known workaround that allowed users to bypass the Microsoft account requirement while installing […]

Thank you for being a Ghacks reader. The post Microsoft is killing another workaround to set up Windows 11 without a Microsoft account appeared first on gHacks Technology News.

Just Look How Bloody Gross Universal’s ‘Terrifier’ Haunted House Is

Halloween Horror Nights went all out for Art the Clown at Universal Studios Orlando and Universal Studios Hollywood.

Prince William Walking Around the TARDIS Set Is a Nice Reminder That ‘Doctor Who’ Is Not Dead Yet

Don't expect him to take over from Billie Piper any time soon. Maybe. Anything goes after 'The Reality War' frankly.

OpenAI Goes All-In on Vibe Coding, Says ‘Mature Experiences’ Are on the Horizon

Vibe-coded apps offering 18+ content is a recipe for success, right?

Trump’s Tariffs Are Now Screwing Up America’s Libraries

Add that to the list of things that have been "improved" by tariffs.

‘Reading Rainbow’ Is Back, This Time Without ‘Star Trek’ Legend LeVar Burton

The reboot aims to bring back the book magic of the beloved 'Star Trek' star's time on the PBS series.

Buckle Up, the Smart Glasses Backlash Is Coming

Is this #glassholes 2.0?

Here’s Why ‘A Knight of the Seven Kingdoms’ Won’t Re-Use the ‘Game of Thrones’ Theme (Again)

'House of the Dragon' episodes open with that familiar music but the latest Westeros-set series will try something new.

Taylor Swift, Defender of Artist Ownership, Allegedly Uses AI in Videos

Taylor Swift (AI Version)

‘TMNT: Mutant Mayhem’ Is Returning Sooner Than You Think (And Not in the Way You Thought)

'Teenage Mutant Ninja Turtles: Chrome Alone 2 — Lost in New Jersey' is a short film hitting theaters this year, and here's a first look.

Sean Duffy Says He Doesn’t Want Air Traffic Controllers Working as Uber Drivers During Shutdown

Sean Duffy tried to blame the Democrats for the government shutdown.

Roborock’s Ready for Prime Time With Savings of Up to 45% Off Their Award-Winning Robot Vacuums

Amazon's big Fall Prime Day event is the perfect opportunity to score a huge deal on one of Roborock's AI-powered cleaning machines.

Here’s Your Chance to Own Lucille, the Most Gruesome Prop From ‘The Walking Dead’

Negan's bat is going up for auction along with tons of other props, costumes, set pieces, and memorabilia from AMC's 'Walking Dead' shows.

Teen Arrested After Asking ChatGPT How to Kill His Friend, Police Say

AI meets surveillance state in America's new educational dystopia.

‘The Leftovers’ Is Still One of TV’s Great Miracles

Eight years later, the questions and emotions raised by the HBO series feel as impactful as ever.

A New ‘Bugonia’ Trailer Leans Heavy Into That Alien Question

Emma Stone and Jesse Plemons star in the latest from Yorgos Lanthimos, which opens October 24.

Hundreds of Hikers Still Stranded on Mount Everest After Brutal Blizzard Hits

Hikers suddenly found themselves in a life-threatening situation as a rare October blizzard buried their camps in snow.

The 2x2 Matrix That Explains AI Adoption

AI-powered products seem to come in different flavors. Some are big and dramatic, talking about reimagining the world and a bold new future.

Read All

Mistral’s Devstral Beats Giants on SWE-Bench, Sets New Bar for Open-Weight Coding Models

Mistral's new Devstral model tops SWE-Bench Verified for open-weight LLMs.

Read All

The Billion Dollars Problem: Can Quantum Qubits Cure Pharma's Innovation Slowdown?

A look at the future of drug discovery, where hybrid quantum-AI systems are already starting to reshape how new medicines are developed.

Read All

From Regulation to Resilience: Best Practices for Securing Healthcare Data in an AI Era

Healthcare’s intrusion frequency is up 23% year-over-year, according to the CrowdStrike 2025 Threat Hunting Report, and artificial intelligence is only making it easier for bad actors to gain access to an organization’s network. With patient outcomes and data at risk, health systems need all the resources and support they can get to protect their environments from malicious attacks. There are several proposed guidelines and regulations seeking to bolster cybersecurity in the healthcare sector, including the Healthcare Cybersecurity Act of 2025, America’s AI Action Plan and an updated HIPAA…

Claude Sonnet 4.5 Ranked Safest LLM From Open-Source Audit Tool Petri

Claude Sonnet 4.5 has emerged as the best-performing model in ‘risky tasks’, narrowly edging out GPT-5 in early evaluations by Petri --- Anthropic’s new open-source AI auditing tool.

Dreamer 4: Learning to Achieve Goals from Offline Data Through Imagination Training

Researchers from DeepMind have described a new approach for teaching intelligent agents to solve complex, long-term tasks by training them exclusively on video footage rather than through direct interaction with the environment. Their new agent, called Dreamer 4, demonstrated the ability to mine diamonds playing Minecraft after being trained on videos, without ever actually playing the game.

Extreme Rains Hit India's Premier Darjeeling Tea Estates

Torrential rains that triggered deadly landslides and floods in India's Darjeeling region also destroyed swathes of premier tea estates, officials said Tuesday.

UK's Starmer Condemns Pro-Palestinian Protests On Oct 7 Anniversary

British Prime Minister Keir Starmer urged students to skip pro-Palestinian protests planned for the second anniversary on Tuesday of Hamas's October 7, 2023, attack in Israel, suggesting they were disrespectful.

Japan's Takaichi Eyes Expanding Coalition, Reports Say

Japan's incoming new prime minister Sanae Takaichi has approached an opposition party with a view to expanding the governing coalition to give it a working majority in parliament, reports said Tuesday.

Canadian PM To Visit White House To Talk Tariffs

Canadian Prime Minister Mark Carney is slated to visit Washington again Tuesday in hopes of convincing President Donald Trump to ease US tariffs that are negatively impacting Canada's economy.

EU Channels Trump With Tariffs To Shield Steel Sector

The European Union will take a page out of US President Donald Trump's tariff playbook on Tuesday with a plan to hike levies and slash quotas for imported steel -- to protect the bloc's struggling industry from cheap Chinese competition.

Open AI's Fidji Simo Says AI Investment Frenzy 'New Normal,' Not Bubble

The dizzying investments in artificial intelligence infrastructure do not constitute a bubble but rather represent today's "new normal" to meet skyrocketing user demand, Fidji Simo, OpenAI's de facto number two, said on Monday.

Tokyo Stocks Hit New Record As Asian Markets Extend Global Rally

Japanese stocks hit another record Tuesday, building on the previous day's surge following the election of a pro-stimulus advocate to lead the country's ruling party, while gold also pushed to a new high amid the US government shutdown and French political upheaval.

Computer Advances And 'Invisibility Cloak' Vie For Physics Nobel

A maths theory powering computer image compression, an "invisibility cloak" or the science behind the James Webb Space Telescope are some achievements that could be honoured when the Nobel physics prize is awarded Tuesday.

Nobel Literature Buzz Tips Swiss Postmodernist, Australians For Prize

Swiss postmodernist novelist Christian Kracht and Australia's Gerald Murnane and Alexis Wright are among the favourites for the Nobel literature prize, experts told AFP ahead of the Swedish Academy's much-anticipated announcement on Thursday.

China Exiles In Thailand Lose Hope, Fearing Beijing's Long Reach

With tears rolling down his cheeks, detained Chinese political exile Zhou Junyi struggles to make his voice heard over dozens of families visiting Bangkok's main immigration detention centre.

Israel Marks October 7 Anniversary As Talks Held To End Gaza War

Israel marks the second anniversary of the October 7, 2023 attack on Tuesday, as Hamas and Israeli negotiators hold indirect talks to end the two-year war in Gaza under a US proposed peace plan.

Indians Lead Drop In US University Visas

The United States issued nearly one-fifth fewer student visas in August following a crackdown by President Donald Trump, led by a steep drop for India which was overtaken by China as top country of origin, data showed Monday.

Colombia's Armed Groups 'Expanding,' Warns Watchdog

Colombia's once-weakened armed groups are growing their power again, the country's ombudswoman told AFP Monday, as the country heads toward its most violent year in a decade.

UNESCO Board Backs Egyptian For Chief After US Row

UNESCO executives voted Monday to make an Egyptian ex-antiquities minister the UN culture body's next leader, as it grapples with accusations of pro-Israel bias that prompted the United States to say it would quit the organisation.

UEFA 'Reluctantly' Approves European League Games In US, Australia

European football governing body UEFA on Monday said it had "reluctantly" approved the staging of a Spanish La Liga and an Italian Serie A match in the United States and Australia respectively despite fan protests.

Kristi Noem Says Chicago Is 'a War Zone'; Illinois Governor Fires Back: 'They're the Ones Making It One'

Homeland Security Secretary Kristi Noem on Sunday called Chicago "a war zone," defending the Trump administration's deployment of federal forces

Fifth Third Bank Buys Comerica: What the $10.9 Billion Deal Really Means for US Banking and Your Money

Fifth Third Bancorp has struck a $10.9 billion all‑stock deal to acquire Comerica, creating the ninth‑largest U.S. bank. What this merger means for regional banking.

'One Piece' Chapter 1162 Release Date, Time, Spoilers and Everything We Know So Far

'One Piece' Chapter 1162 finally gets a confirmed release date: Tuesday, October 14th. Author Eiichiro Oda is recovering after a health scare.

OpenAI's Sora Is the Ultimate Disinformation Weapon, Experts Warn - Here's How

Experts warn OpenAI's Sora could become the most powerful disinformation tool yet, blurring truth and fiction through hyper-real AI-generated video.

Trump Announces 25% Tariff On Imported Heavy Trucks From November 1

The new tariff is aimed at encouraging more vehicle manufacturing within the United States. Trump made the announcement on his social media platform, Truth Social, saying foreign manufacturers should "build in America or pay the price."

Why October Has a Reputation for Market Turbulence — and What the Data Really Says

October's reputation as a 'jinx month' for markets stems from history's biggest crashes — but the data tells a more balanced story about volatility, investor psychology, and long-term returns.

Top Venezuelan Official Claims U.S. 'Threats' Have Made Maduro 'A Thousand Times Stronger'

Venezuelan Attorney General Tarek William Saab said that the "threats" authoritarian President Nicolas Maduro has received from Trump administration officials have made him "a thousand times stronger"

He Drugged and Violated Me': Diddy's Ex-Intern Claims Music Mogul's Brutal Attack Left Him with HIV

A man suing as 'John Doe' claims he was drugged and assaulted at a 2014 Val Verde after-party and later tested HIV-positive; the suit, filed in Los Angeles Superior Court, adds to the dozens of civil cases against Sean 'Diddy' Combs as he begins a federal sentence.

'One Punch Man' Season 3 Release Date, Time, Episode Count, Storyline, Where to Watch and Everything We Know So Far

Get the full release schedule for 'One Punch Man' Season 3, streaming on Crunchyroll. New episodes start October 12.

Miley Cyrus Shares Emotional Moment Dad Billy Ray Cried Over Her Song About Their Rift

Miley Cyrus revealed on CBS Sunday Morning that her father Billy Ray Cyrus cried after hearing her song 'Secrets' from her album Something Beautiful, written as a peace offering to mend their strained relationship after years of family tension.

10 Sifted Summit sessions not to miss

无摘要

The fastest-growing hardware startup teams in Europe

无摘要

Companies are making the same mistake with AI that Tesla made with robots

Some early experiments with AI are revealing the technology's shortcomings - and, by extension, the value of human workers.

This premium Alienware Aurora gaming PC is $600 off for Prime Day

The Alienware Aurora ACT1250 PC is a premium gaming computer with next-gen hardware. During Amazon Prime Day, it's $600 off.

Save 15% off this sleek gaming laptop from Alienware

The Alienware 16 Aurora pulls double duty, excelling as both a gaming PC and an office laptop. During October Prime Day, it received a $220 discount.

Skipping the AirPods Pro 3? Your best alternative are at their lowest price ever

The Powerbeats Pro 2 have similar features to the AirPods Pro 3 - and they're 28% cheaper during Amazon Prime Day.

Why this $350 Google Pixel 9a deal is the only one I'm considering for Prime day

It's not the flashiest model out there, but you really can't go wrong with this phone at this price.

The OnePlus 13 is back at its lowest price yet for Amazon Prime Day

The OnePlus 13 boasts top-tier performance from its Snapdragon 8 Elite chipset. During Amazon Prime Day, the 512GB model is on sale for $800.

One of my favorite immersive gaming PCs is 11% off during Amazon Prime Day

Lenovo's new Legion 5i offers lots of bang for your buck, with the latest in gaming hardware for a sweet sale price during Amazon Prime Day.

Google Pixel Tablet might be the best tablet to buy right now, thanks to an all-time low price

The Google Pixel Tablet is your Android device deal of the season under $250.

My favorite Samsung foldable of all time just hit a record-low price on Amazon - how to get $400 off

The Samsung Galaxy Z Fold 7 hits almost every mark - it's the most complete foldable I've tested to date.

This CyberPowerPC tower is 15% off and perfect for gamers on a budget

This midrange gaming PC houses the latest gaming hardware inside a sleek case. As part of Amazon Prime Day, it hit its lowest price yet.

My favorite modular GoPro action camera is compact and $150 off for Prime Day

The GoPro Hero 13 Black Ultra Wide edition is down from $480 to $330 this sale season.

The Samsung Galaxy S25 just hit an all-time-low price, making it an instant buy

The Samsung Galaxy S25 is now available at FE prices with a $140 discount for a limited time for Amazon Prime Day.

The only lithium button battery brand I recommend now - for serious safety reasons

Button cell battery ingestion causes thousands of injuries and along with a number of deaths each year in the US. Here's the only brand you should buy now - and why.

Looking for a good deal on a new gaming laptop? I found one that's $300 off

Asus' ROG Strix G16 is one of the sleekest 16-inch gaming laptops out there, and right now it's on sale as part of Amazon's Big Prime Deal Days event.

ChatGPT wants to act more like an OS -- as it transforms into an app platform

While ChatGPT started as a chatbot, it is evolving to something more like Mac, Windows, or Android -- and those ambitions became clear at OpenAI DevDay 2025.

Best Amazon Prime Day deals 2025: Our 110+ favorite sales this October

Prime Day kicks off tomorrow, but you don't want to miss these excellent deals on Apple, Google, Samsung, and more.

Best October Prime Day TV deals 2025: All-time-low prices from Samsung, LG, and more

Amazon's October Prime Day sale starts tomorrow, but you don't want to miss these great early deals on TVs from Samsung, LG, Hisense, and more.

Best Amazon Prime Day headphones deals October 2025: Save on Beats, AirPods, Sony, and more

Amazon's October Prime Day kicks off tomorrow, but excellent deals are already available if you are looking for your next pair of headphones or earbuds.

Best Amazon Prime Day Apple deals in October 2025: Save up to 32% on MacBook, AirPods, and more

Amazon's October Prime Day is tomorrow, but there are already some great deals on Apple tech, including iPhones, AirPods, iPads, and Apple Watches.

Looking for an affordable smartwatch? This popular Fitbit is $70 off

The Fitbit Versa 4 is an affordable smartwatch with all the health and fitness metrics you could ask for. It's on sale for $129 for Amazon Prime Day.

Companies are making the same mistake with AI that Tesla made with robots

Some early experiments with AI are revealing the technology's shortcomings - and, by extension, the value of human workers.

This premium Alienware Aurora gaming PC is $600 off for Prime Day

The Alienware Aurora ACT1250 PC is a premium gaming computer with next-gen hardware. During Amazon Prime Day, it's $600 off.

Save 15% off this sleek gaming laptop from Alienware

The Alienware 16 Aurora pulls double duty, excelling as both a gaming PC and an office laptop. During October Prime Day, it received a $220 discount.

Skipping the AirPods Pro 3? Your best alternative are at their lowest price ever

The Powerbeats Pro 2 have similar features to the AirPods Pro 3 - and they're 28% cheaper during Amazon Prime Day.

Why this $350 Google Pixel 9a deal is the only one I'm considering for Prime day

It's not the flashiest model out there, but you really can't go wrong with this phone at this price.

The OnePlus 13 is back at its lowest price yet for Amazon Prime Day

The OnePlus 13 boasts top-tier performance from its Snapdragon 8 Elite chipset. During Amazon Prime Day, the 512GB model is on sale for $800.

One of my favorite immersive gaming PCs is 11% off during Amazon Prime Day

Lenovo's new Legion 5i offers lots of bang for your buck, with the latest in gaming hardware for a sweet sale price during Amazon Prime Day.

Google Pixel Tablet might be the best tablet to buy right now, thanks to an all-time low price

The Google Pixel Tablet is your Android device deal of the season under $250.

My favorite Samsung foldable of all time just hit a record-low price on Amazon - how to get $400 off

The Samsung Galaxy Z Fold 7 hits almost every mark - it's the most complete foldable I've tested to date.

This CyberPowerPC tower is 15% off and perfect for gamers on a budget

This midrange gaming PC houses the latest gaming hardware inside a sleek case. As part of Amazon Prime Day, it hit its lowest price yet.

My favorite modular GoPro action camera is compact and $150 off for Prime Day

The GoPro Hero 13 Black Ultra Wide edition is down from $480 to $330 this sale season.

The Samsung Galaxy S25 just hit an all-time-low price, making it an instant buy

The Samsung Galaxy S25 is now available at FE prices with a $140 discount for a limited time for Amazon Prime Day.

The only lithium button battery brand I recommend now - for serious safety reasons

Button cell battery ingestion causes thousands of injuries and along with a number of deaths each year in the US. Here's the only brand you should buy now - and why.

Looking for a good deal on a new gaming laptop? I found one that's $300 off

Asus' ROG Strix G16 is one of the sleekest 16-inch gaming laptops out there, and right now it's on sale as part of Amazon's Big Prime Deal Days event.

ChatGPT wants to act more like an OS -- as it transforms into an app platform

While ChatGPT started as a chatbot, it is evolving to something more like Mac, Windows, or Android -- and those ambitions became clear at OpenAI DevDay 2025.

Best Amazon Prime Day deals 2025: Our 110+ favorite sales this October

Prime Day kicks off tomorrow, but you don't want to miss these excellent deals on Apple, Google, Samsung, and more.

Best October Prime Day TV deals 2025: All-time-low prices from Samsung, LG, and more

Amazon's October Prime Day sale starts tomorrow, but you don't want to miss these great early deals on TVs from Samsung, LG, Hisense, and more.

Best Amazon Prime Day headphones deals October 2025: Save on Beats, AirPods, Sony, and more

Amazon's October Prime Day kicks off tomorrow, but excellent deals are already available if you are looking for your next pair of headphones or earbuds.

Best Amazon Prime Day Apple deals in October 2025: Save up to 32% on MacBook, AirPods, and more

Amazon's October Prime Day is tomorrow, but there are already some great deals on Apple tech, including iPhones, AirPods, iPads, and Apple Watches.

Looking for an affordable smartwatch? This popular Fitbit is $70 off

The Fitbit Versa 4 is an affordable smartwatch with all the health and fitness metrics you could ask for. It's on sale for $129 for Amazon Prime Day.

Google DeepMind Introduces CodeMender: A New AI Agent that Uses Gemini Deep Think to Automatically Patch Critical Software Vulnerabilities

What if an AI agent could localize a root cause, prove a candidate fix via automated analysis and testing, and proactively rewrite related code to eliminate the entire vulnerability class—then open an upstream patch for review? Google DeepMind introduces CodeMender, an AI agent that generates, validates, and upstreams fixes for real-world vulnerabilities using Gemini “Deep […]

The post Google DeepMind Introduces CodeMender: A New AI Agent that Uses Gemini Deep Think to Automatically Patch Critical Software Vulnerabilities appeared first on MarkTechPost.

Building a Human Handoff Interface for AI-Powered Insurance Agent Using Parlant and Streamlit

Human handoff is a key component of customer service automation—it ensures that when AI reaches its limits, a skilled human can seamlessly take over. In this tutorial, we’ll implement a human handoff system for an AI-powered insurance agent using Parlant. You’ll learn how to create a Streamlit-based interface that allows a human operator (Tier 2) […]

The post Building a Human Handoff Interface for AI-Powered Insurance Agent Using Parlant and Streamlit appeared first on MarkTechPost.

OpenAI Debuts Agent Builder and AgentKit: A Visual-First Stack for Building, Deploying, and Evaluating AI Agents

OpenAI has released AgentKit, a cohesive platform that packages a visual Agent Builder, an embeddable ChatKit UI, and expanded Evals into a single workflow for shipping production agents. The launch includes Agent Builder in beta and the rest generally available. What’s new? Agent Builder (beta). A visual canvas for composing multi-step, multi-agent workflows with drag-and-drop […]

The post OpenAI Debuts Agent Builder and AgentKit: A Visual-First Stack for Building, Deploying, and Evaluating AI Agents appeared first on MarkTechPost.

A New Agency-Focused Supervision Approach Scales Software AI Agents With Only 78 Examples

Do curated, tool-grounded demonstrations build stronger software agents than broad piles of generic instruction data? A team of researchers from Shanghai Jiao Tong University and SII Generative AI Research Lab (GAIR) proposes LIMI (“Less Is More for Agency”), a supervised fine-tuning method that turns a base model into a capable software/research agent using 78 samples. […]

The post A New Agency-Focused Supervision Approach Scales Software AI Agents With Only 78 Examples appeared first on MarkTechPost.

New prediction model could improve the reliability of fusion power plants

The approach combines physics and machine learning to avoid damaging disruptions when powering down tokamak fusion machines.

Printable aluminum alloy sets strength records, may enable lighter aircraft parts

Incorporating machine learning, MIT engineers developed a way to 3D print alloys that are much stronger than conventionally manufactured versions.

The reality of wholesale cheating with AI

By and large, it appears school and work assignments are not being outsourced entirely to ChatGPT. A new working paper by David Deming, Danoff Dean of Harvard College, uncovers the more mundane realities of people's AI habits.

The reality of wholesale cheating with AI

By and large, it appears school and work assignments are not being outsourced entirely to ChatGPT. A new working paper by David Deming, Danoff Dean of Harvard College, uncovers the more mundane realities of people's AI habits.

Elon Musk Gambles on Sexy A.I. Companions

In July, the billionaire’s A.I. company introduced two sexually explicit chatbots, pushing a new version of intimacy.

OpenAI Agrees to Use Computer Chips From AMD

Weeks after a $100 billion agreement to use computer chips from Nvidia, OpenAI has signed a similar deal with a rival chipmaker.

Supreme Court, for Now, Rejects Google Bid to Block Changes to App Store

The emergency order is the latest turn in a longstanding legal dispute between the tech giant and the creator of the popular game Fortnite.

Can Cory Doctorow’s Book ‘Enshittification’ Change the Tech Debate?

Cory Doctorow’s new book looks to offer comfort, and solutions, to the inescapable feeling that digital platforms have gotten worse.

Tamron’s $1,149 70-180mm f/2.8 Lens Completes the ‘G2 Trinity’ on Nikon Z

Tamron is bringing its improved 70-180mm f/2.8 Di III VC VXD G2 lens to the Nikon Z mount, completing the company's f/2.8 "G2 Trinity" for Nikon Z photographers.

[Read More]

Lexar’s New Magnetic SSD Is Probably Good, Just Not as Good as They Claim

Lexar has announced the ES5 MagSafe-compatible external SSD that promises extremely fast read and write speeds when used in tandem with a Samsung Galaxy S25 or Apple iPhone 17 series phone. And while it probably will work just fine, the promised read and write speeds are a bit misleading about what users can actually expect.

[Read More]

Laowa’s New 20mm T2 Proteus 2x Anamorphic Lens Is its Widest Yet

Venus Optics has expanded its Proteus series with a new ultra-wide 20mm T2 2X Anamorphic lens. Set to impress filmmakers with unique features, the 20mm T2 lens offers a blend of cinematic quality and versatility, promising to be a valuable tool for professionals in the film industry, particularly those working in tight spaces or requiring wide-angle shots.

[Read More]

This $100 Studio Mic Is the First to Support USB, XLR, and Wireless Connectivity

The Maono PD200W is what the company calls the first dynamic podcasting microphone with "hybrid triple-mode connectivity," able to transmit a signal via USB, XLR, and a 2.4GHz wireless signal.

[Read More]

Photographers’ Incredible Pangolin Photos Help Save Endangered Animals

The Remembering Wildlife photography book series has released its 10th book, 10 Years of Remembering Wildlife. The series has raised nearly $1.6 million for wildlife conservation to date and features incredible wildlife photos from some of the world's most talented and renowned photographers. This milestone book focuses on pangolins, one of Earth's most threatened species.

[Read More]

The Limits of Tech and Why Smartphones Don’t Operate Under the Same Rules as ‘Real’ Cameras

While smartphone cameras and standalone cameras (which are sometimes referred to as "real" cameras) are often marketed using the same terms, the way they each function is very different. In a video published to TED-Ed, electrical engineer Rachel Yang simplifies and explains why smartphones don't operate by the same rules that standalone cameras do.

[Read More]

CIPA Shows Camera Sales’ Summer Slump Worse Than the Last Two Years

CIPA released updated camera shipment data, adding camera and lens shipment data through August. Although camera shipments were relatively high for much of the year, there was a significant dip from July to August, which was larger than the typical "summer slump" dip observed annually.

[Read More]

runCLAUDErun

Scheduler for Claude Code (macOS)

      <p>
        <a href="https://www.producthunt.com/products/runclauderun?utm_campaign=producthunt-atom-posts-feed&amp;utm_medium=rss-feed&amp;utm_source=producthunt-atom-posts-feed">Discussion</a>
        |
        <a href="https://www.producthunt.com/r/p/1023657?app_id=339">Link</a>
      </p>

Pulse

AI powered tasks, notes, and calendar in one workspace

      <p>
        <a href="https://www.producthunt.com/products/pulse-24?utm_campaign=producthunt-atom-posts-feed&amp;utm_medium=rss-feed&amp;utm_source=producthunt-atom-posts-feed">Discussion</a>
        |
        <a href="https://www.producthunt.com/r/p/1023767?app_id=339">Link</a>
      </p>

TickTag

Smarter time tracking with tags

      <p>
        <a href="https://www.producthunt.com/products/ticktag?utm_campaign=producthunt-atom-posts-feed&amp;utm_medium=rss-feed&amp;utm_source=producthunt-atom-posts-feed">Discussion</a>
        |
        <a href="https://www.producthunt.com/r/p/1023726?app_id=339">Link</a>
      </p>

PromptCompose

The command center for your AI prompts.

      <p>
        <a href="https://www.producthunt.com/products/promptcompose?utm_campaign=producthunt-atom-posts-feed&amp;utm_medium=rss-feed&amp;utm_source=producthunt-atom-posts-feed">Discussion</a>
        |
        <a href="https://www.producthunt.com/r/p/1023721?app_id=339">Link</a>
      </p>

Danish energy player Ørsted raises $9.35bn in discounted share issue

As the Trump administration plays havoc with the wind industry in the US, Denmark’s Orsted has raised $9.35bn to shore up its finances.

Read more: Danish energy player Ørsted raises $9.35bn in discounted share issue

Qualtrics to acquire Press Ganey Forsta in $6.75bn deal

The acquisition will enable Qualtrics to expand its AI capabilities in a space that is increasingly adopting advanced methods and technologies.

Read more: Qualtrics to acquire Press Ganey Forsta in $6.75bn deal

Dublin fintech launches AI tax platform, aims to raise €500,000

Tax doesn’t have to be taxing, claims this start-up that uses AI to support accountants and financial advisers.

Read more: Dublin fintech launches AI tax platform, aims to raise €500,000

OpenAI’s major chips deal with AMD could lead to 10pc stake

OpenAI and AMD have inked a major infrastructure agreement that could be worth billions to the chip maker, and could see it cede a 10pc stake to the the AI giant.

Read more: OpenAI’s major chips deal with AMD could lead to 10pc stake

Deloitte to pay money back to Albanese government after using AI in $440,000 report

Deloitte to pay money back to Albanese government after using AI in $440,000 report

Ouch:

Deloitte will provide a partial refund to the federal government over a $440,000 report that contained several errors, after admitting it used generative artificial intelligence to help produce it.

(I was initially confused by the "Albanese government" reference in the headline since this is a story about the Australian federal government. That's because the current Australia Prime Minister is Anthony Albanese.)

Here's the page for the report. The PDF now includes this note:

This Report was updated on 26 September 2025 and replaces the Report dated 4 July 2025. The Report has been updated to correct those citations and reference list entries which contained errors in the previously issued version, to amend the summary of the Amato proceeding which contained errors, and to make revisions to improve clarity and readability. The updates made in no way impact or affect the substantive content, findings and recommendations in the Report.

<p>Tags: <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/generative-ai">generative-ai</a>, <a href="https://simonwillison.net/tags/llms">llms</a>, <a href="https://simonwillison.net/tags/ai-ethics">ai-ethics</a>, <a href="https://simonwillison.net/tags/hallucinations">hallucinations</a></p>

a system that can do work independently on behalf of the user

I've settled on agents as meaning "LLMs calling tools in a loop to achieve a goal" but OpenAI continue to muddy the waters with much more vague definitions. Swyx spotted this one in the press pack OpenAI sent out for their DevDay announcements today:

How does OpenAl define an "agent"? An Al agent is a system that can do work independently on behalf of the user.

Adding this one to my collection.

<p>Tags: <a href="https://simonwillison.net/tags/ai-agents">ai-agents</a>, <a href="https://simonwillison.net/tags/openai">openai</a>, <a href="https://simonwillison.net/tags/agent-definitions">agent-definitions</a>, <a href="https://simonwillison.net/tags/swyx">swyx</a></p>

gpt-image-1-mini

gpt-image-1-mini

OpenAI released a new image model today: gpt-image-1-mini, which they describe as "A smaller image generation model that’s 80% less expensive than the large model."

They released it very quietly - I didn't hear about this in the DevDay keynote but I later spotted it on the DevDay 2025 announcements page.

It wasn't instantly obvious to me how to use this via their API. I ended up vibe coding a Python CLI tool for it so I could try it out.

I dumped the plain text diff version of the commit to the OpenAI Python library titled feat(api): dev day 2025 launches into ChatGPT GPT-5 Thinking and worked with it to figure out how to use the new image model and build a script for it. Here's the transcript and the the openai_image.py script it wrote.

I had it add inline script dependencies, so you can run it with uv like this:

export OPENAI_API_KEY="$(llm keys get openai)"
uv run https://tools.simonwillison.net/python/openai_image.py "A pelican riding a bicycle"

It picked this illustration style without me specifying it:

(This is a very different test from my normal "Generate an SVG of a pelican riding a bicycle" since it's using a dedicated image generator, not having a text-based model try to generate SVG code.)

My tool accepts a prompt, and optionally a filename (if you don't provide one it saves to a filename like /tmp/image-621b29.png).

It also accepts options for model and dimensions and output quality - the --help output lists those, you can see that here.

OpenAI's pricing is a little confusing. The model page claims low quality images should cost around half a cent and medium quality around a cent and a half. It also lists an image token price of $8/million tokens. It turns out there's a default "high" quality setting - most of the images I've generated have reported between 4,000 and 6,000 output tokens, which costs between 3.2 and 4.8 cents.

One last demo, this time using --quality low:

 uv run https://tools.simonwillison.net/python/openai_image.py \
  'racoon eating cheese wearing a top hat, realistic photo' \
  /tmp/racoon-hat-photo.jpg \
  --size 1024x1024 \
  --output-format jpeg \
  --quality low

This saved the following:

And reported this to standard error:

{
  "background": "opaque",
  "created": 1759790912,
  "generation_time_in_s": 20.87331541599997,
  "output_format": "jpeg",
  "quality": "low",
  "size": "1024x1024",
  "usage": {
    "input_tokens": 17,
    "input_tokens_details": {
      "image_tokens": 0,
      "text_tokens": 17
    },
    "output_tokens": 272,
    "total_tokens": 289
  }
}

This took 21s, but I'm on an unreliable conference WiFi connection so I don't trust that measurement very much.

272 output tokens = 0.2 cents so this is much closer to the expected pricing from the model page.

Tags: python, tools, ai, openai, generative-ai, uv, text-to-image, pelican-riding-a-bicycle, vibe-coding

GPT-5 pro

It has similar base characteristics to GPT-5: both share a September 30, 2024 knowledge cutoff and 400,000 context limit.

GPT-5 pro has maximum output tokens 272,000 max, an increase from 128,000 for GPT-5.

As our most advanced reasoning model, GPT-5 pro defaults to (and only supports) reasoning.effort: high

It's only available via OpenAI's Responses API. My LLM tool doesn't support that in core yet, but the llm-openai-plugin plugin does. I released llm-openai-plugin 0.7 adding support for the new model, then ran this:

llm install -U llm-openai-plugin
llm -m openai/gpt-5-pro "Generate an SVG of a pelican riding a bicycle"

It's very, very slow. The model took 6 minutes 8 seconds to respond and charged me for 16 input and 9,205 output tokens. At $15/million input and $120/million output this pelican cost me $1.10!

Here's the full transcript. It looks visually pretty simpler to the much, much cheaper result I got from GPT-5.

Tags: ai, openai, generative-ai, llms, llm-pricing, pelican-riding-a-bicycle, llm-reasoning, llm-release, gpt-5

I'm at OpenAI DevDay in Fort Mason, San Francisco today. As I did last year, I'm going to be live blogging the announcements from the kenote. Unlike last year, this year there's a livestream.

Tags: ai, openai, generative-ai, llms, live-blog

By Tomasz Magdanski

Introduction:
What if we treated data with the same care and intentionality as a consumer-facing product? Adopting a “data as a product” mindset means viewing data not as an incidental byproduct of systems, but as a core product in its own right. In practice, this means each data product is intentionally designed, built, maintained, and measured to create value. A data product has a clear purpose tied to a business decision, is created for a defined audience, and is continuously evaluated for utility, reliability, and accessibility. It is thoughtfully designed, guided by deliberate lifecycle management that includes both innovation, maintenance and retirement. Each product has explicit ownership, ensuring accuracy and availability, and it earns trust through consistency and quality.

At Netflix, data products — from high leverage telemetry, warehouse tables, key metrics and dashboards to ML models — power critical decisions in areas such as personalized recommendations, content strategy, product feature experimentation, customer acquisition, internal knowledge base, and even financial reporting. This deep reliance on data is how we internally create more value from data products — by using them to inform business decisions that enhance the member experience and drive the company forward. To maximize this value, we are advocating for a formal data product framework to ensure our data assets are managed with the same rigor and strategic focus as traditional products.

Why Treat Data as a Product?
Treating data as a product allows us to apply established product management principles to our data assets. Just as product teams define a clear purpose, target their users, measure success, and iterate on features, we can do the same for data. This shift in perspective yields several benefits: it clarifies the purpose and audience for each dataset, encourages thoughtful design and documentation for usability, enforces quality and reliability standards to build trust, and institutes a lifecycle with innovation and continuous improvement (and eventual retirement) rather than a “set and forget” approach. In short, a product mindset brings strategic alignment, usability, accountability, and better governance to our data investments. Treating data as a product means putting data users first: they are viewed as customers, and their needs guide the design and evolution of data solutions. With this context in mind, below are the key principles that define a data product at Netflix.

Key Principles of Data Products

• Clear Purpose: Every data product should exist for a reason — it should solve a problem or enable a specific business use case. If a data product doesn’t have a clear problem to solve or question to answer, it’s likely not worth creating. Owners of a data product must be able to articulate what value it delivers: for example, what decision does it inform or what insight does it provide? Defining a strong purpose prevents the creation of “data for data’s sake” and keeps our efforts focused on high-impact needs. In practice, this reflects the mantra of starting with business value when developing any new data asset.
• Defined Users (Audience): Just as consumer products have target customers, data products need clearly defined users. Identifying the who — the analysts, data scientists, engineers, tools, or business teams that will consume the data — helps shape the design. If no real users or use cases can be identified, the data product might be redundant. Thinking of data users as customers means understanding their needs and pain points upfront, so we can build data solutions that are actually useful and adopted. A data product designed without a specific audience in mind is likely to miss the mark or go unused.
• Measurable Value and Quality: A data product’s value is best demonstrated through its health, trustworthiness, and ability to support long-term innovation rather than prescriptive metrics. Netflix’s Data Health initiative frames this effort: a cross-functional push to reduce complexity and data debt, ensure adherence to standards, and make intentional tradeoffs about how data is treated. This focus is essential as generative AI models, automation, and new analytics use cases demand higher-quality, well-documented data. Measuring and improving data health across domains — rather than targeting hard-to-quantify impact — helps teams prioritize investments, simplify systems, and strengthen trust in our data ecosystem. Healthy data products enable faster discovery, better reliability, and a stronger foundation for AI-driven innovation.
  | Stay tuned: we’ll publish a dedicated blog post diving deeper into the Data Health program soon.
• Thoughtful Design and Documentation: A good data product is designed with the user in mind. This means data assets should be well-structured, intuitive to understand, and easy to access. Poorly designed or undocumented datasets can frustrate users just as a badly designed app would frustrate customers. To avoid this, we consider aspects like useful schema design, consistent definitions, and convenient access patterns when building data products. We also recommend documenting our data products — describing what the data represents, how to use it, and any known limitations — so that users can quickly trust and leverage the data without guesswork. In essence, we aim to make consuming a data product as straightforward as using a well-designed software product. Just as product design focuses on user experience, data product design should minimize friction and confusion for data consumers.
• Lifecycle Management: A data product’s evolution doesn’t stop at launch. Just as software products undergo updates and eventual deprecation, data products have a lifecycle that requires ongoing evaluation and maintenance. They should be designed with flexibility in mind, allowing extension or adaptation to new needs without costly rewrites. This balance of adaptability and discipline ensures we can add enhancements, adjust to emerging use cases, or retire assets when they become obsolete or lose value. Proactive lifecycle management prevents the accumulation of “data debt” — for example, stale tables or reports that linger without clear ownership or purpose. By planning for the full lifecycle, including eventual sunsetting, we keep our data ecosystem relevant, cost-effective, and resilient. Data products should not live forever by default: if a metric or report is outdated, we treat it like outdated software and phase it out to make room for better solutions.
• Clear Ownership: Every data product needs a clearly designated owner responsible for its success. Ownership entails ensuring the data product’s accuracy, availability, and usability on an ongoing basis. Just as a product manager or product team is accountable for a software service, a data product owner is accountable for the quality and reliability of a dataset or analytic product. Having explicit ownership prevents the common problem of “everyone and no one” being responsible for data issues. It provides a point of contact for questions or problems and drives accountability for maintenance. At Netflix, this often means assigning data products to the teams most closely aligned with their content — for example, a content data product is owned by a content data science and engineering teams. Clear ownership also facilitates faster issue resolution and continuous improvement, because someone is empowered to act when the data product needs attention.
• Trust and Reliability: Users will only treat data as a product if they can trust it. Therefore, reliability isn’t just a technical concern — it’s a core product feature. Data products must be dependable (e.g. correct, up-to-date, consistent) to earn and keep the trust of their users. If a data product is frequently wrong or unavailable, users quickly lose confidence and abandon it (just as they would uninstall a buggy app). Building trust means instituting strong data quality controls, validation tests, monitoring, and alerts as part of the product’s design. It also means being transparent about the data’s limitations and lineage. An excellent data product provides an exceptional user experience and trustworthiness by design. In fact, these two traits — great user experience and trust — are often cited as the defining characteristics of a true data product. By prioritizing reliability, we treat data quality issues as product issues, addressing root causes and preventing bad data from eroding stakeholder confidence.

Conclusion:
In summary, treating data as a product reframes how we build and manage data at Netflix. We produce data products to create value — meaning every important dataset, metric, or report is developed with purpose and maintained with rigor throughout its lifecycle. By recognizing data’s users, use cases, and lifecycle as fundamental attributes, we elevate data to a first-class entity in our organization’s thinking. This product-oriented approach emphasizes reliability, clear ownership, proper governance, and proactive management of data debt. Ultimately, the data-as-a-product mindset ensures that our data products are valuable, trusted, and aligned with Netflix’s strategic goals — just like any successful consumer product would be. We’re excited to adopt this framework across our teams, continuing to improve how we deliver insight and impact through data.

We’re thrilled to announce the first wave of keynote speakers for ODSC West 2025, taking place October 28–30 in San Francisco. This year’s lineup brings together pioneering founders, influential leaders, and inspiring educators who are shaping the future of AI and data science.

From building foundational AI platforms to scaling enterprise adoption, these keynote speakers will share insights, lessons, and predictions that you won’t find anywhere else. These keynotes are the people who build what you use, making their insights hyper relevant to your everyday work.

Here’s your first look at who will be taking the keynote stage this October:

Harrison Chase — CEO and Co-founder, LangChain