012026 Papers156 lessons156 lessons02Machine Learning for Health (ML4H) 2024Machine Learning for Health (ML4H) 2024 — published at ML4H@NeurIPS 2024.03Machine Learning for Health, ML4H@NeurIPS 2024, Vanc...Machine Learning for Health, ML4H@NeurIPS 2024, Vancouver, Canada, 15-16 December 2024 — published at ML4H@NeurIPS 2024.04The Second Tiny Papers Track at ICLR 2024, Tiny Pape...The Second Tiny Papers Track at ICLR 2024, Tiny Papers @ ICLR 2024, Vienna, Austria, May 11, 2024 - published at Tiny Papers @ ICLR 2024.05MpoxVLM: A Vision-Language Model for Diagnosing Skin...In the aftermath of the COVID-19 pandemic and amid accelerating climate change, emerging infectious diseases, particularly those arising from zoonotic s...06A Learning-based Multi-Frame Visual Feature Framewor...A Learning-based Multi-Frame Visual Feature Framewor... - published at NAACL 2025.07A Practical Analysis of Human Alignment with *PO.A Practical Analysis of Human Alignment with *PO. - published at NAACL 2025.08A Semantic-Aware Layer-Freezing Approach to Computat...A Semantic-Aware Layer-Freezing Approach to Computat... - published at ACL 2025.09A Training-free LLM-based Approach to General Chines...A Training-free LLM-based Approach to General Chines... - published at ACL 2025.10AbGen: Evaluating Large Language Models in Ablation...AbGen: Evaluating Large Language Models in Ablation... - published at ACL 2025.11Active Few-Shot Learning for Text Classification.How to intelligently select which examples to annotate when you only have a handful of labeled samples per class. Combines active learning with few-shot text classification to minimize annotation cost - directly applicable to intent detection, content moderation, and domain-specific NLP tasks.12Advancing Language Models through Instruction Tuning...Advancing Language Models through Instruction Tuning... - published at EMNLP 2025.13AERA Chat: An Interactive Platform for Automated Exp...AERA Chat: An Interactive Platform for Automated Exp... - published at EMNLP 2025.14AgentCPM-GUI: Building Mobile-Use Agents with Reinfo...AgentCPM-GUI: Building Mobile-Use Agents with Reinfo... - published at EMNLP 2025.15AIPOM: Agent-aware Interactive Planning for Multi-Ag...AIPOM: Agent-aware Interactive Planning for Multi-Ag... - published at EMNLP 2025.16An Address Intelligence Framework for E-commerce Del...An Address Intelligence Framework for E-commerce Del... - published at EMNLP 2025.17ASRank: Zero-Shot Re-Ranking with Answer Scent for D...ASRank: Zero-Shot Re-Ranking with Answer Scent for D... - published at NAACL 2025.18Automatically Discovering How Misogyny is Framed on...Automatically Discovering How Misogyny is Framed on... - published at NAACL 2025.19AUTOSUMM: A Comprehensive Framework for LLM-Based Co...AUTOSUMM: A Comprehensive Framework for LLM-Based Co... - published at ACL 2025.20Benchmarking and Building Zero-Shot Hindi Retrieval...Benchmarking and Building Zero-Shot Hindi Retrieval... - published at NAACL 2025.21BOOKCOREF: Coreference Resolution at Book Scale.BOOKCOREF: Coreference Resolution at Book Scale. - published at ACL 2025.22CFSP: An Efficient Structured Pruning Framework for...CFSP: An Efficient Structured Pruning Framework for... - published at COLING 2025.23CodeGenWrangler: Data Wrangling task automation usin...CodeGenWrangler: Data Wrangling task automation usin... - published at NAACL 2025.24CodeTaxo: Enhancing Taxonomy Expansion with Limited...CodeTaxo: Enhancing Taxonomy Expansion with Limited... - published at ACL 2025.25Cognitive Kernel: An Open-source Agent System toward...Cognitive Kernel: An Open-source Agent System toward... - published at NAACL 2025.26Compress to Impress: Unleashing the Potential of Com...Compress to Impress: Unleashing the Potential of Com... - published at COLING 2025.27Controllable Style Arithmetic with Language Models.Controllable Style Arithmetic with Language Models. - published at ACL 2025.28Craw4LLM: Efficient Web Crawling for LLM Pretraining.Craw4LLM: Efficient Web Crawling for LLM Pretraining. - published at ACL 2025.29CUFE@NLU of Devanagari Script Languages 2025: Langua...CUFE@NLU of Devanagari Script Languages 2025: Langua... - published at COLING 2025.30CUFE@VarDial 2025 NorSID: Multilingual BERT for Norw...CUFE@VarDial 2025 NorSID: Multilingual BERT for Norw... - published at COLING 2025.31DASR: Distributed Adaptive Scene Recognition - A Mul...DASR: Distributed Adaptive Scene Recognition - A Mul... - published at EMNLP 2025.32DEMO: Reframing Dialogue Interaction with Fine-grain...DEMO: Reframing Dialogue Interaction with Fine-grain... - published at ACL 2025.33Does Generative AI speak Nigerian-Pidgin?: Issues ab...Does Generative AI speak Nigerian-Pidgin?: Issues ab... - published at NAACL 2025.34Does RAG Introduce Unfairness in LLMs? Evaluating Fa...Does RAG Introduce Unfairness in LLMs? Evaluating Fa... - published at COLING 2025.35Driving Chinese Spelling Correction from a Fine-Grai...Driving Chinese Spelling Correction from a Fine-Grai... - published at COLING 2025.36Dual Debiasing for Noisy In-Context Learning for Tex...Dual Debiasing for Noisy In-Context Learning for Tex... - published at ACL 2025.37EasyDistill: A Comprehensive Toolkit for Effective K...EasyDistill: A Comprehensive Toolkit for Effective K... - published at EMNLP 2025.38Efficiency-Effectiveness Reranking FLOPs for LLM-bas...Efficiency-Effectiveness Reranking FLOPs for LLM-bas... - published at EMNLP 2025.39Empathy Prediction from Diverse Perspectives.Empathy Prediction from Diverse Perspectives. - published at ACL 2025.40Enhancing Open-Domain Task-Solving Capability of LLM...Enhancing Open-Domain Task-Solving Capability of LLM... - published at ACL 2025.41FACT-AUDIT: An Adaptive Multi-Agent Framework for Dy...FACT-AUDIT: An Adaptive Multi-Agent Framework for Dy... - published at ACL 2025.42From Feedback to Checklists: Grounded Evaluation of...From Feedback to Checklists: Grounded Evaluation of... - published at EMNLP 2025.43From Long Videos to Engaging Clips: A Human-Inspired...From Long Videos to Engaging Clips: A Human-Inspired... - published at EMNLP 2025.44GADFA: Generator-Assisted Decision-Focused Approach...GADFA: Generator-Assisted Decision-Focused Approach... - published at COLING 2025.45Goal-Driven Data Story, Narrations and Explanations.Goal-Driven Data Story, Narrations and Explanations. - published at NAACL 2025.46GRAM: Generative Recommendation via Semantic-aware M...GRAM: Generative Recommendation via Semantic-aware M... - published at ACL 2025.47How Credible Is an Answer From Retrieval-Augmented L...How Credible Is an Answer From Retrieval-Augmented L... - published at COLING 2025.48H-STAR: LLM-driven Hybrid SQL-Text Adaptive Reasonin...H-STAR: LLM-driven Hybrid SQL-Text Adaptive Reasonin... - published at NAACL 2025.49Hybrid Graphs for Table-and-Text based Question Answ...Hybrid Graphs for Table-and-Text based Question Answ... - published at NAACL 2025.50InTriage: Intelligent Telephone Triage in Pre-Hospit...InTriage: Intelligent Telephone Triage in Pre-Hospit... - published at EMNLP 2025.51IrokoBench: A New Benchmark for African Languages in...IrokoBench: A New Benchmark for African Languages in... - published at NAACL 2025.52Leveraging Language-based Representations for Better...Leveraging Language-based Representations for Better... - published at COLING 2025.53LLM-Coordination: Evaluating and Analyzing Multi-age...LLM-Coordination: Evaluating and Analyzing Multi-age... - published at NAACL 2025.54LLMInit: A Free Lunch from Large Language Models for...LLMInit: A Free Lunch from Large Language Models for... - published at EMNLP 2025.55Loki: An Open-Source Tool for Fact Verification.Loki: An Open-Source Tool for Fact Verification. - published at COLING 2025.56MCPEval: Automatic MCP-based Deep Evaluation for AI...MCPEval: Automatic MCP-based Deep Evaluation for AI... - published at EMNLP 2025.57Meta-Reasoning Improves Tool Use in Large Language M...Meta-Reasoning Improves Tool Use in Large Language M... - published at NAACL 2025.58Mirror in the Model: Ad Banner Image Generation via...Mirror in the Model: Ad Banner Image Generation via... - published at EMNLP 2025.59Multi-Task Pre-Finetuning of Lightweight Transformer...Multi-Task Pre-Finetuning of Lightweight Transformer... - published at EMNLP 2025.60Narrative Media Framing in Political Discourse.Narrative Media Framing in Political Discourse. - published at ACL 2025.61NormAL LoRA: What is the perfect size?NormAL LoRA: What is the perfect size? - published at EMNLP 2025.62Odysseus Navigates the Sirens' Song: Dynamic Fo...Odysseus Navigates the Sirens' Song: Dynamic Fo... - published at ACL 2025.63Open Political Corpora: Structuring, Searching, and...Open Political Corpora: Structuring, Searching, and... - published at EMNLP 2025.64pEBR: A Probabilistic Approach to Embedding Based Re...pEBR: A Probabilistic Approach to Embedding Based Re... - published at EMNLP 2025.65Persona-SQ: A Personalized Suggested Question Genera...Persona-SQ: A Personalized Suggested Question Genera... - published at NAACL 2025.66PledgeTracker: A System for Monitoring the Fulfilmen...PledgeTracker: A System for Monitoring the Fulfilmen... - published at EMNLP 2025.67Position-Aware Depth Decay Decoding (D³): Boosting L...Position-Aware Depth Decay Decoding (D³): Boosting L... - published at ACL 2025.68Problem-Solving Logic Guided Curriculum In-Context L...Problem-Solving Logic Guided Curriculum In-Context L... - published at ACL 2025.69Proceedings of Bridging Neurons and Symbols for Natu...Proceedings of Bridging Neurons and Symbols for Natu... - published at COLING 2025.70Proceedings of Context and Meaning: Navigating Disag...Proceedings of Context and Meaning: Navigating Disag... - published at COLING 2025.71Proceedings of the 5th Celtic Language Technology Wo...Proceedings of the 5th Celtic Language Technology Wo... - published at COLING 2025.72Proceedings of the Joint Workshop of the 9th Financi...Proceedings of the Joint Workshop of the 9th Financi... - published at COLING 2025.73Reasoning Knowledge Filter for Logical Table-to-Text...Reasoning Knowledge Filter for Logical Table-to-Text... - published at COLING 2025.74Reasoning-Enhanced Domain-Adaptive Pretraining of Mu...Reasoning-Enhanced Domain-Adaptive Pretraining of Mu... - published at EMNLP 2025.75Red Queen: Exposing Latent Multi-Turn Risks in Large...Red Queen: Exposing Latent Multi-Turn Risks in Large... - published at ACL 2025.76RedOne: Revealing Domain-specific LLM Post-Training...RedOne: Revealing Domain-specific LLM Post-Training... - published at EMNLP 2025.77Registering Source Tokens to Target Language Spaces...Registering Source Tokens to Target Language Spaces... - published at ACL 2025.78Reinforcement Learning for Aligning Large Language M...Reinforcement Learning for Aligning Large Language M... - published at NAACL 2025.79Representing the Under-Represented: Cultural and Cor...Representing the Under-Represented: Cultural and Cor... - published at COLING 2025.80RevieWeaver: Weaving Together Review Insights by Lev...RevieWeaver: Weaving Together Review Insights by Lev... - published at NAACL 2025.81RichRAG: Crafting Rich Responses for Multi-faceted Q...RichRAG: Crafting Rich Responses for Multi-faceted Q... - published at COLING 2025.82RTSM: Knowledge Distillation with Diverse Signals fo...RTSM: Knowledge Distillation with Diverse Signals fo... - published at NAACL 2025.83Safe: Enhancing Mathematical Reasoning in Large Lang...Safe: Enhancing Mathematical Reasoning in Large Lang... - published at ACL 2025.84SciClaims: An End-to-End Generative System for Biome...SciClaims: An End-to-End Generative System for Biome... - published at EMNLP 2025.85Script-Agnosticism and its Impact on Language Identi...Script-Agnosticism and its Impact on Language Identi... - published at NAACL 2025.86sDPO: Don't Use Your Data All at Once.sDPO: Don't Use Your Data All at Once. - published at COLING 2025.87SeaLLMs 3: Open Foundation and Chat Multilingual Lar...SeaLLMs 3: Open Foundation and Chat Multilingual Lar... - published at NAACL 2025.88Semi-automatic Sequential Sentence Classification in...Semi-automatic Sequential Sentence Classification in... - published at NAACL 2025.89Sens-Merging: Sensitivity-Guided Parameter Balancing...Sens-Merging: Sensitivity-Guided Parameter Balancing... - published at ACL 2025.90SlackAgents: Scalable Collaboration of AI Agents in...SlackAgents: Scalable Collaboration of AI Agents in... - published at EMNLP 2025.91SoftCoT: Soft Chain-of-Thought for Efficient Reasoni...SoftCoT: Soft Chain-of-Thought for Efficient Reasoni... - published at ACL 2025.92Structured Tender Entities Extraction from Complex T...Structured Tender Entities Extraction from Complex T... - published at COLING 2025.93TableCoder: Table Extraction from Text via Reliable...TableCoder: Table Extraction from Text via Reliable... - published at ACL 2025.94TelAgentBench: A Multi-faceted Benchmark for Evaluat...TelAgentBench: A Multi-faceted Benchmark for Evaluat... - published at EMNLP 2025.95Text-Attributed Graph Learning with Coupled Augmenta...Text-Attributed Graph Learning with Coupled Augmenta... - published at COLING 2025.96The Invalsi Benchmarks: measuring the Linguistic and...The Invalsi Benchmarks: measuring the Linguistic and... - published at COLING 2025.97The LLM Language Network: A Neuroscientific Approach...The LLM Language Network: A Neuroscientific Approach... - published at NAACL 2025.98The Role of Handling Attributive Nouns in Improving...The Role of Handling Attributive Nouns in Improving... - published at COLING 2025.99Thinking with DistilQwen: A Tale of Four Distilled R...Thinking with DistilQwen: A Tale of Four Distilled R... - published at EMNLP 2025.100TRUSTEVAL: A Dynamic Evaluation Toolkit on Trustwort...TRUSTEVAL: A Dynamic Evaluation Toolkit on Trustwort... - published at NAACL 2025.101VCRMNER: Visual Cue Refinement in Multimodal NER usi...VCRMNER: Visual Cue Refinement in Multimodal NER usi... - published at COLING 2025.102Vision-Language Models Struggle to Align Entities ac...Vision-Language Models Struggle to Align Entities ac... - published at ACL 2025.103VoxpopuliTTS: a large-scale multilingual TTS corpus...VoxpopuliTTS: a large-scale multilingual TTS corpus... - published at COLING 2025.104Watching the AI Watchdogs: A Fairness and Robustness...Watching the AI Watchdogs: A Fairness and Robustness... - published at NAACL 2025.105What Makes for Good Visual Instructions? Synthesizin...What Makes for Good Visual Instructions? Synthesizin... - published at COLING 2025.106Analysing LLM Persona Generation and Fairness Interp...Analysing LLM Persona Generation and Fairness Interp... - published at EACL 2026.107Argumentation and Judgement Factors: LLM-based Disco...Argumentation and Judgement Factors: LLM-based Disco... - published at EACL 2026.108Beyond Grid Search: Leveraging Bayesian Optimization...Beyond Grid Search: Leveraging Bayesian Optimization... - published at EACL 2026.109Beyond "Not Novel Enough": Enriching Schol...Beyond "Not Novel Enough": Enriching Schol... - published at EACL 2026.110Bridging Attribution and Open-Set Detection using Gr...Bridging Attribution and Open-Set Detection using Gr... - published at EACL 2026.111Cards Against Contamination: TCG-Bench for Difficult...Cards Against Contamination: TCG-Bench for Difficult... - published at EACL 2026.112Different Time, Different Language: Revisiting the B...Different Time, Different Language: Revisiting the B... - published at EACL 2026.113DIVINE : Coordinating Multimodal Disentangled Repres...DIVINE : Coordinating Multimodal Disentangled Repres... - published at EACL 2026.114Enhancing Reliability in Community Question Answerin...Enhancing Reliability in Community Question Answerin... - published at EACL 2026.115Evaluation of Deontic Conditional Reasoning in Large...Evaluation of Deontic Conditional Reasoning in Large... - published at EACL 2026.116From Paper to Structured JSON: An Agentic AI Workflo...From Paper to Structured JSON: An Agentic AI Workflo... - published at EACL 2026.117Generating Multi-Aspect Queries for Conversational S...Generating Multi-Aspect Queries for Conversational S... - published at EACL 2026.118I know you are different! Towards Persona Driven Kno...I know you are different! Towards Persona Driven Kno... - published at EACL 2026.119LEMUR: Robust Fine-Tuning for Multilingual Embedding...LEMUR: Robust Fine-Tuning for Multilingual Embedding... - published at EACL 2026.120Leveraging LLM-GNN Integration for Open-World Questi...Leveraging LLM-GNN Integration for Open-World Questi... - published at EACL 2026.121McMining: Automated Discovery of Misconceptions in S...McMining: Automated Discovery of Misconceptions in S... - published at EACL 2026.122Mitigating Copy Bias in In-Context Learning through...Mitigating Copy Bias in In-Context Learning through... - published at EACL 2026.123MULSUM: A Multimodal Summarization System with Vis-A...MULSUM: A Multimodal Summarization System with Vis-A... - published at EACL 2026.124Multilingual Self-Taught Faithfulness Evaluators.Multilingual Self-Taught Faithfulness Evaluators. - published at EACL 2026.125Nemotron-CrossThink: Scaling Self-Learning beyond Ma...Nemotron-CrossThink: Scaling Self-Learning beyond Ma... - published at EACL 2026.126PromptLab: A Collaborative Platform for Prompt Engin...PromptLab: A Collaborative Platform for Prompt Engin... - published at EACL 2026.127Rad-Flamingo: A Multimodal Prompt driven Radiology R...Rad-Flamingo: A Multimodal Prompt driven Radiology R... - published at EACL 2026.128RECIPE-TKG: From Sparse History to Structured Reason...RECIPE-TKG: From Sparse History to Structured Reason... - published at EACL 2026.129TIPA: Typologically Informed Parameter Aggregation.TIPA: Typologically Informed Parameter Aggregation. - published at EACL 2026.130Where Do LLMs Compose Meaning? A Layerwise Analysis...Where Do LLMs Compose Meaning? A Layerwise Analysis... - published at EACL 2026.131VenusBench-Mobile: A Challenging and User-Centric Be...Existing online benchmarks for mobile GUI agents remain largely app-centric and task-homogeneous, failing to reflect the diversity and instability of re...132SPEED-Bench: A Unified and Diverse Benchmark for Spe...Speculative Decoding (SD) has emerged as a critical technique for accelerating Large Language Model (LLM) inference. Unlike deterministic system optimiz...133Turing Test on Screen: A Benchmark for Mobile GUI Ag...The rise of autonomous GUI agents has triggered adversarial countermeasures from digital platforms, yet existing research prioritizes utility and robust...134A Constrained RL Approach for Cost-Efficient Deliver...Next-generation networks aim to provide performance guarantees to real-time interactive services that require timely and cost-efficient packet delivery....135A Dual-Helix Governance Approach Towards Reliable Ag...WebGIS development requires rigor, yet agentic AI frequently fails due to five large language model (LLM) limitations: context constraints, cross-sessio...136A Stein Identity for q-Gaussians with Bounded SupportStein's identity is a fundamental tool in machine learning with applications in generative models, stochastic optimization, and other problems involving...137Accurate and Efficient Hybrid-Ensemble Atmospheric D...Data assimilation (DA) combines model forecasts and observations to estimate the optimal state of the atmosphere with its uncertainty, providing initial...138AgentIR: Reasoning-Aware Retrival for Deep Research...Deep Research agents are rapidly emerging as primary consumers of modern retrieval systems. Unlike human users who issue and refine queries without docu...139AILS-NTUA at SemEval-2026 Task 12: Graph-Based Retri...We present a winning three-stage system for SemEval 2026 Task~12: Abductive Event Reasoning that combines graph-based retrieval, LLM-driven abductive re...140ArtHOI: Articulated Human-Object Interaction Synthes...Synthesizing physically plausible articulated human-object interactions (HOI) without 3D/4D supervision remains a fundamental challenge. While recent ze...141Balancing Fidelity, Utility, and Privacy in Syntheti...Deep learning in cardiac MRI (CMR) is fundamentally constrained by both data scarcity and privacy regulations. This study systematically benchmarks thre...142Beyond Mixtures and Products for Ensemble Aggregatio...Density aggregation is a central problem in machine learning, for instance when combining predictions from a Deep Ensemble. The choice of aggregation re...143Causality Elicitation from Large Language ModelsLarge language models (LLMs) are trained on enormous amounts of data and encode knowledge in their parameters. We propose a pipeline to elicit causal re...144Code Fingerprints: Disentangled Attribution of LLM-G...The rapid adoption of Large Language Models (LLMs) has transformed modern software development by enabling automated code generation at scale. While the...145Direct Bayesian Additive Regression Trees for Condit...Regression discontinuity designs (RDD) are widely used for causal inference. In many empirical applications, treatment effects vary substantially with c...146Dissecting Quantization Error: A Concentration-Align...Quantization can drastically increase the efficiency of large language and vision models, but typically incurs an accuracy drop. Recently, function-pres...147Dual-Modality Multi-Stage Adversarial Safety Trainin...Multimodal web agents that process both screenshots and accessibility trees are increasingly deployed to interact with web interfaces, yet their dual-st...148Efficient Refusal Ablation in LLM through Optimal Tr...Safety-aligned language models refuse harmful requests through learned refusal behaviors encoded in their internal representations. Recent activation-ba...149Enhancing Authorship Attribution with Synthetic Pain...Attributing authorship to paintings is a historically complex task, and one of its main challenges is the limited availability of real artworks for trai...150Exploiting Subgradient Sparsity in Max-Plus Neural N...Deep Neural Networks are powerful tools for solving machine learning problems, but their training often involves dense and costly parameter updates. In...151Fixed-Budget Constrained Best Arm Identification in...We study fixed budget constrained best-arm identification in grouped bandits, where each arm consists of multiple independent attributes with stochastic...152FocusGraph: Graph-Structured Frame Selection for Emb...The ability to understand long videos is vital for embodied intelligent agents, because their effectiveness depends on how well they can accumulate, org...153Generalization Properties of Score-matching Diffusio...Despite the remarkable empirical success of score-based diffusion models, their statistical guarantees remain underdeveloped. Existing analyses often pr...154Helios: Real Real-Time Long Video Generation ModelWe introduce Helios, the first 14B video generation model that runs at 19.5 FPS on a single NVIDIA H100 GPU and supports minute-scale generation while m...155Hierarchical Inference and Closure Learning via Adap...Inverse problems are the task of calibrating models to match data. They play a pivotal role in diverse engineering applications by allowing practitioner...156Hold-One-Shot-Out (HOSO) for Validation-Free Few-Sho...In many CLIP adaptation methods, a blending ratio hyperparameter controls the trade-off between general pretrained CLIP knowledge and the limited, datas...157Invariance-Based Dynamic Regret MinimizationWe consider stochastic non-stationary linear bandits where the linear parameter connecting contexts to the reward changes over time. Existing algorithms...158Inverse Contextual Bandits without Rewards: Learning...We study the Inverse Contextual Bandit (ICB) problem, in which a learner seeks to optimize a policy while an observer, who cannot access the learner's r...159$τ$-Knowledge: Evaluating Conversational Agents over...Conversational agents are increasingly deployed in knowledge-intensive settings, where correct behavior depends on retrieving and applying domain-specif...160Low-Resource Guidance for Controllable Latent Audio...Generative audio requires fine-grained controllable outputs, yet most existing methods require model retraining on specific controls or inference-time c...161Memex(RL): Scaling Long-Horizon LLM Agents via Index...Large language model (LLM) agents are fundamentally bottlenecked by finite context windows on long-horizon tasks. As trajectories grow, retaining tool o...162MOO: A Multi-view Oriented Observations Dataset for...Animal re-identification (ReID) faces critical challenges due to viewpoint variations, particularly in Aerial-Ground (AG-ReID) settings where models mus...163Observationally Informed Adaptive Causal Experimenta...Randomized Controlled Trials (RCTs) represent the gold standard for causal inference yet remain a scarce resource. While large-scale observational data...164Out-of-distribution transfer of PDE foundation model...Most PDE foundation models are pretrained and fine-tuned on fluid-centric benchmarks. Their utility under extreme-loading material dynamics remains uncl...165Pointer-CAD: Unifying B-Rep and Command Sequences vi...Constructing computer-aided design (CAD) models is labor-intensive but essential for engineering and manufacturing. Recent advances in Large Language Mo...166Position: Vector Prompt Interfaces Should Be Exposed...As large language models (LLMs) transition from research prototypes to real-world systems, customization has emerged as a central bottleneck. While text...167PTOPOFL: Privacy-Preserving Personalised Federated L...Federated learning (FL) faces two structural tensions: gradient sharing enables data-reconstruction attacks, while non-IID client distributions degrade...168RANGER: Sparsely-Gated Mixture-of-Experts with Adapt...Pathology report generation remains a relatively under-explored downstream task, primarily due to the gigapixel scale and complex morphological heteroge...169Retrieval or Representation? Reassessing Benchmark G...Retrieval-augmented generation (RAG) is a common way to ground language models in external documents and up-to-date information. Classical retrieval sys...170RoboCasa365: A Large-Scale Simulation Framework for...Recent advances in robot learning have accelerated progress toward generalist robots that can perform everyday tasks in human environments. Yet it remai...171Robust Unscented Kalman Filtering via Recurrent Meta...The Unscented Kalman Filter (UKF) is a ubiquitous tool for nonlinear state estimation; however, its performance is limited by the static parameterizatio...172Robustness of Agentic AI Systems via Adversarially-A...As Large Language Models (LLMs) transition into autonomous multi-agent ecosystems, robust minimax training becomes essential yet remains prone to instab...173Scalable Evaluation of the Realism of Synthetic Envi...Evaluation of AI systems often requires synthetic test cases, particularly for rare or safety-critical conditions that are difficult to observe in opera...174SELDON: Supernova Explosions Learned by Deep ODE Net...The discovery rate of optical transients will explode to 10 million public alerts per night once the Vera C. Rubin Observatory's Legacy Survey of Space...175Self-Sovereign AgentWe investigate the emerging prospect of self-sovereign agents -- AI systems that can economically sustain and extend their own operation without human i...176Semi-Supervised Generative Learning via Latent Space...We introduce Latent Space Distribution Matching (LSDM), a novel framework for semi-supervised generative modeling of conditional distributions. LSDM ope...177SimpliHuMoN: Simplifying Human Motion PredictionHuman motion prediction combines the tasks of trajectory forecasting and human pose prediction. For each of the two tasks, specialized models have been...178SpotIt+: Verification-based Text-to-SQL Evaluation w...We present SpotIt+, an open-source tool for evaluating Text-to-SQL systems via bounded equivalence verification. Given a generated SQL query and the gro...179SPRINT: Semi-supervised Prototypical Representation...Real-world systems must continuously adapt to novel concepts from limited data without forgetting previously acquired knowledge. While Few-Shot Class-In...180Stable and Steerable Sparse Autoencoders with Weight...Sparse autoencoders (SAEs) are widely used to extract human-interpretable features from neural network activations, but their learned features can vary...181Statistical Inference for Score DecompositionsWe introduce inference methods for score decompositions, which partition scoring functions for predictive assessment into three interpretable components...182TaxonRL: Reinforcement Learning with Intermediate Re...Traditional vision-language models struggle with contrastive fine-grained taxonomic reasoning, particularly when distinguishing between visually similar...183Testing Full Mediation of Treatment Effects and the...In causal analysis, understanding the causal mechanisms through which an intervention or treatment affects an outcome is often of central interest. We p...184The Company You Keep: How LLMs Respond to Dark Triad...Large Language Models (LLMs) often exhibit highly agreeable and reinforcing conversational styles, also known as AI-sycophancy. Although this behavior i...185Turning Trust to Transactions: Tracking Affiliate Ma...YouTube has evolved into a powerful platform that where creators monetize their influence through affiliate marketing, raising concerns about transparen...186Underrepresented in Foundation Model Pretraining Dat...Large-scale Vision-Language Foundation Models (VLFMs), such as CLIP, now underpin a wide range of computer vision research and applications. VLFMs are o...187$V_1$: Unifying Generation and Self-Verification for...Test-time scaling for complex reasoning tasks shows that leveraging inference-time compute, by methods such as independently sampling and aggregating mu...188What Does Flow Matching Bring To TD Learning?Recent work shows that flow matching can be effective for scalar Q-value function estimation in reinforcement learning (RL), but it remains unclear why...189When Do Language Models Endorse Limitations on Human...As Large Language Models (LLMs) increasingly mediate global information access with the potential to shape public discourse, their alignment with univer...190World Properties without World Models: Recovering Sp...Recent work interprets the linear recoverability of geographic and temporal variables from large language model (LLM) hidden states as evidence for worl...191ZipMap: Linear-Time Stateful 3D Reconstruction with...Feed-forward transformer models have driven rapid progress in 3D vision, but state-of-the-art methods such as VGGT and $π^3$ have a computational cost t...192Behavior-dLDS: A decomposed linear dynamical systems...Brain-wide recordings of large-scale networks of neurons now provide an unprecedented view into how the brain drives behavior. However, brain activity c...193Improved Scaling Laws via Weak-to-Strong Generalizat...It is increasingly common in machine learning to use learned models to label data and then employ such data to train more capable models. The phenomenon...194Spatial Competence BenchmarkSpatial competence is the quality of maintaining a consistent internal representation of an environment and using it to infer discrete structure and pla...195SurvHTE-Bench: A Benchmark for Heterogeneous Treatme...Estimating heterogeneous treatment effects (HTEs) from right-censored survival data is critical in high-stakes applications such as precision medicine a...196Thermodynamic Response Functions in Singular Bayesia...Singular statistical models-including mixtures, matrix factorization, and neural networks-violate regular asymptotics due to parameter non-identifiabili...197A recipe for scalable attention-based MLIPs: unlocki...Machine-learning interatomic potentials (MLIPs) have advanced rapidly, with many top models relying on strong physics-based inductive biases. However, a...198A Reference Architecture of Reinforcement Learning F...The surge in reinforcement learning (RL) applications gave rise to diverse supporting technology, such as RL frameworks. However, the architectural patt...199Abductive Reasoning with Syllogistic Forms in Large...Research in AI using Large-Language Models (LLMs) is rapidly evolving, and the comparison of their performance with human reasoning has become a key con...200Agnostic learning in (almost) optimal time via Gauss...The complexity of learning a concept class under Gaussian marginals in the difficult agnostic model is closely related to its $L_1$-approximability by l...201Artificial Intelligence for Detecting Fetal Orofacia...Orofacial clefts are among the most common congenital craniofacial abnormalities, yet accurate prenatal detection remains challenging due to the scarcit...202AV-Unified: A Unified Framework for Audio-visual Sce...When humans perceive the world, they naturally integrate multiple audio-visual tasks within dynamic, real-world scenes. However, current works such as e...203Bayesian Additive Distribution RegressionDistribution regression, where the goal is to predict a scalar response from a distribution-valued predictor, arises naturally in settings where observa...204BEVLM: Distilling Semantic Knowledge from LLMs into...The integration of Large Language Models (LLMs) into autonomous driving has attracted growing interest for their strong reasoning and semantic understan...205Beyond Rows to Reasoning: Agentic Retrieval for Mult...Recent advances in multimodal Retrieval-Augmented Generation (RAG) enable Large Language Models (LLMs) to analyze enterprise spreadsheet workbooks conta...206Boosting deep Reinforcement Learning using pretraini...Deep reinforcement learning agents are often misaligned, as they over-exploit early reward signals. Recently, several symbolic approaches have addressed...207Causal Interpretation of Neural Network Computations...Understanding how neural networks transform inputs into outputs is crucial for interpreting and manipulating their behavior. Most existing approaches an...208Certified and accurate computation of function space...Neural network methods for PDEs require reliable error control in function space norms. However, trained neural networks can typically only be probed at...209CLoPA: Continual Low Parameter Adaptation of Interac...Interactive segmentation enables clinicians to guide annotation, but existing zero-shot models like nnInteractive fail to consistently reach expert-leve...210COLD-Steer: Steering Large Language Models via In-Co...Activation steering methods enable inference-time control of large language model (LLM) behavior without retraining, but current approaches face a funda...211Continual Adaptation for Pacific Indigenous Speech R...Speech foundation models struggle with low-resource Pacific Indigenous languages because of severe data scarcity. Furthermore, full fine-tuning risks ca...212Design Experiments to Compare Multi-armed Bandit Alg...Online platforms routinely compare multi-armed bandit algorithms, such as UCB and Thompson Sampling, to select the best-performing policy. Unlike standa...213Do Foundation Models Know Geometry? Probing Frozen F...Vision-language models encode continuous geometry that their text pathway fails to express: a 6,000-parameter linear probe extracts hand joint angles at...214EgoReasoner: Learning Egocentric 4D Reasoning via Ta...Egocentric video understanding is inherently complex due to the dynamic 4D nature of the environment, where camera motion and object displacements neces...215Evaluation of Deontic Conditional Reasoning in Large...As large language models (LLMs) advance in linguistic competence, their reasoning abilities are gaining increasing attention. In humans, reasoning often...216Fly360: Omnidirectional Obstacle Avoidance within Dr...Obstacle avoidance in unmanned aerial vehicles (UAVs), as a fundamental capability, has gained increasing attention with the growing focus on spatial in...217From Prompting to Preference Optimization: A Compara...Large language models (LLMs) have recently reshaped Automated Essay Scoring (AES), yet prior studies typically examine individual techniques in isolatio...218Gradient Flow Polarizes Softmax Outputs towards Low-...Understanding the intricate non-convex training dynamics of softmax-based models is crucial for explaining the empirical success of transformers. In thi...219Hierarchical Industrial Demand Forecasting with Temp...Hierarchical time-series forecasting is essential for demand prediction across various industries. While machine learning models have obtained significa...220KCLarity at SemEval-2026 Task 6: Encoder and Zero-Sh...This paper describes the KCLarity team's participation in CLARITY, a shared task at SemEval 2026 on classifying ambiguity and evasion techniques in poli...221Large deviation principles for convolutional Bayesia...While suitably scaled CNNs with Gaussian initialization are known to converge to Gaussian processes as the number of channels diverges, little is known...222LiveSense: A Real-Time Wi-Fi Sensing Platform for Ra...We present LiveSense - a cross-platform that transforms a commercial off-the-shelf (COTS) Wi-Fi Network Interface Card (NIC) on a laptop into a centimet...223Mind the Gap: Pitfalls of LLM Alignment with Asian P...Large Language Models (LLMs) are increasingly being deployed in multilingual, multicultural settings, yet their reliance on predominantly English-centri...224Modeling and Measuring Redundancy in Multisource Mul...Next-generation autonomous vehicles (AVs) rely on large volumes of multisource and multimodal ($M^2$) data to support real-time decision-making. In prac...225Multimodal Large Language Models as Image ClassifiersMultimodal Large Language Models (MLLM) classification performance depends critically on evaluation protocol and ground truth quality. Studies comparing...226NEGATE: Constrained Semantic Guidance for Linguistic...Negation is a fundamental linguistic operator, yet it remains inadequately modeled in diffusion-based generative systems. In this work, we present a for...227NOBLE: Accelerating Transformers with Nonlinear Low-...We introduce NOBLE (Nonlinear lOw-rank Branch for Linear Enhancement), an architectural augmentation that adds nonlinear low-rank branches to transforme...228Omni-Diffusion: Unified Multimodal Understanding and...While recent multimodal large language models (MLLMs) have made impressive strides, they predominantly employ a conventional autoregressive architecture...229Penguin-VL: Exploring the Efficiency Limits of VLM w...Vision Language Model (VLM) development has largely relied on scaling model size, which hinders deployment on compute-constrained mobile and edge device...230PONTE: Personalized Orchestration for Natural Langua...Explainable Artificial Intelligence (XAI) seeks to enhance the transparency and accountability of machine learning systems, yet most methods follow a on...231Predictive Coding Graphs are a Superset of Feedforwa...Predictive coding graphs (PCGs) are a recently introduced generalization to predictive coding networks, a neuroscience-inspired probabilistic latent var...232Prosodic Boundary-Aware Streaming Generation for LLM...Streaming TTS that receives streaming text is essential for interactive systems, yet this scheme faces two major challenges: unnatural prosody due to mi...233Quantum Diffusion Models: Score Reversal Is Not Free...Diffusion-based generative modeling suggests reversing a noising semigroup by adding a score drift. For continuous-variable Gaussian Markov dynamics, co...234RAMoEA-QA: Hierarchical Specialization for Robust Re...Conversational generative AI is rapidly entering healthcare, where general-purpose models must integrate heterogeneous patient signals and support diver...235Robust support vector model based on bounded asymmet...In this paper, we propose a novel bounded asymmetric elastic net ($L_{baen}$) loss function and combine it with the support vector machine (SVM), result...236SAHOO: Safeguarded Alignment for High-Order Optimiza...Recursive self-improvement is moving from theory to practice: modern systems can critique, revise, and evaluate their own outputs, yet iterative self-mo...237SCOPE: Scene-Contextualized Incremental Few-Shot 3D...Incremental Few-Shot (IFS) segmentation aims to learn new categories over time from only a few annotations. Although widely studied in 2D, it remains un...238Self-Supervised Flow Matching for Scalable Multi-Mod...Strong semantic representations improve the convergence and generation quality of diffusion and flow models. Existing approaches largely rely on externa...239Semantics-Aware Caching for Concept LearningConcept learning is a form of supervised machine learning that operates on knowledge bases in description logics. State-of-the-art concept learners ofte...240SG-DOR: Learning Scene Graphs with Direction-Conditi...Robotic harvesting in dense crop canopies requires effective interventions that depend not only on geometry, but also on explicit, direction-conditioned...241Spatial Calibration of Diffuse LiDARsDiffuse direct time-of-flight LiDARs report per-pixel depth histograms formed by aggregating photon returns over a wide instantaneous field of view, vio...242Speak in Context: Multilingual ASR with Speech Conte...Automatic speech recognition (ASR) has benefited from advances in pretrained speech and language models, yet most systems remain constrained to monoling...243SPPCSO: Adaptive Penalized Estimation Method for Hig...With the rise of high-dimensional correlated data, multicollinearity poses a significant challenge to model stability, often leading to unstable estimat...244SUREON: A Benchmark and Vision-Language-Model for Su...Surgeons don't just see -- they interpret. When an expert observes a surgical scene, they understand not only what instrument is being used, but why it...245SurgFormer: Scalable Learning of Organ Deformation w...We introduce SurgFormer, a multiresolution gated transformer for data driven soft tissue simulation on volumetric meshes. High fidelity biomechanical so...246Synthetic Monitoring Environments for Reinforcement...Reinforcement Learning (RL) lacks benchmarks that enable precise, white-box diagnostics of agent behavior. Current environments often entangle complexit...247The Art That Poses Back: Assessing AI Pastiches afte...This study explores artificial visual creativity, focusing on ChatGPT's ability to generate new images intentionally pastiching original artworks such a...248The EpisTwin: A Knowledge Graph-Grounded Neuro-Symbo...Personal Artificial Intelligence is currently hindered by the fragmentation of user data across isolated silos. While Retrieval-Augmented Generation off...249Toward Generative Quantum Utility via Correlation-Co...We propose a Correlation-Complexity Map as a practical diagnostic tool for determining when real-world data distributions are structurally aligned with...250Transparent AI for Mathematics: Transformer-Based La...Mathematical text understanding is a challenging task due to the presence of specialized entities and complex relationships between them. This study for...251When One Modality Rules Them All: Backdoor Modality...While diffusion models have revolutionized visual content generation, their rapid adoption has underscored the critical need to investigate vulnerabilit...252Self-Execution Simulation Improves Coding ModelsA promising research direction in enabling LLMs to generate consistently correct code involves addressing their inability to properly estimate program e...253A Quantitative Characterization of Forgetting in Pos...Continual post-training of generative models is widely used, yet a principled understanding of when and why forgetting occurs remains limited. We develo...254Adaptive Conditional Forest Sampling for Spectral Ri...Minimising a spectral risk objective, defined as a convex combination of expected cost and Conditional Value-at-Risk (CVaR), is challenging when the unc...255Budget-Sensitive Discovery Scoring: A Formally Verif...Scientific discovery increasingly relies on AI systems to select candidates for expensive experimental validation, yet no principled, budget-aware evalu...256EB-RANSAC: Random Sample Consensus based on Energy-B...Random sample consensus (RANSAC), which is based on a repetitive sampling from a given dataset, is one of the most popular robust estimation methods. In...257Probabilistic Joint and Individual Variation Explain...Collecting multiple types of data on the same set of subjects is common in modern scientific applications including, genomics, metabolomics, and neuroim...2583DTCR: A Physics-Based Generative Framework for Vort...Tropical cyclone (TC) intensity forecasting remains challenging as current numerical and AI-based weather models fail to satisfactorily represent extrem...259A New Kernel Regularity Condition for Distributed Mi...Existing convergence of distributed optimization methods in non-Euclidean geometries typically rely on kernel assumptions: (i) global Lipschitz smoothne...260A theory of learning data statistics in diffusion mo...While diffusion models have emerged as a powerful class of generative models, their learning dynamics remain poorly understood. We address this issue fi...261Asymptotic and Finite-Time Guarantees for Langevin-B...The InfoNCE loss in contrastive learning depends critically on a temperature parameter, yet its dynamics under fixed versus annealed schedules remain po...262Batched Kernelized Bandits: Refinements and ExtensionsIn this paper, we consider the problem of black-box optimization with noisy feedback revealed in batches, where the unknown function to optimize has a b...263BenDFM: A taxonomy and synthetic CAD dataset for man...Predicting the manufacturability of CAD designs early, in terms of both feasibility and required effort, is a key goal of Design for Manufacturing (DFM)...264Beyond Final Answers: CRYSTAL Benchmark for Transpar...We introduce **CRYSTAL** (*__C__lear __R__easoning via __Y__ielded __S__teps, __T__raceability and __L__ogic*), a diagnostic benchmark with 6,372 instan...265BoSS: A Best-of-Strategies Selector as an Oracle for...Active learning (AL) aims to reduce annotation costs while maximizing model performance by iteratively selecting valuable instances. While foundation mo...266Breaking the Tuning Barrier: Zero-Hyperparameters Yi...Yield Multi-Corner Analysis validates circuits across 25+ Process-Voltage-Temperature corners, resulting in a combinatorial simulation cost of $O(K im...267Causal Cellular Context Transfer Learning (C3TL): An...Predicting the effects of chemical and genetic perturbations on quantitative cell states is a central challenge in computational biology, molecular medi...268CLARIN-PT-LDB: An Open LLM Leaderboard for Portugues...This paper reports on the development of a leaderboard of Open Large Language Models (LLM) for European Portuguese (PT-PT), and on its associated benchm...269Clustering Astronomical Orbital Synthetic Data Using...The dynamics of Saturn's satellite system offer a rich framework for studying orbital stability and resonance interactions. Traditional methods for anal...270Competition-Aware CPC Forecasting with Near-Market C...Cost-per-click (CPC) in paid search is a volatile auction outcome generated by a competitive landscape that is only partially observable from any single...271daVinci-Env: Open SWE Environment Synthesis at ScaleTraining capable software engineering (SWE) agents demands large-scale, executable, and verifiable environments that provide dynamic feedback loops for...272Developing and evaluating a chatbot to support mater...The ability to provide trustworthy maternal health information using phone-based chatbots can have a significant impact, particularly in low-resource se...273Developing the PsyCogMetrics AI Lab to Evaluate Larg...This study presents the development of the PsyCogMetrics AI Lab (psycogmetrics.ai), an integrated, cloud-based platform that operationalizes psychometri...274Diffusion-Based Feature Denoising and Using NNMF for...Brain tumor classification from magnetic resonance imaging, which is also known as MRI, plays a sensitive role in computer-assisted diagnosis systems. I...275DiT-IC: Aligned Diffusion Transformer for Efficient...Diffusion-based image compression has recently shown outstanding perceptual fidelity, yet its practicality is hindered by prohibitive sampling overhead...276DS$^2$-Instruct: Domain-Specific Data Synthesis for...Adapting Large Language Models (LLMs) to specialized domains requires high-quality instruction tuning datasets, which are expensive to create through hu...277ESG-Bench: Benchmarking Long-Context ESG Reports for...As corporate responsibility increasingly incorporates environmental, social, and governance (ESG) criteria, ESG reporting is becoming a legal requiremen...278FDeID-Toolbox: Face De-Identification ToolboxFace de-identification (FDeID) aims to remove personally identifiable information from facial images while preserving task-relevant utility attributes s...279Finite Difference Flow Optimization for RL Post-Trai...Reinforcement learning (RL) has become a standard technique for post-training diffusion-based image synthesis models, as it enables learning from reward...280Fractals made Practical: Denoising Diffusion as Part...What is a diffusion model actually doing when it turns noise into a photograph? We show that the deterministic DDIM reverse chain operates as a Partitio...281From Experiments to Expertise: Scientific Knowledge...While large language models (LLMs) have transformed AI agents into proficient executors of computational materials science, performing a hundred simulat...282GeoChemAD: Benchmarking Unsupervised Geochemical Ano...Geochemical anomaly detection plays a critical role in mineral exploration as deviations from regional geochemical baselines may indicate mineralization...283Geometry-Guided Camera Motion Understanding in Video...Camera motion is a fundamental geometric signal that shapes visual perception and cinematic style, yet current video-capable vision-language models (Vid...284HMS-BERT: Hybrid Multi-Task Self-Training for Multil...Cyberbullying on social media is inherently multilingual and multi-faceted, where abusive behaviors often overlap across multiple categories. Existing m...285Influence Malleability in Linearized Attention: Dual...Understanding the theoretical foundations of attention mechanisms remains challenging due to their complex, non-linear dynamics. This work reveals a fun...286Interpretable Semantic Gradients in SSD: A PCA Sweep...Supervised Semantic Differential (SSD) is a mixed quantitative-interpretive method that models how text meaning varies with continuous individual-differ...287Is Human Annotation Necessary? Iterative MBR Distill...Error Span Detection (ESD) is a crucial subtask in Machine Translation (MT) evaluation, aiming to identify the location and severity of translation erro...288L2GTX: From Local to Global Time Series ExplanationsDeep learning models achieve high accuracy in time series classification, yet understanding their class-level decision behaviour remains challenging. Ex...289Learnability and Privacy Vulnerability are Entangled...Prior approaches for membership privacy preservation usually update or retrain all weights in neural networks, which is costly and can lead to unnecessa...290Learning from Child-Directed Speech in Two-Language...Research on developmentally plausible language models has largely focused on English, leaving open questions about multilingual settings. We present a s...291LLM Constitutional Multi-Agent GovernanceLarge Language Models (LLMs) can generate persuasive influence strategies that shift cooperative behavior in multi-agent populations, but a critical que...292Long-form RewardBench: Evaluating Reward Models for...The widespread adoption of reinforcement learning-based alignment highlights the growing importance of reward models. Various benchmarks have been built...293Mending the Holes: Mitigating Reward Hacking in Rein...Large Language Models (LLMs) have demonstrated remarkable capability in machine translation on high-resource language pairs, yet their performance on lo...294MXNorm: Reusing MXFP block scales for efficient tens...Matrix multiplication performance has long been the major bottleneck to scaling deep learning workloads, which has stimulated the design of new accelera...295Neuron-Aware Data Selection In Instruction Tuning Fo...Instruction Tuning (IT) has been proven to be an effective approach to unlock the powerful capabilities of large language models (LLMs). Recent studies...296NOIR: Neural Operator mapping for Implicit Represent...This paper presents NOIR, a framework that reframes core medical imaging tasks as operator learning between continuous function spaces, challenging the...297Out of Sight, Out of Mind? Evaluating State Evolutio...Evolutions in the world, such as water pouring or ice melting, happen regardless of being observed. Video world models generate worlds via 2D frame ob...298Panoramic Multimodal Semantic Occupancy Prediction f...Panoramic imagery provides holistic 360° visual coverage for perception in quadruped robots. However, existing occupancy prediction methods are mainly d...299Perceive What Matters: Relevance-Driven Scheduling f...In modern human-robot collaboration (HRC) applications, multiple perception modules jointly extract visual, auditory, and contextual cues to achieve com...300PhysMoDPO: Physically-Plausible Humanoid Motion with...Recent progress in text-conditioned human motion generation has been largely driven by diffusion models trained on large-scale human motion data. Buildi...301Representation Learning for Spatiotemporal Physical...Machine learning approaches to spatiotemporal physical systems have primarily focused on next-frame prediction, with the goal of learning an accurate em...302Rethinking Multiple-Choice Questions for RLVR: Unloc...Reinforcement Learning with Verifiable Rewards (RLVR) significantly enhances the reasoning capabilities of Large Language Models. When applied to RLVR,...303Semantic Invariance in Agentic AILarge Language Models (LLMs) increasingly serve as autonomous reasoning agents in decision support, scientific problem-solving, and multi-agent coordina...304Steve-Evolving: Open-World Embodied Self-Evolution v...Open-world embodied agents must solve long-horizon tasks where the main bottleneck is not single-step planning quality but how interaction experience is...305Structured Distillation for Personalized Agent Memor...Long conversations with an AI agent create a simple problem for one user: the history is useful, but carrying it verbatim is expensive. We study persona...306Towards Faithful Multimodal Concept Bottleneck ModelsConcept Bottleneck Models (CBMs) are interpretable models that route predictions through a layer of human-interpretable concepts. While widely studied i...307Towards Spatio-Temporal World Scene Graph Generation...Spatio-temporal scene graphs provide a principled representation for modeling evolving object interactions, yet existing methods remain fundamentally fr...308Variational Garrote for Sparse Inverse ProblemsSparse regularization plays a central role in solving inverse problems arising from incomplete or corrupted measurements. Different regularizers corresp...309VecMol: Vector-Field Representations for 3D Molecule...Generative modeling of three-dimensional (3D) molecules is a fundamental yet challenging problem in drug discovery and materials science. Existing appro...310Visual-ERM: Reward Modeling for Visual EquivalenceVision-to-code tasks require models to reconstruct structured visual inputs, such as charts, tables, and SVGs, into executable or structured representat...311When Right Meets Wrong: Bilateral Context Conditioni...Group Relative Policy Optimization (GRPO) has emerged as an effective method for training reasoning models. While it computes advantages based on group...312When Your Model Stops Working: Anytime-Valid Calibra...Practitioners monitoring deployed probabilistic models face a fundamental trap: any fixed-sample test applied repeatedly over an unbounded stream will e...313ZO-SAM: Zero-Order Sharpness-Aware Minimization for...Deep learning models, despite their impressive achievements, suffer from high computational costs and memory requirements, limiting their usability in r...314AgriIR: A Scalable Framework for Domain-Specific Kno...This paper introduces AgriIR, a configurable retrieval augmented generation (RAG) framework designed to deliver grounded, domain-specific answers while...315Emergent Compositional Communication for Latent Worl...Can multi-agent communication pressure extract discrete, compositional representations of invisible physical properties from frozen video features? We s...316SPRITE: From Static Mockups to Engine-Ready Game UIGame UI implementation requires translating stylized mockups into interactive engine entities. However, current 'Screenshot-to-Code' tools often struggl...317Multi-User Large Language Model AgentsLarge language models (LLMs) and LLM-based agents are increasingly deployed as assistants in planning and decision making, yet most existing systems are...318A Federated Many-to-One Hopfield model for associati...Federated learning enables collaborative training without sharing raw data, but struggles under client heterogeneity and streaming distribution shifts,...319A two-step sequential approach for hyperparameter se...Finite-context models (FCMs) are widely used for compressing symbolic sequences such as DNA, where predictive performance depends critically on the cont...320Adaptive Greedy Frame Selection for Long Video Under...Large vision--language models (VLMs) are increasingly applied to long-video question answering, yet inference is often bottlenecked by the number of inp...321AI Agents Can Already Autonomously Perform Experimen...Large language model-based AI agents are now able to autonomously execute substantial portions of a high energy physics (HEP) analysis pipeline with min...322An Agentic Approach to Generating XAI-NarrativesExplainable AI (XAI) research has experienced substantial growth in recent years. Existing XAI methods, however, have been criticized for being technica...323An Agentic Multi-Agent Architecture for Cybersecurit...Getting a real cybersecurity risk assessment for a small organization is expensive -- a NIST CSF-aligned engagement runs $15,000 on the low end, takes w...324An Empirical Study of SFT-DPO Interaction and Parame...Direct Preference Optimization (DPO) is widely used after supervised fine-tuning (SFT) to align language models, yet empirical behavior under small back...325Beyond Single Tokens: Distilling Discrete Diffusion...It is currently difficult to distill discrete diffusion models. In contrast, continuous diffusion literature has many distillation approaches methods th...326Can Large Multimodal Models Inspect Buildings? A Hie...Automated building facade inspection is a critical component of urban resilience and smart city maintenance. Traditionally, this field has relied on spe...327Chain-of-Adaptation: Surgical Vision-Language Adapta...Conventional fine-tuning on domain-specific datasets can inadvertently alter a model's pretrained multimodal priors, leading to reduced generalization....328Conditioning Protein Generation via Hopfield Pattern...Protein sequence generation via stochastic attention produces plausible family members from small alignments without training, but treats all stored seq...329CoVR-R:Reason-Aware Composed Video RetrievalComposed Video Retrieval (CoVR) aims to find a target video given a reference video and a textual modification. Prior work assumes the modification text...330Current LLMs still cannot 'talk much' about grammar...We aim to examine the extent to which Large Language Models (LLMs) can 'talk much' about grammar modules, providing evidence from syntax core properties...331Deep Autocorrelation Modeling for Time-Series Foreca...Autocorrelation is a defining characteristic of time-series data, where each observation is statistically dependent on its predecessors. In the context...332Design-OS: A Specification-Driven Framework for Engi...Engineering system design -- whether mechatronic, control, or embedded -- often proceeds in an ad hoc manner, with requirements left implicit and tracea...333Deterministic Mode Proposals: An Efficient Alternati...Many segmentation tasks, such as medical image segmentation or future state prediction, are inherently ambiguous, meaning that multiple predictions are...334EgoForge: Goal-Directed Egocentric World SimulatorGenerative world models have shown promise for simulating dynamic environments, yet egocentric video remains challenging due to rapid viewpoint changes,...335Enhancing Hyperspace Analogue to Language (HAL) Repr...The Hyperspace Analogue to Language (HAL) model relies on global word co-occurrence matrices to construct distributional semantic representations. While...336Evaluating Evidence Grounding Under User Pressure in...In contested domains, instruction-tuned language models must balance user-alignment pressures against faithfulness to the in-context evidence. To evalua...337Evolving Jailbreaks: Automated Multi-Objective Long-...Large Language Models (LLMs) have been widely deployed, especially through free Web-based applications that expose them to diverse user-generated inputs...338Explainable cluster analysis: a bagging approachA major limitation of clustering approaches is their lack of explainability: methods rarely provide insight into which features drive the grouping of si...339From Masks to Pixels and Meaning: A New Taxonomy, Be...Existing tampering detection benchmarks largely rely on object masks, which severely misalign with the true edit signal: many pixels inside a mask are u...340GO-GenZip: Goal-Oriented Generative Sampling and Hyb...Current network data telemetry pipelines consist of massive streams of fine-grained Key Performance Indicators (KPIs) from multiple distributed sources...341Graph-Informed Adversarial Modeling: Infimal Subaddi...We study adversarial learning when the target distribution factorizes according to a known Bayesian network. For interpolative divergences, including $(...342Heavy-Tailed and Long-Range Dependent Noise in Stoch...Stochastic approximation (SA) is a fundamental iterative framework with broad applications in reinforcement learning and optimization. Classical analyse...343Improving Generalization on Cybersecurity Tasks with...The use of ML in cybersecurity has long been impaired by generalization issues: Models that work well in controlled scenarios fail to maintain performan...344Improving Image-to-Image Translation via a Rectified...In this work, we propose Image-to-Image Rectified Flow Reformulation (I2I-RFR), a practical plug-in reformulation that recasts standard I2I regression n...345Kolmogorov-Arnold causal generative modelsCausal generative models provide a principled framework for answering observational, interventional, and counterfactual queries from observational data....346LagerNVS: Latent Geometry for Fully Neural Real-time...Recent work has shown that neural networks can perform 3D tasks such as Novel View Synthesis (NVS) without explicit 3D reconstruction. Even so, we argue...347Learning Dynamic Belief Graphs for Theory-of-mind Re...Theory of Mind (ToM) reasoning with Large Language Models (LLMs) requires inferring how people's implicit, evolving beliefs shape what they seek and how...348LoASR-Bench: Evaluating Large Speech Language Models...Large language models (LLMs) have driven substantial advances in speech language models (SpeechLMs), yielding strong performance in automatic speech rec...349LumosX: Relate Any Identities with Their Attributes...Recent advances in diffusion models have significantly improved text-to-video generation, enabling personalized content creation with fine-grained contr...350MeanFlow Meets Control: Scaling Sampled-Data Control...Steering large-scale swarms in only a few control updates is challenging because real systems operate in sampled-data form: control inputs are updated i...351Measuring Faithfulness Depends on How You Measure: C...Recent work on chain-of-thought (CoT) faithfulness reports single aggregate numbers (e.g., DeepSeek-R1 acknowledges hints 39% of the time), implying tha...352Minimax Generalized Cross-EntropyLoss functions play a central role in supervised classification. Cross-entropy (CE) is widely used, whereas the mean absolute error (MAE) loss can offer...353MME-CoF-Pro: Evaluating Reasoning Coherence in Video...Video generative models show emerging reasoning behaviors. It is essential to ensure that generated events remain causally consistent across frames for...354Model Selection and Parameter Estimation of Multi-di...In this paper, we study the problem of learning multi-dimensional Gaussian Mixture Models (GMMs), with a specific focus on model order selection and eff...355MuSteerNet: Human Reaction Generation from Videos vi...Video-driven human reaction generation aims to synthesize 3D human motions that directly react to observed video sequences, which is crucial for buildin...356PersonaVLM: Long-Term Personalized Multimodal LLMsMultimodal Large Language Models (MLLMs) serve as daily assistants for millions. However, their ability to generate responses aligned with individual pr...357Predicting States of Understanding in Explanatory In...We investigate how verbal and nonverbal linguistic features, exhibited by speakers and listeners in dialogue, can contribute to predicting the listener'...358Reasoning Gets Harder for LLMs Inside A DialogueLarge Language Models (LLMs) achieve strong performance on many reasoning benchmarks, yet these evaluations typically focus on isolated tasks that diffe...359Regularity of Solutions to Beckmann's Parametric Opt...Beckmann's problem in optimal transport minimizes the total squared flux in a continuous transport problem from a source to a target distribution. In th...360Revisiting Gene Ontology Knowledge Discovery with Hi...Large language models have achieved great success in multiple challenging tasks, and their capacity can be further boosted by the emerging agentic AI te...361ReViSQL: Achieving Human-Level Text-to-SQLTranslating natural language to SQL (Text-to-SQL) is a critical challenge in both database research and data analytics applications. Recent efforts have...362RouterKGQA: Specialized--General Model Routing for C...Knowledge graph question answering (KGQA) is a promising approach for mitigating LLM hallucination by grounding reasoning in structured and verifiable k...363Scalable Learning of Multivariate Distributions via...Efficient and scalable non-parametric or semi-parametric regression analysis and density estimation are of crucial importance to the fields of statistic...364Semantic Token Clustering for Efficient Uncertainty...Large language models (LLMs) have demonstrated remarkable capabilities across diverse tasks. However, the truthfulness of their outputs is not guarantee...365Spectral Alignment in Forward-Backward Representatio...Forward-backward (FB) representations provide a powerful framework for learning the successor representation (SR) in continuous spaces by enforcing a lo...366The $\mathbf{Y}$-Combinator for LLMs: Solving Long-C...LLMs are increasingly used as general-purpose reasoners, but long inputs remain bottlenecked by a fixed context window. Recursive Language Models (RLMs)...367The monotonicity of the Franz-Parisi potential is eq...Over the last decades, two distinct approaches have been instrumental to our understanding of the computational complexity of statistical estimation. Th...368The Robot's Inner Critic: Self-Refinement of Social...Conventional robot social behavior generation has been limited in flexibility and autonomy, relying on predefined motions or human feedback. This study...369TinyML Enhances CubeSat Mission CapabilitiesEarth observation (EO) missions traditionally rely on transmitting raw or minimally processed imagery from satellites to ground stations for computation...370Trojan horse hunt in deep forecasting models: Insigh...Forecasting plays a crucial role in modern safety-critical applications, such as space operations. However, the increasing use of deep forecasting model...371Two-Time-Scale Learning Dynamics: A Population View...Population-based learning paradigms, including evolutionary strategies, Population-Based Training (PBT), and recent model-merging methods, combine fast...372Uncertainty Quantification Via the Posterior Predict...We use the law of total variance to generate multiple expansions for the posterior predictive variance. These expansions are sums of terms involving con...373Var-JEPA: A Variational Formulation of the Joint-Emb...The Joint-Embedding Predictive Architecture (JEPA) is often seen as a non-generative alternative to likelihood-based self-supervised learning, emphasizi...374VideoSeek: Long-Horizon Video Agent with Tool-Guided...Video agentic models have advanced challenging video-language tasks. However, most agentic approaches still heavily rely on greedy parsing over densely...375When Contextual Inference Fails: Cancelability in In...We investigate the separation of literal interpretation from contextual inference in a collaborative block-building task where a builder must resolve un...376Wildfire Spread Scenarios: Increasing Sample Diversi...Predicting future states in uncertain environments, such as wildfire spread, medical diagnosis, or autonomous driving, requires models that can consider...377How to Fine-Tune a Reasoning Model? A Teacher-Studen...A widely adopted strategy for model enhancement is to use synthetic data generated by a stronger model for supervised fine-tuning (SFT). However, for em...378QuanBench+: A Unified Multi-Framework Benchmark for...Large Language Models (LLMs) are increasingly used for code generation, yet quantum code generation is still evaluated mostly within single frameworks,...379Robust Reasoning BenchmarkWhile Large Language Models (LLMs) achieve high performance on standard mathematical benchmarks, their underlying reasoning processes remain highly over...380SEVerA: Verified Synthesis of Self-Evolving AgentsRecent advances have shown the effectiveness of self-evolving LLM agents on tasks such as program repair and scientific discovery. In this paradigm, a p...381Scaling Teams or Scaling Time? Memory Enabled Lifelo...Large language model (LLM) multi-agent systems can scale along two distinct dimensions: by increasing the number of agents and by improving through accu...382Tunable Soft Equivariance with GuaranteesEquivariance is a fundamental property in computer vision models, yet strict equivariance is rarely satisfied in real-world data, which can limit a mode...383Mimic Intent, Not Just TrajectoriesWhile imitation learning (IL) has achieved impressive success in dexterous manipulation through generative modeling and pretraining, state-of-the-art ap...384Structural Graph Probing of Vision-Language ModelsVision-language models (VLMs) achieve strong multimodal performance, yet how computation is organized across populations of neurons remains poorly under...385AgentSwing: Adaptive Parallel Context Management Rou...As large language models (LLMs) evolve into autonomous agents for long-horizon information-seeking, managing finite context capacity has become a critic...386Learning to Retrieve from Agent TrajectoriesInformation retrieval (IR) systems have traditionally been designed and trained for human users, with learning-to-rank methods relying heavily on large-...387LIBERO-Para: A Diagnostic Benchmark and Metrics for...Vision-Language-Action (VLA) models achieve strong performance in robotic manipulation by leveraging pre-trained vision-language backbones. However, in...388Backdoor Attacks on Decentralised Post-TrainingDecentralised post-training of large language models utilises data and pipeline parallelism techniques to split the data and the model. Unfortunately, d...389Learning to Hint for Reinforcement LearningGroup Relative Policy Optimization (GRPO) is widely used for reinforcement learning with verifiable rewards, but it often suffers from advantage collaps...390Locally Confident, Globally Stuck: The Quality-Explo...Diffusion large language models (dLLMs) theoretically permit token decoding in arbitrary order, a flexibility that could enable richer exploration of re...391ONE-SHOT: Compositional Human-Environment Video Synt...Recent advances in Video Foundation Models (VFMs) have revolutionized human-centric video synthesis, yet fine-grained and independent editing of subject...392Type-Checked Compliance: Deterministic Guardrails fo...The rapid evolution of autonomous, agentic artificial intelligence within financial services has introduced an existential architectural crisis: large l...393Adam's Law: Textual Frequency Law on Large Language...While textual frequency has been validated as relevant to human cognition in reading speed, its relatedness to Large Language Models (LLMs) is seldom st...394Automating Database-Native Function Code Synthesis w...Database systems incorporate an ever-growing number of functions in their kernels (a.k.a., database native functions) for scenarios like new application...395BidirLM: From Text to Omnimodal Bidirectional Encode...Transforming causal generative language models into bidirectional encoders offers a powerful alternative to BERT-style architectures. However, current a...396Expert-Choice Routing Enables Adaptive Computation i...Diffusion language models (DLMs) enable parallel, non-autoregressive text generation, yet existing DLM mixture-of-experts (MoE) models inherit token-cho...397Learning interacting particle systems from unlabeled...Learning the potentials of interacting particle systems is a fundamental task across various scientific disciplines. A major challenge is that unlabeled...398Learning to Learn-at-Test-Time: Language Agents with...Test-Time Learning (TTL) enables language agents to iteratively refine their performance through repeated interactions with the environment at inference...399PLUME: Latent Reasoning Based Universal Multimodal E...Universal multimodal embedding (UME) maps heterogeneous inputs into a shared retrieval space with a single model. Recent approaches improve UME by gener...400ThinkTwice: Jointly Optimizing Large Language Models...We introduce ThinkTwice, a simple two-phase framework that jointly optimizes LLMs to solve reasoning problems and refine the answers, based on Group Rel...401Unifying Group-Relative and Self-Distillation Policy...Reinforcement learning with verifiable rewards (RLVR) has become a standard paradigm for post-training large language models. While Group Relative Polic...402A Systematic Security Evaluation of OpenClaw and Its...Tool-augmented AI agents substantially extend the practical capabilities of large language models, but they also introduce security risks that cannot be...403A Tsetlin Machine-driven Intrusion Detection System...The rapid adoption of the Internet of Medical Things (IoMT) is transforming healthcare by enabling seamless connectivity among medical devices, systems,...404AI-Assisted Unit Test Writing and Test-Driven Code R...Many software systems originate as prototypes or minimum viable products (MVPs), developed with an emphasis on delivery speed and responsiveness to chan...405An Independent Safety Evaluation of Kimi K2.5Kimi K2.5 is an open-weight LLM that rivals closed models across coding, multimodal, and agentic benchmarks, but was released without an accompanying sa...406BAS: A Decision-Theoretic Approach to Evaluating Lar...Large language models (LLMs) often produce confident but incorrect answers in settings where abstention would be safer. Standard evaluation protocols, h...407Beyond Precision: Importance-Aware Recall for Factua...Evaluating the factuality of long-form output generated by large language models (LLMs) remains challenging, particularly when responses are open-ended...408Beyond the Parameters: A Technical Survey of Context...Large language models (LLMs) encode vast world knowledge in their parameters, yet they remain fundamentally limited by static knowledge, finite context...409BibTeX Citation Hallucinations in Scientific Publish...Large language models with web search are increasingly used in scientific publishing agents, yet they still produce BibTeX entries with pervasive field-...410CAMEO: A Conditional and Quality-Aware Multi-Agent I...Conditional image editing aims to modify a source image according to textual prompts and optional reference guidance. Such editing is crucial in scenari...411Characterization of Gaussian Universality Breakdown...We study high-dimensional convex empirical risk minimization (ERM) under general non-Gaussian data designs. By heuristically extending the Convex Gaussi...412Chart-RL: Policy Optimization Reinforcement Learning...The recent advancements in Vision Language Models (VLMs) have demonstrated progress toward true intelligence requiring robust reasoning capabilities. Be...413CoME-VL: Scaling Complementary Multi-Encoder Vision-...Recent vision-language models (VLMs) typically rely on a single vision encoder trained with contrastive image-text objectives, such as CLIP-style pretra...414Coupled Control, Structured Memory, and Verifiable A...Agentic AI is increasingly judged not by fluent output alone but by whether it can act, remember, and verify under partial observability, delay, and str...415Detecting and Correcting Reference Hallucinations in...Large language models and deep research agents supply citation URLs to support their claims, yet the reliability of these citations has not been systema...416Do Audio-Visual Large Language Models Really See and...Audio-Visual Large Language Models (AVLLMs) are emerging as unified interfaces to multimodal perception. We present the first mechanistic interpretabili...417Domain-Adapted Retrieval for In-Context Annotation o...Automated annotation of pedagogical dialogue is a high-stakes task where LLMs often fail without sufficient domain grounding. We present a domain-adapte...418DSBD: Dual-Aligned Structural Basis Distillation for...Graph domain adaptation (GDA) aims to transfer knowledge from a labeled source graph to an unlabeled target graph under distribution shifts. However, ex...419EffiMiniVLM: A Compact Dual-Encoder Regression Frame...Predicting product quality from multimodal item information is critical in cold-start scenarios, where user interaction history is unavailable and predi...420Enhancing Robustness of Federated Learning via Serve...This paper explores the use of server learning for enhancing the robustness of federated learning against malicious attacks even when clients' training...421Escape dynamics and implicit bias of one-pass SGD in...We analyze the one-pass stochastic gradient descent dynamics of a two-layer neural network with quadratic activations in a teacher--student framework. I...422Frame Theoretical Derivation of Three Factor Learnin...We show that the error-gated Hebbian rule for PCA (EGHR-PCA), a three-factor learning rule equivalent to Oja's subspace rule under Gaussian inputs, can...423GBQA: A Game Benchmark for Evaluating LLMs as Qualit...The autonomous discovery of bugs remains a significant challenge in modern software development. Compared to code generation, the complexity of dynamic...424Generating DDPM-based Samples from Tilted DistributionsGiven $n$ independent samples from a $d$-dimensional probability distribution, our aim is to generate diffusion-based samples from a distribution obtain...425Gradient Boosting within a Single Attention LayerTransformer attention computes a single softmax-weighted average over values -- a one-pass estimate that cannot correct its own errors. We introduce \em...426Hierarchical Planning with Latent World ModelsModel predictive control (MPC) with learned world models has emerged as a promising paradigm for embodied control, particularly for its ability to gener...427High-dimensional Many-to-many-to-many Mediation Anal...We study high-dimensional mediation analysis in which exposures, mediators, and outcomes are all multivariate, and both exposures and mediators may be h...428HyperCT: Low-Rank Hypernet for Unified Chest CT Anal...Non-contrast chest CTs offer a rich opportunity for both conventional pulmonary and opportunistic extra-pulmonary screening. While Multi-Task Learning (...429HyperFitS -- Hypernetwork Fitting Spectra for metabo...Purpose: Proton magnetic resonance spectroscopic imaging ($^1$H MRSI) enables the mapping of whole-brain metabolites concentrations in-vivo. However, a...430InCoder-32B-Thinking: Industrial Code World Model fo...Industrial software development across chip design, GPU optimization, and embedded systems lacks expert reasoning traces showing how engineers reason ab...431Inversion-Free Natural Gradient Descent on Riemannia...The natural gradient method is widely used in statistical optimization, but its standard formulation assumes a Euclidean parameter space. This paper pro...432Large Language Models Align with the Human Brain dur...Creative thinking is a fundamental aspect of human cognition, and divergent thinking-the capacity to generate novel and varied ideas-is widely regarded...433Learning the Signature of Memorization in Autoregres...All prior membership inference attacks for fine-tuned language models use hand-crafted heuristics (e.g., loss thresholding, Min-K\%, reference calibrati...434Lipschitz bounds for integral kernelsFeature maps associated with positive definite kernels play a central role in kernel methods and learning theory, where regularity properties such as Li...435Low-Rank Compression of Pretrained Models via Random...The massive scale of pretrained models has made efficient compression essential for practical deployment. Low-rank decomposition based on the singular v...436Multi-Aspect Knowledge Distillation for Language Mod...Knowledge distillation is an effective technique for pre-trained language model compression. However, existing methods only focus on the knowledge distr...437Multi-View Video Diffusion Policy: A 3D Spatio-Tempo...Robotic manipulation requires understanding both the 3D spatial structure of the environment and its temporal evolution, yet most existing policies over...438Power one sequential tests exist for weakly compact...Suppose we observe data from a distribution $P$ and we wish to test the composite null hypothesis that $P\in\mathscr P$ against a composite alternative...439PR3DICTR: A modular AI framework for medical 3D imag...Three-dimensional medical image data and computer-aided decision making, particularly using deep learning, are becoming increasingly important in the me...440PRISM: LLM-Guided Semantic Clustering for High-Preci...In this paper, we propose Precision-Informed Semantic Modeling (PRISM), a structured topic modeling framework combining the benefits of rich representat...441ProtoFlow: Mitigating Forgetting in Class-Incrementa...Remote sensing segmentation in real deployment is inherently continual: new semantic categories emerge, and acquisition conditions shift across seasons,...442Quotient-Based Posterior Analysis for Euclidean Late...Latent space models are widely used in statistical network analysis and are often fit by Markov chain Monte Carlo. However, posterior summaries of laten...443Real-Time Surrogate Modeling for Personalized Blood...Cardiovascular modeling has rapidly advanced over the past few decades due to the rising needs for health tracking and early detection of cardiovascular...444Reflective Context Learning: Studying the Optimizati...Generally capable agents must learn from experience in ways that generalize across tasks and environments. The fundamental problems of learning, includi...445Reliability Gated Multi-Teacher Distillation for Low...We study multiteacher knowledge distillation for low resource abstractive summarization from a reliability aware perspective. We introduce EWAD (Entropy...446Rethinking Forward Processes for Score-Based Data As...Data assimilation is the process of estimating the time-evolving state of a dynamical system by integrating model predictions and noisy observations. It...447Salt: Self-Consistent Distribution Matching with Cac...Distilling video generation models to extremely low inference budgets (e.g., 2--4 NFEs) is crucial for real-time deployment, yet remains challenging. Tr...448SCC-Loc: A Unified Semantic Cascade Consensus Framew...Cross-modal Thermal Geo-localization (TG) provides a robust, all-weather solution for Unmanned Aerial Vehicles (UAVs) in Global Navigation Satellite Sys...449SD-FSMIS: Adapting Stable Diffusion for Few-Shot Med...Few-Shot Medical Image Segmentation (FSMIS) aims to segment novel object classes in medical images using only minimal annotated examples, addressing the...450Self-Distilled RLVROn-policy distillation (OPD) has become a popular training paradigm in the LLM community. This paradigm selects a larger model as the teacher to provide...451Semantic Richness or Geometric Reasoning? The Fragil...This work investigates the fundamental fragility of state-of-the-art Vision-Language Models (VLMs) under basic geometric transformations. While modern V...452SFFNet: Synergistic Feature Fusion Network With Dual...Object detection in unmanned aerial vehicle (UAV) images remains a highly challenging task, primarily caused by the complexity of background noise and t...453Significance and Stability Analysis of Gene-Environm...Genotype-by-Environment (GxE) interactions influence the performance of genotypes across diverse environments, reducing the predictability of phenotypes...454State estimations and noise identifications with int...This paper focuses on the state estimation problem in distributed sensor networks, where intermittent packet dropouts, corrupted observations, and unkno...455StoryScope: Investigating idiosyncrasies in AI fictionAs AI-generated fiction becomes increasingly prevalent, questions of authorship and originality are becoming central to how written work is evaluated. W...456Structure-Preserving Multi-View Embedding Using Grom...Multi-view data analysis seeks to integrate multiple representations of the same samples in order to recover a coherent low-dimensional structure. Class...457The Compression Gap: Why Discrete Tokenization Limit...Scaling Vision-Language-Action (VLA) models by upgrading the vision encoder is expected to improve downstream manipulation performance--as it does in vi...458The Eleventh NTIRE 2026 Efficient Super-Resolution C...This paper reviews the NTIRE 2026 challenge on efficient single-image super-resolution with a focus on the proposed solutions and results. The aim of th...459Transfer Learning for Meta-analysis Under Covariate...Randomized controlled trials often do not represent the populations where decisions are made, and covariate shift across studies can invalidate standard...460Understanding the Role of Hallucination in Reinforce...The recent success of reinforcement learning (RL) in large reasoning models has inspired the growing adoption of RL for post-training Multimodal Large L...461Valence-Arousal Subspace in LLMs: Circular Emotion G...We present a method to identify a valence-arousal (VA) subspace within large language model representations. From 211k emotion-labeled texts, we derive...462VOSR: A Vision-Only Generative Model for Image Super...Most of the recent generative image super-resolution (SR) methods rely on adapting large text-to-image (T2I) diffusion models pretrained on web-scale te...463Can Natural Image Autoencoders Compactly Tokenize fM...Modeling long-range spatiotemporal dynamics in functional Magnetic Resonance Imaging (fMRI) remains a key challenge due to the high dimensionality of th...464LightThinker++: From Reasoning Compression to Memory...Large language models (LLMs) excel at complex reasoning, yet their efficiency is limited by the surging cognitive overhead of long thought traces. In th...465POEMetric: The Last Stanza of HumanityLarge Language Models (LLMs) can compose poetry, but how far are they from human poets? In this paper, we introduce POEMetric, the first comprehensive f...466SciLT: Long-Tailed Classification in Scientific Imag...Long-tailed recognition has benefited from foundation models and fine-tuning paradigms, yet existing studies and benchmarks are mainly confined to natur...467Squeez: Task-Conditioned Tool-Output Pruning for Cod...Coding agents repeatedly consume long tool observations even though only a small fraction of each observation matters for the next step. We study task-c...468Training a Student Expert via Semi-Supervised Founda...Foundation models deliver strong perception but are often too computationally heavy to deploy, and adapting them typically requires costly annotations....469A Systematic Study of Cross-Modal Typographic Attack...As audio-visual multi-modal large language models (MLLMs) are increasingly deployed in safety-critical applications, understanding their vulnerabilities...470ACES: Who Tests the Tests? Leave-One-Out AUC Consist...Selecting LLM-generated code candidates using LLM-generated tests is challenging because the tests themselves may be incorrect. Existing methods either...471AURA: Always-On Understanding and Real-Time Assistan...Video Large Language Models (VideoLLMs) have achieved strong performance on many video understanding tasks, but most existing systems remain offline and...472Cactus: Accelerating Auto-Regressive Decoding with C...Speculative sampling (SpS) has been successful in accelerating the decoding throughput of auto-regressive large language models by leveraging smaller dr...473Can LLMs Learn to Reason Robustly under Noisy Superv...Reinforcement Learning with Verifiable Rewards (RLVR) effectively trains reasoning models that rely on abundant perfect labels, but its vulnerability to...474ClawArena: Benchmarking AI Agents in Evolving Inform...AI agents deployed as persistent assistants must maintain correct beliefs as their information environment evolves. In practice, evidence is scattered a...475Combee: Scaling Prompt Learning for Self-Improving L...Recent advances in prompt learning allow large language model agents to acquire task-relevant knowledge from inference-time context without parameter ch...476DARE: Diffusion Large Language Models Alignment and...Diffusion large language models (dLLMs) are emerging as a compelling alternative to dominant autoregressive models, replacing strictly sequential token...477QEIL v2: Heterogeneous Computing for Edge Intelligen...Deploying large language models (LLMs) on heterogeneous edge devices demands frameworks that jointly optimize energy efficiency, inference quality, and...478StateSMix: Online Lossless Compression via Mamba Sta...We present StateSMix, a fully self-contained lossless compressor that couples an online-trained Mamba-style State Space Model (SSM) with sparse n-gram c...479The Geometric Alignment Tax: Tokenization vs. Contin...Foundation models for biology and physics optimize predictive accuracy, but their internal representations systematically fail to preserve the continuou...480A Frame is Worth One Token: Efficient Generative Wor...Anticipating diverse future states is a central challenge in video world modeling. Discriminative world models produce a deterministic prediction that i...481AvatarPointillist: AutoRegressive 4D Gaussian Avatar...We introduce AvatarPointillist, a novel framework for generating dynamic 4D Gaussian avatars from a single portrait image. At the core of our method is...482Beyond Hard Negatives: The Importance of Score Distr...Transferring knowledge from a cross-encoder teacher via Knowledge Distillation (KD) has become a standard paradigm for training retrieval models. While...483ClawsBench: Evaluating Capability and Safety of LLM...Large language model (LLM) agents are increasingly deployed to automate productivity tasks (e.g., email, scheduling, document management), but evaluatin...484CLEAR: Unlocking Generative Potential for Degraded I...Image degradation from blur, noise, compression, and poor illumination severely undermines multimodal understanding in real-world settings. Unified mult...485Cog-DRIFT: Exploration on Adaptively Reformulated In...Reinforcement learning from verifiable rewards (RLVR) has improved the reasoning abilities of LLMs, yet a fundamental limitation remains: models cannot...486Demystifying When Pruning Works via Representation H...Network pruning, which removes less important parameters or architectures, is often expected to improve efficiency while preserving performance. However...487DeonticBench: A Benchmark for Reasoning over RulesReasoning with complex, context-specific rules remains challenging for large language models (LLMs). In legal and policy settings, this manifests as deo...488FileGram: Grounding Agent Personalization in File-Sy...Coworking AI agents operating within local file systems are rapidly emerging as a paradigm in human-AI interaction; however, effective personalization r...489General Multimodal Protein Design Enables DNA-Encodi...Evolution is an extraordinary engine for enzymatic diversity, yet the chemistry it has explored remains a narrow slice of what DNA can encode. Deep gene...490HDP: A Lightweight Cryptographic Protocol for Human...Agentic AI systems increasingly execute consequential actions on behalf of human principals, delegating tasks through multi-step chains of autonomous ag...491How Well Do Agentic Skills Work in the Wild: Benchma...Agent skills, which are reusable, domain-specific knowledge artifacts, have become a popular mechanism for extending LLM-based agents, yet formally benc...492Less Detail, Better Answers: Degradation-Driven Prom...Recent advancements in Vision-Language Models (VLMs) have significantly pushed the boundaries of Visual Question Answering (VQA).However,high-resolution...493MedGemma 1.5 Technical ReportWe introduce MedGemma 1.5 4B, the latest model in the MedGemma collection. MedGemma 1.5 expands on MedGemma 1 by integrating additional capabilities: hi...494MegaTrain: Full Precision Training of 100B+ Paramete...We present MegaTrain, a memory-centric system that efficiently trains 100B+ parameter large language models at full precision on a single GPU. Unlike tr...495Memory Intelligence AgentDeep research agents (DRAs) integrate LLM reasoning with external tools. Memory systems enable DRAs to leverage historical experiences, which are essent...496MinerU2.5-Pro: Pushing the Limits of Data-Centric Do...Current document parsing methods compete primarily on model architecture innovation, while systematic engineering of training data remains underexplored...497OpenWorldLib: A Unified Codebase and Definition of A...World models have garnered significant attention as a promising research direction in artificial intelligence, yet a clear and unified definition remain...498Paper Espresso: From Paper Overload to Research InsightThe accelerating pace of scientific publishing makes it increasingly difficult for researchers to stay current. We present Paper Espresso, an open-sourc...499R3PM-Net: Real-time, Robust, Real-world Point Matchi...Accurate Point Cloud Registration (PCR) is an important task in 3D data processing, involving the estimation of a rigid transformation between two point...500REAM: Merging Improves Pruning of Experts in LLMsMixture-of-Experts (MoE) large language models (LLMs) are among the top-performing architectures. The largest models, often with hundreds of billions of...501SkillX: Automatically Constructing Skill Knowledge B...Learning from experience is critical for building capable large language model (LLM) agents, yet prevailing self-evolving paradigms remain inefficient:...502SkVM: Compiling Skills for Efficient Execution Every...LLM agents increasingly adopt skills as a reusable unit of composition. While skills are shared across diverse agent platforms, current systems treat th...503SpatialEdit: Benchmarking Fine-Grained Image Spatial...Image spatial editing performs geometry-driven transformations, allowing precise control over object layout and camera viewpoints. Current models are in...504Structured Causal Video Reasoning via Multi-Objectiv...Human understanding of video dynamics is typically grounded in a structured mental representation of entities, actions, and temporal relations, rather t...505SuperLocalMemory V3.3: The Living Brain -- Biologica...AI coding agents operate in a paradox: they possess vast parametric knowledge yet cannot remember a conversation from an hour ago. Existing memory syste...506Synthetic Sandbox for Training Machine Learning Engi...As large language model agents advance beyond software engineering (SWE) tasks toward machine learning engineering (MLE), verifying agent behavior becom...507TriAttention: Efficient Long Reasoning with Trigonom...Extended reasoning in large language models (LLMs) creates severe KV cache memory bottlenecks. Leading KV cache compression methods estimate KV importan...508Vanast: Virtual Try-On with Human Image Animation vi...We present Vanast, a unified framework that generates garment-transferred human animation videos directly from a single human image, garment images, and...509Vero: An Open RL Recipe for General Visual ReasoningWhat does it take to build a visual reasoner that works across charts, science, spatial understanding, and open-ended tasks? The strongest vision-langua...510Video-MME-v2: Towards the Next Stage in Benchmarks f...With the rapid advancement of video understanding, existing benchmarks are becoming increasingly saturated, exposing a critical discrepancy between infl...511Watch Before You Answer: Learning from Visually Grou...It is critical for vision-language models (VLMs) to comprehensively understand visual, temporal, and textual cues. However, despite rapid progress in mu...512Your Agent, Their Asset: A Real-World Safety Analysi...OpenClaw, the most widely deployed personal AI agent in early 2026, operates with full local system access and integrates with sensitive services such a...513Action Images: End-to-End Policy Learning via Multiv...World action models (WAMs) have emerged as a promising direction for robot policy learning, as they can leverage powerful video backbones to model the f...514AgentGL: Towards Agentic Graph Learning with LLMs vi...Large Language Models (LLMs) increasingly rely on agentic capabilities-iterative retrieval, tool use, and decision-making-to overcome the limits of stat...515Beyond Accuracy: Unveiling Inefficiency Patterns in...In real-world Tool-Integrated Reasoning (TIR) scenarios, where LLMs interleave reasoning with external tool calls, a major source of inefficiency is tha...516Claw-Eval: Toward Trustworthy Evaluation of Autonomo...Large language models are increasingly deployed as autonomous agents executing multi-step workflows in real-world software environments. However, existi...517Context-Value-Action Architecture for Value-Driven L...Large Language Models (LLMs) have shown promise in simulating human behavior, yet existing agents often exhibit behavioral rigidity, a flaw frequently m...518CUE-R: Beyond the Final Answer in Retrieval-Augmente...As language models shift from single-shot answer generation toward multi-step reasoning that retrieves and consumes evidence mid-inference, evaluating t...519Experience Transfer for Multimodal LLM Agents in Min...Multimodal LLM agents operating in complex game environments must continually reuse past experience to solve new tasks efficiently. In this work, we pro...520FactReview: Evidence-Grounded Reviews with Literatur...Peer review in machine learning is under growing pressure from rising submission volume and limited reviewer time. Most LLM-based reviewing systems read...521Graph of Skills: Dependency-Aware Structural Retriev...Skill usage has become a core component of modern agent systems and can substantially improve agents' ability to complete complex tasks. In real-world s...522Graph-Based Chain-of-Thought Pruning for Reducing Re...Extending CoT through RL has been widely used to enhance the reasoning capabilities of LLMs. However, due to the sparsity of reward signals, it can also...523Improving Semantic Proximity in Information Retrieva...With the increasing accessibility and utilization of multilingual documents, Cross-Lingual Information Retrieval (CLIR) has emerged as an important rese...524In-Place Test-Time TrainingThe static ``train then deploy' paradigm fundamentally limits Large Language Models (LLMs) from dynamically adapting their weights in response to contin...525MedConclusion: A Benchmark for Biomedical Conclusion...Large language models (LLMs) are widely explored for reasoning-intensive research tasks, yet resources for testing whether they can infer scientific con...526MMEmb-R1: Reasoning-Enhanced Multimodal Embedding wi...MLLMs have been successfully applied to multimodal embedding tasks, yet their generative reasoning capabilities remain underutilized. Directly incorpora...527Neural ComputersWe propose a new frontier: Neural Computers (NCs) -- an emerging machine form that unifies computation, memory, and I/O in a learned runtime state. Unli...528Paper Circle: An Open-source Multi-agent Research Di...The rapid growth of scientific literature has made it increasingly difficult for researchers to efficiently discover, evaluate, and synthesize relevant...529QiMeng-PRepair: Precise Code Repair via Edit-Aware R...Large Language Models (LLMs) achieve strong program repair performance but often suffer from over-editing, where excessive modifications overwrite corre...530Qualixar OS: A Universal Operating System for AI Age...We present Qualixar OS, the first application-layer operating system for universal AI agent orchestration. Unlike kernel-level approaches (AIOS) or sing...531RAGEN-2: Reasoning Collapse in Agentic RLRL training of multi-turn LLM agents is inherently unstable, and reasoning quality directly determines task performance. Entropy is widely used to track...532Scientific Graphics Program Synthesis via Dual Self-...Graphics Program Synthesis is pivotal for interpreting and editing visual data, effectively facilitating the reverse-engineering of static visuals into...533Spec Kit Agents: Context-Grounded Agentic WorkflowsSpec-driven development (SDD) with AI coding agents provides a structured workflow, but agents often remain 'context blind' in large, evolving repositor...534Target Policy OptimizationIn RL, given a prompt, we sample a group of completions from a model and score them. Two questions follow: which completions should gain probability mas...535The Depth Ceiling: On the Limits of Large Language M...The viability of chain-of-thought (CoT) monitoring hinges on models being unable to reason effectively in their latent representations. Yet little is kn...536The Master Key Hypothesis: Unlocking Cross-Model Cap...We investigate whether post-trained capabilities can be transferred across models without retraining, with a focus on transfer across different model sc...537TRACE: Capability-Targeted Agentic TrainingLarge Language Models (LLMs) deployed in agentic environments must exercise multiple capabilities across different task instances, where a capability is...538Appear2Meaning: A Cross-Cultural Benchmark for Struc...Recent advances in vision-language models (VLMs) have improved image captioning for cultural heritage. However, inferring structured cultural metadata (...539ATANT: An Evaluation Framework for AI ContinuityWe present ATANT (Automated Test for Acceptance of Narrative Truth), an open evaluation framework for measuring continuity in AI systems: the ability to...540CylinderDepth: Cylindrical Spatial Attention for Mul...Self-supervised surround-view depth estimation enables dense, low-cost 3D perception with a 360° field of view from multiple minimally overlapping image...541Fast Spatial Memory with Elastic Test-Time TrainingLarge Chunk Test-Time Training (LaCT) has shown strong performance on long-context 3D reconstruction, but its fully plastic inference-time updates remai...542FlowInOne:Unifying Multimodal Generation as Image-in...Multimodal generation has long been dominated by text-driven pipelines where language dictates vision but cannot reason or create within it. We challeng...543Flux Attention: Context-Aware Hybrid Attention for E...The quadratic computational complexity of standard attention mechanisms presents a severe scalability bottleneck for LLMs in long-context scenarios. Whi...544FORGE:Fine-grained Multimodal Evaluation for Manufac...The manufacturing sector is increasingly adopting Multimodal Large Language Models (MLLMs) to transition from simple perception to autonomous execution,...545FP4 Explore, BF16 Train: Diffusion Reinforcement Lea...Reinforcement-Learning-based post-training has recently emerged as a promising paradigm for aligning text-to-image diffusion models with human preferenc...546GameWorld: Towards Standardized and Verifiable Evalu...Towards an embodied generalist for real-world interaction, Multimodal Large Language Model (MLLM) agents still suffer from challenging latency, sparse f...547Generate, Filter, Control, Replay: A Comprehensive S...Reinforcement learning (RL) has become a central post-training tool for improving the reasoning abilities of large language models (LLMs). In these syst...548GenLCA: 3D Diffusion for Full-Body Avatars from In-t...We present GenLCA, a diffusion-based generative model for generating and editing photorealistic full-body avatars from text and image inputs. The genera...549HY-Embodied-0.5: Embodied Foundation Models for Real...We introduce HY-Embodied-0.5, a family of foundation models specifically designed for real-world embodied agents. To bridge the gap between general Visi...550INSPATIO-WORLD: A Real-Time 4D World Simulator via S...Building world models with spatial consistency and real-time interactivity remains a fundamental challenge in computer vision. Current video generation...551MARS: Enabling Autoregressive Models Multi-Token Gen...Autoregressive (AR) language models generate text one token at a time, even when consecutive tokens are highly predictable given earlier context. We int...552MoRight: Motion Control Done RightGenerating motion-controlled videos--where user-specified actions drive physically plausible scene dynamics under freely chosen viewpoints--demands two...553On the Step Length Confounding in LLM Reasoning Data...Large reasoning models have recently demonstrated strong performance on complex tasks that require long chain-of-thought reasoning, through supervised f...554OpenSpatial: A Principled Data Engine for Empowering...Spatial understanding is a fundamental cornerstone of human-level intelligence. Nonetheless, current research predominantly focuses on domain-specific d...555Personalized RewardBench: Evaluating Reward Models w...Pluralistic alignment has emerged as a critical frontier in the development of Large Language Models (LLMs), with reward models (RMs) serving as a centr...556Personalizing Text-to-Image Generation to Individual...Modern text-to-image (T2I) models generate high-fidelity visuals but remain indifferent to individual user preferences. While existing reward models opt...557POS-ISP: Pipeline Optimization at the Sequence Level...Recent work has explored optimizing image signal processing (ISP) pipelines for various tasks by composing predefined modules and adapting them to task-...558Q-Zoom: Query-Aware Adaptive Perception for Efficien...MLLMs require high-resolution visual inputs for fine-grained tasks like document understanding and dense scene perception. However, current global resol...559RefineAnything: Multimodal Region-Specific Refinemen...We introduce region-specific image refinement as a dedicated problem setting: given an input image and a user-specified region (e.g., a scribble mask or...560Rethinking Generalization in Reasoning SFT: A Condit...A prevailing narrative in LLM post-training holds that supervised finetuning (SFT) memorizes while reinforcement learning (RL) generalizes. We revisit t...561TC-AE: Unlocking Token Capacity for Deep Compression...We propose TC-AE, a ViT-based architecture for deep compression autoencoders. Existing methods commonly increase the channel number of latent representa...562Think in Strokes, Not Pixels: Process-Driven Image G...Humans paint images incrementally: they plan a global layout, sketch a coarse draft, inspect, and refine details, and most importantly, each step is gro...5633D-VCD: Hallucination Mitigation in 3D-LLM Embodied...Large multimodal models are increasingly used as the reasoning core of embodied agents operating in 3D environments, yet they remain prone to hallucinat...564Accurate and Reliable Uncertainty Estimates for Dete...Computational models support high-stakes decisions across engineering and science, and practitioners increasingly seek probabilistic predictions to quan...565Act Wisely: Cultivating Meta-Cognitive Tool Use in A...The advent of agentic multimodal models has empowered systems to actively interact with external environments. However, current agents suffer from a pro...566AnomalyVFM -- Transforming Vision Foundation Models...Zero-shot anomaly detection aims to detect and localise abnormal regions in the image without access to any in-domain training images. While recent appr...567AVGen-Bench: A Task-Driven Benchmark for Multi-Granu...Text-to-Audio-Video (T2AV) generation is rapidly becoming a core interface for media creation, yet its evaluation remains fragmented. Existing benchmark...568Beyond Stochastic Exploration: What Makes Training D...Reinforcement learning (RL) has become an effective approach for advancing the reasoning capabilities of large language models (LLMs) through the strate...569ClawBench: Can AI Agents Complete Everyday Online Ta...AI agents may be able to automate your inbox, but can they automate other routine aspects of your life? Everyday online tasks offer a realistic yet unso...570Cram Less to Fit More: Training Data Pruning Improve...Large language models (LLMs) can struggle to memorize factual knowledge in their parameters, often leading to hallucinations and poor performance on kno...571Cross-Modal Emotion Transfer for Emotion Editing in...Talking face generation has gained significant attention as a core application of generative models. To enhance the expressiveness and realism of synthe...572DMax: Aggressive Parallel Decoding for dLLMsWe present DMax, a new paradigm for efficient diffusion language models (dLLMs). It mitigates error accumulation in parallel decoding, enabling aggressi...573Efficient RL Training for LLMs with Experience ReplayWhile Experience Replay - the practice of storing rollouts and reusing them multiple times during training - is a foundational technique in general RL,...574EXAONE 4.5 Technical ReportThis technical report introduces EXAONE 4.5, the first open-weight vision language model released by LG AI Research. EXAONE 4.5 is architected by integr...575Externalization in LLM Agents: A Unified Review of M...Large language model (LLM) agents are increasingly built less by changing model weights than by reorganizing the runtime around them. Capabilities that...576Faithful GRPO: Improving Visual Spatial Reasoning in...Multimodal reasoning models (MRMs) trained with reinforcement learning with verifiable rewards (RLVR) show improved accuracy on visual reasoning benchma...577FIT: A Large-Scale Dataset for Fit-Aware Virtual Try-OnGiven a person and a garment image, virtual try-on (VTO) aims to synthesize a realistic image of the person wearing the garment, while preserving their...578ImplicitMemBench: Measuring Unconscious Behavioral A...Existing memory benchmarks for LLM agents evaluate explicit recall of facts, yet overlook implicit memory where experience becomes automated behavior wi...579Initialisation Determines the Basin: Efficient Codeb...Additive quantization enables extreme LLM compression with O(1) lookup-table dequantization, making it attractive for edge deployment. Yet at 2-bit prec...580KnowU-Bench: Towards Interactive, Proactive, and Per...Personalized mobile agents that infer user preferences and calibrate proactive assistance hold great promise as everyday digital assistants, yet existin...581Lighting-grounded Video Generation with Renderer-bas...Diffusion models have achieved remarkable progress in video generation, but their controllability remains a major limitation. Key scene factors such as...582LPM 1.0: Video-based Character Performance ModelPerformance, the externalization of intent, emotion, and personality through visual, vocal, and temporal behavior, is what makes a character alive. Lear...583MegaStyle: Constructing Diverse and Scalable Style D...In this paper, we introduce MegaStyle, a novel and scalable data curation pipeline that constructs an intra-style consistent, inter-style diverse and hi...584Meta-learning In-Context Enables Training-Free Cross...Visual decoding from brain signals is a key challenge at the intersection of computer vision and neuroscience, requiring methods that bridge neural repr...585MolmoWeb: Open Visual Web Agent and Open Data for th...Web agents--autonomous systems that navigate and execute tasks on the web on behalf of users--have the potential to transform how people interact with t...586OmniJigsaw: Enhancing Omni-Modal Reasoning via Modal...To extend the reinforcement learning post-training paradigm to omni-modal models for concurrently bolstering video-audio understanding and collaborative...587On Semiotic-Grounded Interpretive Evaluation of Gene...Interpretation is essential to deciphering the language of art: audiences communicate with artists by recovering meaning from visual artifacts. However,...588On the Global Photometric Alignment for Low-Level Vi...Supervised low-level vision models rely on pixel-wise losses against paired references, yet paired training sets exhibit per-pair photometric inconsiste...589OpenVLThinkerV2: A Generalist Multimodal Reasoning M...Group Relative Policy Optimization (GRPO) has emerged as the de facto Reinforcement Learning (RL) objective driving recent advancements in Multimodal La...590p1: Better Prompt Optimization with Fewer PromptsPrompt optimization improves language models without updating their weights by searching for a better system prompt, but its effectiveness varies widely...591Phantom: Physics-Infused Video Generation via Joint...Recent advances in generative video modeling, driven by large-scale datasets and powerful architectures, have yielded remarkable visual realism. However...592PokeGym: A Visually-Driven Long-Horizon Benchmark fo...While Vision-Language Models (VLMs) have achieved remarkable progress in static visual understanding, their deployment in complex 3D embodied environmen...593Policy-Aware Design of Large-Scale Factorial Experim...Digital firms routinely run many online experiments on shared user populations. When product decisions are compositional, such as combinations of interf...594ReconPhys: Reconstruct Appearance and Physical Attri...Reconstructing non-rigid objects with physical plausibility remains a significant challenge. Existing approaches leverage differentiable rendering for p...595RewardFlow: Generate Images by Optimizing What You R...We introduce RewardFlow, an inversion-free framework that steers pretrained diffusion and flow-matching models at inference time through multi-reward La...596SIM1: Physics-Aligned Simulator as Zero-Shot Data Sc...Robotic manipulation with deformable objects represents a data-intensive regime in embodied learning, where shape, contact, and topology co-evolve in wa...597SkillClaw: Let Skills Evolve Collectively with Agent...Large language model (LLM) agents such as OpenClaw rely on reusable skills to perform complex tasks, yet these skills remain largely static after deploy...598Small Vision-Language Models are Smart Compressors f...Adapting Multimodal Large Language Models (MLLMs) for hour-long videos is bottlenecked by context limits. Dense visual streams saturate token budgets an...599Structured Distillation of Web Agent Capabilities En...Frontier LLMs can navigate complex websites, but their cost and reliance on third-party APIs make local deployment impractical. We introduce Agent-as-An...600Towards Real-world Human Behavior Simulation: Benchm...The emergence of Large Language Models (LLMs) has illuminated the potential for a general-purpose user simulator. However, existing benchmarks remain co...601Uni-ViGU: Towards Unified Video Generation and Under...Unified multimodal models integrating visual understanding and generation face a fundamental challenge: visual generation incurs substantially higher co...602ViVa: A Video-Generative Value Model for Robot Reinf...Vision-language-action (VLA) models have advanced robot manipulation through large-scale pretraining, but real-world deployment remains challenging due...603What do Language Models Learn and When? The Implicit...Large language models (LLMs) can perform remarkably complex tasks, yet the fine-grained details of how these capabilities emerge during pretraining rema...604When Numbers Speak: Aligning Textual Numerals and Vi...Text-to-video diffusion models have enabled open-ended video synthesis, but often struggle with generating the correct number of objects specified in a...605WildDet3D: Scaling Promptable 3D Detection in the WildUnderstanding objects in 3D from a single image is a cornerstone of spatial intelligence. A key step toward this goal is monocular 3D object detection--...606A novel hybrid approach for positive-valued DAG lear...Causal discovery from observational data remains a fundamental challenge in machine learning and statistics, particularly when variables represent inher...607A Predictive View on Streaming Hidden Markov ModelsWe develop a predictive-first optimisation framework for streaming hidden Markov models. Unlike classical approaches that prioritise full posterior reco...608Across the Levels of Analysis: Explaining Predictive...Under the lens of Marr's levels of analysis, we critique and extend two claims about language models (LMs) and language processing: first, that predicti...609AdaCubic: An Adaptive Cubic Regularization Optimizer...A novel regularization technique, AdaCubic, is proposed that adapts the weight of the cubic term. The heart of AdaCubic is an auxiliary optimization pro...610Agentic Jackal: Live Execution and Semantic Value Gr...Translating natural language into Jira Query Language (JQL) requires resolving ambiguous field references, instance-specific categorical values, and com...611An Open-Source, Open Data Approach to Activity Class...The accelerometer has become an almost ubiquitous device, providing enormous opportunities in healthcare monitoring beyond step counting or other averag...612ANTIC: Adaptive Neural Temporal In-situ CompressorThe persistent storage requirements for high-resolution, spatiotemporally evolving fields governed by large-scale and high-dimensional partial different...613AsymLoc: Towards Asymmetric Feature Matching for Eff...Precise and real-time visual localization is critical for applications like AR/VR and robotics, especially on resource-constrained edge devices such as...614Automated Instruction Revision (AIR): A Structured C...This paper studies Automated Instruction Revision (AIR), a rule-induction-based method for adapting large language models (LLMs) to downstream tasks usi...615BERT-as-a-Judge: A Robust Alternative to Lexical Met...Accurate evaluation is central to the large language model (LLM) ecosystem, guiding model selection and downstream adoption across diverse use cases. In...616Beyond Augmented-Action Surrogates for Multi-Expert...Learning-to-Defer routes each input to the expert that minimizes expected cost, but it assumes that the information available to every expert is fixed a...617Case-Grounded Evidence Verification: A Framework for...Evidence-grounded reasoning requires more than attaching retrieved text to a prediction: a model should make decisions that depend on whether the provid...618CONSCIENTIA: Can LLM Agents Learn to Strategize? Eme...As large language models (LLMs) are increasingly deployed as autonomous agents, understanding how strategic behavior emerges in multi-agent environments...619Continuous Orthogonal Mode Decomposition: Haptic Sig...The Tactile Internet demands sub-millisecond latency and ultra-high reliability, as high latency or packet loss could lead to haptic control instability...620CT-1: Vision-Language-Camera Models Transfer Spatial...Camera-controllable video generation aims to synthesize videos with flexible and physically plausible camera movements. However, existing methods either...621Data-Efficient Non-Gaussian Semi-Nonparametric Densi...Accurate representation of non-Gaussian distributions of quantities of interest in nonlinear dynamical systems is critical for estimation, control, and...622Do AI Coding Agents Log Like Humans? An Empirical StudySoftware logging is essential for maintaining and debugging complex systems, yet it remains unclear how AI coding agents handle this non-functional requ...623DSVTLA: Deep Swin Vision Transformer-Based Transfer...In this study, we proposed a deep Swin-Vision Transformer-based transfer learning architecture for robust multi-cancer histopathological image classific...624E3-TIR: Enhanced Experience Exploitation for Tool-In...While Large Language Models (LLMs) have demonstrated significant potential in Tool-Integrated Reasoning (TIR), existing training paradigms face signific...625ECHO: Efficient Chest X-ray Report Generation with O...Chest X-ray report generation (CXR-RG) has the potential to substantially alleviate radiologists' workload. However, conventional autoregressive vision-...626EgoTL: Egocentric Think-Aloud Chains for Long-Horizo...Large foundation models have made significant advances in embodied intelligence, enabling synthesis and reasoning over egocentric input for household ta...627ELT: Elastic Looped Transformers for Visual GenerationWe introduce Elastic Looped Transformers (ELT), a highly parameter-efficient class of visual generative models based on a recurrent transformer architec...628Envisioning the Future, One Step at a TimeAccurately anticipating how complex, diverse scenes will evolve requires models that represent uncertainty, simulate along extended interaction chains,...629EquiformerV3: Scaling Efficient, Expressive, and Gen...As SE(3)-equivariant graph neural networks mature as a core tool for 3D atomistic modeling, improving their efficiency, expressivity, and physical consi...630Event-Driven Temporal Graph Networks for Asynchronou...The transition of Multi-Agent Reinforcement Learning (MARL) policies from simulated cyber wargames to operational Security Operations Centers (SOCs) is...631From Reasoning to Agentic: Credit Assignment in Rein...Reinforcement learning (RL) for large language models (LLMs) increasingly relies on sparse, outcome-level rewards -- yet determining which actions withi...632Generalization and Scaling Laws for Mixture-of-Exper...We develop a theory of generalization and scaling for Mixture-of-Experts (MoE) Transformers that cleanly separates \emph{active} per-input capacity from...633Hierarchical Kernel Transformer: Multi-Scale Attenti...The Hierarchical Kernel Transformer (HKT) is a multi-scale attention mechanism that processes sequences at L resolution levels via trainable causal down...634Hierarchical SVG Tokenization: Learning Compact Visu...Recent large language models have shifted SVG generation from differentiable rendering optimization to autoregressive program synthesis. However, existi...635High-dimensional Adaptive MCMC with Reduced Computat...We propose an adaptive MCMC method that learns a linear preconditioner which is dense in its off-diagonal elements but sparse in its parametrisation. Du...636Identifying Causal Effects Using a Single Proxy Vari...Unobserved confounding is a key challenge when estimating causal effects from a treatment on an outcome in scientific applications. In this work, we ass...637Incremental Semantics-Aided Meshing from LiDAR-Inert...Geometric high-fidelity mesh reconstruction from LiDAR-inertial scans remains challenging in large, complex indoor environments -- such as cultural buil...638Integrated electro-optic attention nonlinearities fo...Transformers have emerged as the dominant neural-network architecture, achieving state-of-the-art performance in language processing and computer vision...639Is More Data Worth the Cost? Dataset Scaling Laws in...Training Transformer language models is expensive, as performance typically improves with increasing dataset size and computational budget. Although sca...640Iterative Identification Closure: Amplifying Causal...The Half-Trek Criterion (HTC) is the primary graphical tool for determining generic identifiability of causal effect coefficients in linear structural e...641Large Language Models Generate Harmful Content Using...Large language models (LLMs) undergo alignment training to avoid harmful behaviors, yet the resulting safeguards remain brittle: jailbreaks routinely by...642Many Ways to Be Fake: Benchmarking Fake News Detecti...Recent advances in large language models (LLMs) have enabled the large-scale generation of highly fluent and deceptive news-like content. While prior wo...643Many-Tier Instruction Hierarchy in LLM AgentsLarge language model agents receive instructions from many sources-system messages, user prompts, tool outputs, and more-each carrying different levels...644Matrix-Game 3.0: Real-Time and Streaming Interactive...With the advancement of interactive video generation, diffusion models have increasingly demonstrated their potential as world models. However, existing...645MixFlow: Mixed Source Distributions Improve Rectifie...Diffusion models and their variations, such as rectified flows, generate diverse and high-quality images, but they are still hindered by slow iterative...646Online Quantile Regression for Nonparametric Additiv...This paper introduces a projected functional gradient descent algorithm (P-FGD) for training nonparametric additive quantile regression models in online...647Online3R: Online Learning for Consistent Sequential...We present Online3R, a new sequential reconstruction framework that is capable of adapting to new scenes through online learning, effectively resolving...648Process Reward Agents for Steering Knowledge-Intensi...Reasoning in knowledge-intensive domains remains challenging as intermediate steps are often not locally verifiable: unlike math or code, evaluating ste...649Rays as Pixels: Learning A Joint Distribution of Vid...Recovering camera parameters from images and rendering scenes from novel viewpoints have long been treated as separate tasks in computer vision and grap...650Realizing Immersive Volumetric Video: A Multimodal F...Fully immersive experiences that tightly integrate 6-DoF visual and auditory interaction are essential for virtual and augmented reality. While such exp...651RecaLLM: Addressing the Lost-in-Thought Phenomenon w...We propose RecaLLM, a set of reasoning language models post-trained to make effective use of long-context information. In-context retrieval, which ident...652RIRF: Reasoning Image Restoration FrameworkUniversal image restoration (UIR) aims to recover clean images from diverse and unknown degradations using a unified model. Existing UIR methods primari...653SafeAdapt: Provably Safe Policy Updates in Deep Rein...Safety guarantees are a prerequisite to the deployment of reinforcement learning (RL) agents in safety-critical tasks. Often, deployment environments ex...654SafeMind: A Risk-Aware Differentiable Control Framew...Learning-based quadruped controllers achieve impressive agility but typically lack formal safety guarantees under model uncertainty, perception noise, a...655ScheMatiQ: From Research Question to Structured Data...Many disciplines pose natural-language research questions over large document collections whose answers typically require structured evidence, tradition...656SCoRe: Clean Image Generation from Diffusion Models...Diffusion models trained on noisy datasets often reproduce high-frequency training artifacts, significantly degrading generation quality. To address thi...657Seeing is Believing: Robust Vision-Guided Cross-Moda...Prompt learning is a parameter-efficient approach for vision-language models, yet its robustness under label noise is less investigated. Visual content...658Semantic Rate-Distortion for Bounded Multi-Agent Com...When two agents of different computational capacities interact with the same environment, they need not compress a common semantic alphabet differently;...659Sharp description of local minima in the loss landsc...We study the population loss landscape of two-layer ReLU networks of the form $\sum_{k=1}^K \mathrm{ReLU}(w_k^\top x)$ in a realisable teacher-student s...660Sim-to-Real Transfer for Muscle-Actuated Robots via...Tendon drives paired with soft muscle actuation enable faster and safer robots while potentially accelerating skill acquisition. Still, these systems ar...661SPASM: Stable Persona-driven Agent Simulation for Mu...Large language models are increasingly deployed in multi-turn settings such as tutoring, support, and counseling, where reliability depends on preservin...662SPPO: Sequence-Level PPO for Long-Horizon Reasoning...Proximal Policy Optimization (PPO) is central to aligning Large Language Models (LLMs) in reasoning tasks with verifiable rewards. However, standard tok...663Strategic Algorithmic Monoculture:Experimental Evide...AI agents increasingly operate in multi-agent environments where outcomes depend on coordination. We distinguish primary algorithmic monoculture -- base...664Strips as Tokens: Artist Mesh Generation with Native...Recent advancements in autoregressive transformers have demonstrated remarkable potential for generating artist-quality meshes. However, the token order...665TAIHRI: Task-Aware 3D Human Keypoints Localization f...Accurate 3D human keypoints localization is a critical technology enabling robots to achieve natural and safe physical interaction with users. Conventio...666Tango: Taming Visual Signals for Efficient Video Lar...Token pruning has emerged as a mainstream approach for developing efficient Video Large Language Models (Video LLMs). This work revisits and advances th...667Toward World Models for EpidemiologyWorld models have emerged as a unifying paradigm for learning latent dynamics, simulating counterfactual futures, and supporting planning under uncertai...668U-Cast: A Surprisingly Simple and Efficient Frontier...AI-based weather forecasting now rivals traditional physics-based ensembles, but state-of-the-art (SOTA) models rely on specialized architectures and ma...669UIPress: Bringing Optical Token Compression to UI-to...UI-to-Code generation requires vision-language models (VLMs) to produce thousands of tokens of structured HTML/CSS from a single screenshot, making visu...670VisionFoundry: Teaching VLMs Visual Perception with...Vision-language models (VLMs) still struggle with visual perception tasks such as spatial understanding and viewpoint recognition. One plausible contrib...671VISOR: Agentic Visual Retrieval-Augmented Generation...Visual Retrieval-Augmented Generation (VRAG) empowers Vision-Language Models to retrieve and reason over visually rich documents. To tackle complex quer...672VL-Calibration: Decoupled Confidence Calibration for...Large Vision Language Models (LVLMs) achieve strong multimodal reasoning but frequently exhibit hallucinations and incorrect responses with high certain...673XFED: Non-Collusive Model Poisoning Attack Against B...Model poisoning attacks pose a significant security threat to Federated Learning (FL). Most existing model poisoning attacks rely on collusion, requirin...674You Can't Fight in Here! This is BBS!Norm, the formal theoretical linguist, and Claudette, the computational language scientist, have a lovely time discussing whether modern language models...675A Temporally Augmented Graph Attention Network for A...Graph attention networks (GATs) provide one of the best frameworks for learning node representations in relational data; but, existing variants such as...676Attention Sink in Transformers: A Survey on Utilizat...As the foundational architecture of modern machine learning, Transformers have driven remarkable progress across diverse AI domains. Despite their trans...677Counting to Four is still a Chore for VLMsVision--language models (VLMs) have achieved impressive performance on complex multimodal reasoning tasks, yet they still fail on simple grounding skill...678EditCrafter: Tuning-free High-Resolution Image Editi...We propose EditCrafter, a high-resolution image editing method that operates without tuning, leveraging pretrained text-to-image (T2I) diffusion models...679Not All Denoising Steps Are Equal: Model Scheduling...Recent advances in masked diffusion language models (MDLMs) narrow the quality gap to autoregressive LMs, but their sampling remains expensive because g...680Prompt Relay: Inference-Time Temporal Control for Mu...Video diffusion models have achieved remarkable progress in generating high-quality videos. However, these models struggle to represent the temporal suc...681Zero-shot World Models Are Developmentally Efficient...Young children demonstrate early abilities to understand their physical world, estimating depth, motion, object coherence, interactions, and many other...682Advancing Polish Language Modeling through Tokenizer...The development of the Bielik v3 PL series, encompassing both the 7B and 11B parameter variants, represents a significant milestone in the field of lang...683Audio-Omni: Extending Multi-modal Understanding to V...Recent progress in multimodal models has spurred rapid advances in audio understanding, generation, and editing. However, these capabilities are typical...684BMdataset: A Musicologically Curated LilyPond DatasetSymbolic music research has relied almost exclusively on MIDI-based datasets; text-based engraving formats such as LilyPond remain unexplored for music...685DiningBench: A Hierarchical Multi-view Benchmark for...Recent advancements in Vision-Language Models (VLMs) have revolutionized general visual understanding. However, their application in the food domain rem...686IceCache: Memory-efficient KV-cache Management for L...Key-Value (KV) cache plays a crucial role in accelerating inference in large language models (LLMs) by storing intermediate attention states and avoidin...687PersonalAI: A Systematic Comparison of Knowledge Gra...Personalizing language models by effectively incorporating user interaction history remains a central challenge in the development of adaptive AI system...688PokeRL: Reinforcement Learning for Pokemon RedPokemon Red is a long-horizon JRPG with sparse rewards, partial observability, and quirky control mechanics that make it a challenging benchmark for rei...689Rethinking the Diffusion Model from a Langevin Persp...Diffusion models are often introduced from multiple perspectives, such as VAEs, score matching, or flow matching, accompanied by dense and technically d...690SciPredict: Can LLMs Predict the Outcomes of Scienti...Accelerating scientific discovery requires the identification of which experiments would yield the best outcomes before committing resources to costly p...691SCOPE: Signal-Calibrated On-Policy Distillation Enha...On-policy reinforcement learning has become the dominant paradigm for reasoning alignment in large language models, yet its sparse, outcome-level reward...692The Blind Spot of Agent Safety: How Benign User Inst...Computer-use agents (CUAs) can now autonomously complete complex tasks in real digital environments, but when misled, they can also be used to automate...693TorchUMM: A Unified Multimodal Model Codebase for Ev...Recent advances in unified multimodal models (UMMs) have led to a proliferation of architectures capable of understanding, generating, and editing acros...694Tracing the Roots: A Multi-Agent Framework for Uncov...Post-training data plays a pivotal role in shaping the capabilities of Large Language Models (LLMs), yet datasets are often treated as isolated artifact...695When Reasoning Models Hurt Behavioral Simulation: A...Large language models are increasingly used as agents in social, economic, and policy simulations. A common assumption is that stronger reasoning should...6963DTV: A Feedforward Interpolation Network for Real-T...Real-time free-viewpoint rendering requires balancing multi-camera redundancy with the latency constraints of interactive applications. We address this...697ADD for Multi-Bit Image WatermarkingAs generative models enable rapid creation of high-fidelity images, societal concerns about misinformation and authenticity have intensified. A promisin...698Agentic Aggregation for Parallel Scaling of Long-Hor...We study parallel test-time scaling for long-horizon agentic tasks such as agentic search and deep research, where multiple rollouts are generated in pa...699Anthropogenic Regional Adaptation in Multimodal Visi...While the field of vision-language (VL) has achieved remarkable success in integrating visual and textual information across multiple languages and doma...700Audio Flamingo Next: Next-Generation Open Audio-Lang...We present Audio Flamingo Next (AF-Next), the next-generation and most capable large audio-language model in the Audio Flamingo series, designed to adva...701Beyond Perception Errors: Semantic Fixation in Large...Large vision-language models (VLMs) often rely on familiar semantic priors, but existing evaluations do not cleanly separate perception failures from ru...702ClawGUI: A Unified Framework for Training, Evaluatin...GUI agents drive applications through their visual interfaces instead of programmatic APIs, interacting with arbitrary software via taps, swipes, and ke...703CocoaBench: Evaluating Unified Digital Agents in the...LLM agents now perform strongly in software engineering, deep research, GUI automation, and various other applications, while recent agent scaffolds and...704CodeTracer: Towards Traceable Agent StatesCode agents are advancing rapidly, but debugging them is becoming increasingly difficult. As frameworks orchestrate parallel tool calls and multi-stage...705Continuous Adversarial Flow ModelsWe propose continuous adversarial flow models, a type of continuous-time flow model trained with an adversarial objective. Unlike flow matching, which u...706Cross-Tokenizer LLM Distillation through a Byte-Leve...Cross-tokenizer distillation (CTD), the transfer of knowledge from a teacher to a student language model when the two use different tokenizers, remains...707Do Thought Streams Matter? Evaluating Reasoning in G...We benchmark how internal reasoning traces, which we call thought streams, affect video scene understanding in vision-language models. Using four config...708Eliciting Medical Reasoning with Knowledge-enhanced...While large language models hold promise for complex medical applications, their development is hindered by the scarcity of high-quality reasoning data....709From Reasoning to Agentic: Credit Assignment in Rein...Reinforcement learning (RL) for large language models (LLMs) increasingly relies on sparse, outcome-level rewards -- yet determining which actions withi...710General365: Benchmarking General Reasoning in Large...Contemporary large language models (LLMs) have demonstrated remarkable reasoning capabilities, particularly in specialized domains like mathematics and...711HDR Video Generation via Latent Alignment with Logar...High dynamic range (HDR) imagery offers a rich and faithful representation of scene radiance, but remains challenging for generative models due to its m...712How Alignment Routes: Localizing, Scaling, and Contr...This paper localizes the policy routing mechanism in alignment-trained language models. An intermediate-layer attention gate reads detected content and...713Introspective Diffusion Language ModelsDiffusion language models promise parallel generation, yet still lag behind autoregressive (AR) models in quality. We stem this gap to a failure of intr...714LARY: A Latent Action Representation Yielding Benchm...While the shortage of explicit action data limits Vision-Language-Action (VLA) models, human action videos offer a scalable yet unlabeled data source. A...715LASA: Language-Agnostic Semantic Alignment at the Se...Large language models (LLMs) often demonstrate strong safety performance in high-resource languages, yet exhibit severe vulnerabilities when queried in...716Learning Long-term Motion Embeddings for Efficient K...Understanding and predicting motion is a fundamental component of visual intelligence. Although modern video models exhibit strong comprehension of scen...717Low-rank Optimization Trajectories Modeling for LLM...Recently, scaling reinforcement learning with verifiable rewards (RLVR) for large language models (LLMs) has emerged as an effective training paradigm f...718Mobile GUI Agent Privacy Personalization with Trajec...Mobile GUI agents powered by Multimodal Large Language Models (MLLMs) can execute complex tasks on mobile devices. Despite this progress, most existing...719Narrative-Driven Paper-to-Slide Generation via ArcDeckWe introduce ArcDeck, a multi-agent framework that formulates paper-to-slide generation as a structured narrative reconstruction task. Unlike existing m...720OccuBench: Evaluating AI Agents on Real-World Profes...AI agents are expected to perform professional work across hundreds of occupational domains (from emergency department triage to nuclear reactor safety...721OmniScript: Towards Audio-Visual Script Generation f...Current multimodal large language models (MLLMs) have demonstrated remarkable capabilities in short-form video understanding, yet translating long-form...722OmniShow: Unifying Multimodal Conditions for Human-O...In this work, we study Human-Object Interaction Video Generation (HOIVG), which aims to synthesize high-quality human-object interaction videos conditio...723Panoptic Pairwise Distortion GraphIn this work, we introduce a new perspective on comparative image assessment by representing an image pair as a structured composition of its regions. I...724Playing Along: Learning a Double-Agent Defender for...As large language models (LLMs) become the engine behind conversational systems, their ability to reason about the intentions and states of their dialog...725Polyglot Teachers: Evaluating Language Models for Mu...Synthesizing supervised finetuning (SFT) data from language models (LMs) to teach smaller models multilingual tasks has become increasingly common. Howe...726Predicting integers from continuous parametersWe study the problem of predicting numeric labels that are constrained to the integers or to a subrange of the integers. For example, the number of up-v...727Pseudo-Unification: Entropy Probing Reveals Divergen...Unified multimodal models (UMMs) were designed to combine the reasoning ability of large language models (LLMs) with the generation capability of vision...728RationalRewards: Reasoning Rewards Scale Visual Gene...Most reward models for visual generation reduce rich human judgments to a single unexplained score, discarding the reasoning that underlies preference....729Representations Before Pixels: Semantics-Guided Hier...Accurate future video prediction requires both high visual fidelity and consistent scene semantics, particularly in complex dynamic environments such as...730Seeing Through Touch: Tactile-Driven Visual Localiza...We address the problem of tactile localization, where the goal is to identify image regions that share the same material properties as a tactile input....731Self-Distillation Zero: Self-Revision Turns Binary R...Current post-training methods in verifiable settings fall into two categories. Reinforcement learning (RLVR) relies on binary rewards, which are broadly...732Self-Evolving LLM Memory Extraction Across Heterogen...As LLM-based assistants become persistent and personalized, they must extract and retain useful information from past conversations as memory. However,...733Sema Code: Decoupling AI Coding Agents into Programm...AI coding agents have become central to developer workflows, yet every existing solution locks its reasoning capabilities within a specific delivery for...734SemaClaw: A Step Towards General-Purpose Personal AI...The rise of OpenClaw in early 2026 marks the moment when millions of users began deploying personal AI agents into their daily lives, delegating tasks r...735SHARE: Social-Humanities AI for Research and EducationThis intermediate technical report introduces the SHARE family of base models and the MIRROR user interface. The SHARE models are the first causal langu...736Solving Physics Olympiad via Reinforcement Learning...We have witnessed remarkable advances in LLM reasoning capabilities with the advent of DeepSeek-R1. However, much of this progress has been fueled by th...737SWE-AGILE: A Software Agent Framework for Efficientl...Prior representative ReAct-style approaches in autonomous Software Engineering (SWE) typically lack the explicit System-2 reasoning required for deep an...738The Past Is Not Past: Memory-Enhanced Dynamic Reward...Despite the success of reinforcement learning for large language models, a common failure mode is reduced sampling diversity, where the policy repeatedl...739Time is Not a Label: Continuous Phase Rotation for T...Structured memory representations such as knowledge graphs are central to autonomous agents and other long-lived systems. However, most existing approac...740You Only Judge Once: Multi-response Reward Modeling...We present a discriminative multimodal reward model that scores all candidate responses in a single forward pass. Conventional discriminative reward mod...741Accelerating Speculative Decoding with Block Diffusi...Speculative decoding accelerates autoregressive language models by using a lightweight drafter to propose multiple future tokens, which the target model...742AgentSPEX: An Agent SPecification and EXecution Lang...Language-model agent systems commonly rely on reactive prompting, in which a single instruction guides the model through an open-ended sequence of reaso...743ASGuard: Activation-Scaling Guard to Mitigate Target...Large language models (LLMs), despite being safety-aligned, exhibit brittle refusal behaviors that can be circumvented by simple linguistic changes. As...744Balanced Aggregation: Understanding and Fixing Aggre...Reinforcement learning with verifiable rewards (RLVR) has become a central paradigm for improving reasoning and code generation in large language models...745Boosting Visual Instruction Tuning with Self-Supervi...Multimodal large language models (MLLMs) perform well on many vision-language tasks but often struggle with vision-centric problems that require fine-gr...746Concrete Jungle: Towards Concreteness Paved Contrast...Vision-Language Models demonstrate remarkable capabilities but often struggle with compositional reasoning, exhibiting vulnerabilities regarding word or...747Dive into Claude Code: The Design Space of Today's a...Claude Code is an agentic coding tool that can run shell commands, edit files, and call external services on behalf of the user. This study describes it...748Domain-Specific Latent Representations Improve the F...Latent diffusion models for medical image super-resolution universally inherit variational autoencoders designed for natural photographs. We show that t...749Exploration and Exploitation Errors Are Measurable f...Language Model (LM) agents are increasingly used in complex open-ended decision-making tasks, from AI coding to physical AI. A core requirement in these...750Forge-UGC: FX optimization and register-graph engine...We present Forge-UGC (FX Optimization and Register-Graph Engine for Universal Graph Compilation), a four-phase compiler for transformer deployment on he...751Generative Refinement Networks for Visual SynthesisWhile diffusion models dominate the field of visual generation, they are computationally inefficient, applying a uniform computational effort regardless...752GlotOCR Bench: OCR Models Still Struggle Beyond a Ha...Optical character recognition (OCR) has advanced rapidly with the rise of vision-language models, yet evaluation has remained concentrated on a small cl...753Grid2Matrix: Revealing Digital Agnosia in Vision-Lan...Vision-Language Models (VLMs) excel on many multimodal reasoning benchmarks, but these evaluations often do not require an exhaustive readout of the ima...754Habitat-GS: A High-Fidelity Navigation Simulator wit...Training embodied AI agents depends critically on the visual fidelity of simulation environments and the ability to model dynamic humans. Current simula...755InfiniteScienceGym: An Unbounded, Procedurally-Gener...Large language models are emerging as scientific assistants, but evaluating their ability to reason from empirical data remains challenging. Benchmarks...756KnowRL: Boosting LLM Reasoning via Reinforcement Lea...RLVR improves reasoning in large language models, but its effectiveness is often limited by severe reward sparsity on hard problems. Recent hint-based R...757KV Packet: Recomputation-Free Context-Independent KV...Large Language Models (LLMs) rely heavily on Key-Value (KV) caching to minimize inference latency. However, standard KV caches are context-dependent: re...758Learning Versatile Humanoid Manipulation with Touch...Humanoid robots promise general-purpose assistance, yet real-world humanoid loco-manipulation remains challenging because it requires whole-body stabili...759Lightning OPD: Efficient Post-Training for Large Rea...On-policy distillation (OPD) has emerged as an efficient post-training paradigm for large language models. However, standard OPD requires a live teacher...760Lyra 2.0: Explorable Generative 3D WorldsRecent advances in video generation enable a new paradigm for 3D scene creation: generating camera-controlled videos that simulate scene walkthroughs, t...761Masked by Consensus: Disentangling Privileged Knowle...Humans use introspection to evaluate their understanding through private internal states inaccessible to external observers. We investigate whether larg...762Mobile GUI Agents under Real-world Threats: Are We T...Recent years have witnessed a rapid development of mobile GUI agents powered by large language models (LLMs), which can autonomously execute diverse dev...763Nemotron 3 Super: Open, Efficient Mixture-of-Experts...We describe the pre-training, post-training, and quantization of Nemotron 3 Super, a 120 billion (active 12 billion) parameter hybrid Mamba-Attention Mi...764Parcae: Scaling Laws For Stable Looped Language ModelsTraditional fixed-depth architectures scale quality by increasing training FLOPs, typically through increased parameterization, at the expense of a high...765Rethinking On-Policy Distillation of Large Language...On-policy distillation (OPD) has become a core technique in the post-training of large language models, yet its training dynamics remain poorly understo...766ROSE: An Intent-Centered Evaluation Metric for NL2SQLExecution Accuracy (EX), the widely used metric for evaluating the effectiveness of Natural Language to SQL (NL2SQL) solutions, is becoming increasingly...767Self-Adversarial One Step Generation via Condition S...The push for efficient text to image synthesis has moved the field toward one step sampling, yet existing methods still face a three way tradeoff among...768SpotSound: Enhancing Large Audio-Language Models wit...Large Audio-Language Models (ALMs) have recently demonstrated remarkable capabilities in holistic audio understanding, yet they remain unreliable for te...769Toward Autonomous Long-Horizon Engineering for ML Re...Autonomous AI research has advanced rapidly, but long-horizon ML research engineering remains difficult: agents must sustain coherent progress across ta...770Towards Autonomous Mechanistic Reasoning in Virtual...Large language models (LLMs) have recently gained significant attention as a promising approach to accelerate scientific discovery. However, their appli...771Towards Long-horizon Agentic Multimodal SearchMultimodal deep search agents have shown great potential in solving complex tasks by iteratively collecting textual and visual evidence. However, managi...772VideoFlexTok: Flexible-Length Coarse-to-Fine Video T...Visual tokenizers map high-dimensional raw pixels into a compressed representation for downstream modeling. Beyond compression, tokenizers dictate what...773C2: Scalable Rubric-Augmented Reward Modeling from B...Rubric-augmented verification guides reward models with explicit evaluation criteria, yielding more reliable judgments than single-model verification. H...774Feed-Forward 3D Scene Modeling: A Problem-Driven Per...Reconstructing 3D representations from 2D inputs is a fundamental task in computer vision and graphics, serving as a cornerstone for understanding and i...775Free Geometry: Refining 3D Reconstruction from Longe...Feed-forward 3D reconstruction models are efficient but rigid: once trained, they perform inference in a zero-shot manner and cannot adapt to the test s...776From P(y|x) to P(y): Investigating Reinforcement Lea...While reinforcement learning with verifiable rewards (RLVR) significantly enhances LLM reasoning by optimizing the conditional distribution P(y|x), its...777Geometric Context Transformer for Streaming 3D Recon...Streaming 3D reconstruction aims to recover 3D information, such as camera poses and point clouds, from a video stream, which necessitates geometric acc...778GFT: From Imitation to Reward Fine-Tuning with Unbia...Large language models are typically post-trained using supervised fine-tuning (SFT) and reinforcement learning (RL), yet effectively unifying efficient...779HiVLA: A Visual-Grounded-Centric Hierarchical Embodi...While end-to-end Vision-Language-Action (VLA) models offer a promising paradigm for robotic manipulation, fine-tuning them on narrow control data often...780HY-World 2.0: A Multi-Modal World Model for Reconstr...We introduce HY-World 2.0, a multi-modal world model framework that advances our prior project HY-World 1.0. HY-World 2.0 accommodates diverse input mod...781LangFlow: Continuous Diffusion Rivals Discrete in La...Continuous diffusion has been the foundation of high-fidelity, controllable, and few-step generation of many data modalities such as images. However, in...782Memory Transfer Learning: How Memories are Transferr...Memory-based self-evolution has emerged as a promising paradigm for coding agents. However, existing approaches typically restrict memory utilization to...783MERRIN: A Benchmark for Multimodal Evidence Retrieva...Motivated by the underspecified, multi-hop nature of search queries and the multimodal, heterogeneous, and often conflicting nature of real-world web re...784OneHOI: Unifying Human-Object Interaction Generation...Human-Object Interaction (HOI) modelling captures how humans act upon and relate to objects, typically expressed as <person, action, object> triplets. E...785Reinforcement Learning via Value Gradient FlowWe study behavior-regularized reinforcement learning (RL), where regularization toward a reference distribution (the dataset in offline RL or the base m...786Reward Hacking in the Era of Large Models: Mechanism...Reinforcement Learning from Human Feedback (RLHF) and related alignment paradigms have become central to steering large language models (LLMs) and multi...787ROSE: Retrieval-Oriented Segmentation EnhancementExisting segmentation models based on multimodal large language models (MLLMs), such as LISA, often struggle with novel or emerging entities due to thei...788Seedance 2.0: Advancing Video Generation for World C...Seedance 2.0 is a new native multi-modal audio-video generation model, officially released in China in early February 2026. Compared with its predecesso...789SpatialEvo: Self-Evolving Spatial Intelligence via D...Spatial reasoning over three-dimensional scenes is a core capability for embodied intelligence, yet continuous model improvement remains bottlenecked by...790Three-Phase TransformerWe present Three-Phase Transformer (3PT), a residual-stream structural prior for decoder-only Transformers on a standard SwiGLU + RMSNorm + RoPE + GQA b...791TIP: Token Importance in On-Policy DistillationOn-policy knowledge distillation (OPD) trains a student on its own rollouts under token-level supervision from a teacher. Not all token positions matter...792TREX: Automating LLM Fine-tuning via Agent-Driven Tr...While Large Language Models (LLMs) have empowered AI research agents to perform isolated scientific tasks, automating complex, real-world workflows, suc...793UI-Copilot: Advancing Long-Horizon GUI Automation vi...MLLM-based GUI agents have demonstrated strong capabilities in complex user interface interaction tasks. However, long-horizon scenarios remain challeng...794UI-Zoomer: Uncertainty-Driven Adaptive Zoom-In for G...GUI grounding, which localizes interface elements from screenshots given natural language queries, remains challenging for small icons and dense layouts...795Amortized Optimal Transport from Sliced PotentialsWe propose a novel amortized optimization method for predicting optimal transport (OT) plans across multiple pairs of measures by leveraging Kantorovich...796An Optimal Transport-driven Approach for Cultivating...In online incremental learning, data continuously arrives with substantial distributional shifts, creating a significant challenge because previous samp...797Beyond Prompts: Unconditional 3D Inversion for Out-o...Text-driven inversion of generative models is a core paradigm for manipulating 2D or 3D content, unlocking numerous applications such as text-based edit...798Don't Retrieve, Navigate: Distilling Enterprise Know...Retrieval-Augmented Generation (RAG) grounds LLM responses in external evidence but treats the model as a passive consumer of search results: it never s...799DR^{3}-Eval: Towards Realistic and Reproducible Deep...Deep Research Agents (DRAs) aim to solve complex, long-horizon research tasks involving planning, retrieval, multimodal understanding, and report genera...800GlobalSplat: Efficient Feed-Forward 3D Gaussian Spla...The efficient spatial allocation of primitives serves as the foundation of 3D Gaussian Splatting, as it directly dictates the synergy between representa...801LeapAlign: Post-Training Flow Matching Models at Any...This paper focuses on the alignment of flow matching models with human preferences. A promising way is fine-tuning by directly backpropagating reward gr...802Learning Adaptive Reasoning Paths for Efficient Visu...Visual reasoning models (VRMs) have recently shown strong cross-modal reasoning capabilities by integrating visual perception with language reasoning. H...803LongAct: Harnessing Intrinsic Activation Patterns fo...Reinforcement Learning (RL) has emerged as a critical driver for enhancing the reasoning capabilities of Large Language Models (LLMs). While recent adva...804MinShap: A Modified Shapley Value Approach for Featu...Feature selection is a classical problem in statistics and machine learning, and it continues to remain an extremely challenging problem especially in t...805MM-WebAgent: A Hierarchical Multimodal Web Agent for...The rapid progress of Artificial Intelligence Generated Content (AIGC) tools enables images, videos, and visualizations to be created on demand for webp...806Model Capability Dominates: Inference-Time Optimizat...Majority voting over multiple LLM attempts improves mathematical reasoning, but correlated errors limit the effective sample size. A natural fix is to a...807One-Shot Generative Flows: Existence and ObstructionsWe study dynamic measure transport for generative modelling in the setting of a stochastic process $X_\bullet$ whose marginals interpolate between a sou...808OpenMobile: Building Open Mobile Agents with Task an...Mobile agents powered by vision-language models have demonstrated impressive capabilities in automating mobile tasks, with recent leading models achievi...809PRIM-cipal components analysisSupervised No Free Lunch Theorems (NFLTs) are well studied, yet unsupervised NFLTs remain underexplored. For elliptical distributions, we prove that the...810PRL-Bench: A Comprehensive Benchmark Evaluating LLMs...The paradigm of agentic science requires AI systems to conduct robust reasoning and engage in long-horizon, autonomous exploration. However, current sci...811Protecting Language Models Against Unauthorized Dist...Knowledge distillation is a widely adopted technique for transferring capabilities from LLMs to smaller, more efficient student models. However, unautho...812RAD-2: Scaling Reinforcement Learning in a Generator...High-level autonomous driving requires motion planners capable of modeling multimodal future uncertainties while remaining robust in closed-loop interac...813RadAgent: A tool-using AI agent for stepwise interpr...Vision-language models (VLM) have markedly advanced AI-driven interpretation and reporting of complex medical imaging, such as computed tomography (CT)....814Scaling Test-Time Compute for Agentic CodingTest-time scaling has become a powerful way to improve large language models. However, existing methods are best suited to short, bounded outputs that c...815Spurious Predictability in Financial Machine LearningAdaptive specification search generates statistically significant backtests even under martingale-difference nulls. We introduce a falsification audit t...816Structural interpretability in SVMs with truncated o...We study post-training interpretability for Support Vector Machines (SVMs) built from truncated orthogonal polynomial kernels. Since the associated repr...817Switch-KD: Visual-Switch Knowledge Distillation for...Vision-Language Models (VLMs) have shown remarkable capabilities in joint vision-language understanding, but their large scale poses significant challen...818Symbolic Guardrails for Domain-Specific Agents: Stro...AI agents that interact with their environments through tools enable powerful applications, but in high-stakes business settings, unintended actions can...819Theta-regularized Kriging: Modelling and AlgorithmsTo obtain more accurate model parameters and improve prediction accuracy, we proposed a regularized Kriging model that penalizes the hyperparameter thet...820TRACER: Trace-Based Adaptive Cost-Efficient Routing...Every call to an LLM classification endpoint produces a labeled input-output pair already retained in production logs. These pairs constitute a free, gr...821UniDoc-RL: Coarse-to-Fine Visual RAG with Hierarchic...Retrieval-Augmented Generation (RAG) extends Large Vision-Language Models (LVLMs) with external visual knowledge. However, existing visual RAG systems t...822WavAlign: Enhancing Intelligence and Expressiveness...End-to-end spoken dialogue models have garnered significant attention because they offer a higher potential ceiling in expressiveness and perceptual abi...823A Bayesian Updating Framework for Long-term Multi-En...In variety testing, multi-environment trials (MET) are essential for evaluating the genotypic performance of crop plants. A persistent challenge in the...824A Two-Stage, Object-Centric Deep Learning Framework...Academic integrity continues to face the persistent challenge of examination cheating. Traditional invigilation relies on human observation, which is in...825Adaptive multi-fidelity optimization with fast learn...In multi-fidelity optimization, biased approximations of varying costs of the target function are available. This paper studies the problem of optimizin...826AIFIND: Artifact-Aware Interpreting Fine-Grained Ali...As forgery types continue to emerge consistently, Incremental Face Forgery Detection (IFFD) has become a crucial paradigm. However, existing methods typ...827ArtifactNet: Detecting AI-Generated Music via Forens...We present ArtifactNet, a lightweight framework that detects AI-generated music by reframing the problem as forensic physics -- extracting and analyzing...828ASMR-Bench: Auditing for Sabotage in ML ResearchAs AI systems are increasingly used to conduct research autonomously, misaligned systems could introduce subtle flaws that produce misleading results wh...829AtManRL: Towards Faithful Reasoning via Differentiab...Large language models (LLMs) increasingly rely on chain-of-thought (CoT) reasoning to solve complex tasks. Yet ensuring that the reasoning trace both co...830BAGEL: Benchmarking Animal Knowledge Expertise in La...Large language models have shown strong performance on broad-domain knowledge and reasoning benchmarks, but it remains unclear how well language models...831Benign Fine-Tuning Breaks Safety Alignment in Audio...Prior work shows that fine-tuning aligned models on benign data degrades safety in text and vision modalities, and that proximity to harmful content in...832Beyond Distribution Sharpening: The Importance of Ta...Frontier models have demonstrated exceptional capabilities following the integration of task-reward-based reinforcement learning (RL) into their trainin...833Beyond Surface Statistics: Robust Conformal Predicti...Large language models are increasingly deployed in settings where reliability matters, yet output-level uncertainty signals such as token probabilities,...834Can LLMs Understand the Impact of Trauma? Costs and...Firearm violence is a pressing public health issue, yet research into survivors' lived experiences remains underfunded and difficult to scale. Qualitati...835C-GenReg: Training-Free 3D Point Cloud Registration...We introduce C-GenReg, a training-free framework for 3D point cloud registration that leverages the complementary strengths of world-scale generative pr...836Chain-of-Thought Degrades Visual Spatial Reasoning C...Multimodal Reasoning Models (MRMs) leveraging Chain-of-Thought (CoT) based thinking have revolutionized mathematical and logical problem-solving. Howeve...837Characterising LLM-Generated Competency Questions: a...Competency Questions (CQs) are a cornerstone of requirement elicitation in ontology engineering. CQs represent requirements as a set of natural language...838ChemGraph-XANES: An Agentic Framework for XANES Simu...Computational X-ray absorption near-edge structure (XANES) is widely used to probe local coordination environments, oxidation states, and electronic str...839Collective Kernel EFT for Pre-activation ResNetsIn finite-width deep neural networks, the empirical kernel $G$ evolves stochastically across layers. We develop a collective kernel effective field theo...840CollideNet: Hierarchical Multi-scale Video Represent...Time-to-Collision (TTC) forecasting is a critical task in collision prevention, requiring precise temporal prediction and comprehending both local and g...841Cut Your Losses! Learning to Prune Paths Early for E...Parallel reasoning enhances Large Reasoning Models (LRMs) but incurs prohibitive costs due to futile paths caused by early errors. To mitigate this, pat...842DENALI: A Dataset Enabling Non-Line-of-Sight Spatial...Consumer LiDARs in mobile devices and robots typically output a single depth value per pixel. Yet internally, they record full time-resolved histograms...843Dental Panoramic Radiograph Analysis Using YOLO26 Fr...Panoramic radiography is a fundamental diagnostic tool in dentistry, offering a comprehensive view of the entire dentition with minimal radiation exposu...844Detecting and Suppressing Reward Hacking with Gradie...Reinforcement learning with verifiable rewards (RLVR) typically optimizes for outcome rewards without imposing constraints on intermediate reasoning. Th...845Do Vision-Language Models Truly Perform Vision Reaso...Reasoning in vision-language models (VLMs) has recently attracted significant attention due to its broad applicability across diverse downstream tasks....846Enhancing AI and Dynamical Subseasonal Forecasts wit...Decision-makers rely on weather forecasts to plant crops, manage wildfires, allocate water and energy, and prepare for weather extremes. Today, such for...847Enhancing Hazy Wildlife Imagery: AnimalHaze3k and In...Atmospheric haze significantly degrades wildlife imagery, impeding computer vision applications critical for conservation, such as animal detection, tra...848Evaluating the Progression of Large Language Model C...Large Language Models (LLMs) have the potential to accelerate small molecule drug design due to their ability to reason about information from diverse s...849Find, Fix, Reason: Context Repair for Video ReasoningReinforcement learning has advanced video reasoning in large multi-modal models, yet dominant pipelines either rely on on-policy self-exploration, which...850FineCog-Nav: Integrating Fine-grained Cognitive Modu...UAV vision-language navigation (VLN) requires an agent to navigate complex 3D environments from an egocentric perspective while following ambiguous mult...851FL-MHSM: Spatially-adaptive Fusion and Ensemble Lear...Existing multi-hazard susceptibility mapping (MHSM) studies often rely on spatially uniform models, treat hazards independently, and provide limited rep...852From Benchmarking to Reasoning: A Dual-Aspect, Large...The complexity of Vietnam's legal texts presents a significant barrier to public access to justice. While Large Language Models offer a promising soluti...853GAViD: A Large-Scale Multimodal Dataset for Context-...Understanding affective dynamics in real-world social systems is fundamental to modeling and analyzing human-human interactions in complex environments....854Geometric coherence of single-cell CRISPR perturbati...Genome engineering has achieved remarkable sequence-level precision, yet predicting the transcriptomic state that a cell will occupy after perturbation...855Geometric regularization of autoencoders via observe...Stochastic dynamical systems with slow or metastable behavior evolve, on long time scales, on an unknown low-dimensional manifold in high-dimensional am...856GTA-2: Benchmarking General Tool Agents from Atomic...The development of general-purpose agents requires a shift from executing simple instructions to completing complex, real-world productivity workflows....857Hero-Mamba: Mamba-based Dual Domain Learning for Und...Underwater images often suffer from severe degradation, such as color distortion, low contrast, and blurred details, due to light absorption and scatter...858Information Router for Mitigating Modality Dominance...Vision Language models (VLMs) have demonstrated strong performance across a wide range of benchmarks, yet they often suffer from modality dominance, whe...859Joint-Centric Dual Contrastive Alignment with Struct...We propose HILBERT (HIerarchical Long-sequence Balanced Embedding with Reciprocal contrastive Training), a cross-attentive multimodal framework for lear...860JumpLoRA: Sparse Adapters for Continual Learning in...Adapter-based methods have become a cost-effective approach to continual learning (CL) for Large Language Models (LLMs), by sequentially learning a low-...861KWBench: Measuring Unprompted Problem Recognition in...We introduce the first version of KWBench (Knowledge Work Bench), a benchmark for unprompted problem recognition in large language models: can an LLM id...862Learning to Reason with Insight for Informal Theorem...Although most of the automated theorem-proving approaches depend on formal proof systems, informal theorem proving can align better with large language...863Mind's Eye: A Benchmark of Visual Abstraction, Trans...Multimodal large language models (MLLMs) have achieved impressive progress on vision language benchmarks, yet their capacity for visual cognitive and vi...864Modeling Sparse and Bursty Vulnerability Sightings:...Understanding and anticipating vulnerability-related activity is a major challenge in cyber threat intelligence. This work investigates whether vulnerab...865MTR-DuplexBench: Towards a Comprehensive Evaluation...Full-Duplex Speech Language Models (FD-SLMs) enable real-time, overlapping conversational interactions, offering a more dynamic user experience compared...866Neuro-Symbolic ODE Discovery with Latent Grammar FlowUnderstanding natural and engineered systems often relies on symbolic formulations, such as differential equations, which provide interpretability and t...867No Universal Courtesy: A Cross-Linguistic, Multi-Mod...This paper explores the response of Large Language Models (LLMs) to user prompts with different degrees of politeness and impoliteness. The Politeness T...868On the Rejection Criterion for Proxy-based Test-time...Recent works proposed test-time alignment methods that rely on a small aligned model as a proxy that guides the generation of a larger base (unaligned)...869On the Robustness of LLM-Based Dense Retrievers: A S...Decoder-only large language models (LLMs) are increasingly replacing BERT-style architectures as the backbone for dense retrieval, achieving substantial...870Optimizing Korean-Centric LLMs via Token PruningThis paper presents a systematic benchmark of state-of-the-art multilingual large language models (LLMs) adapted via token pruning - a compression techn...871OT on the Map: Quantifying Domain Shifts in Geograph...In computer vision and machine learning for geographic data, out-of-domain generalization is a pervasive challenge, arising from uneven global data cove...872Phase transitions in Doi-Onsager, Noisy Transformer,...We study phase transitions for repulsive-attractive mean-field free energies on the circle. For a $\frac{1}{n+1}$-periodic interaction whose Fourier coe...873Qwen3.5-Omni Technical ReportIn this work, we present Qwen3.5-Omni, the latest advancement in the Qwen-Omni model family. Representing a significant evolution over its predecessor,...874Repurposing 3D Generative Model for Autoregressive L...We introduce LaviGen, a framework that repurposes 3D generative models for 3D layout generation. Unlike previous methods that infer object layouts from...875Revisiting a Pain in the Neck: A Semantic Reasoning...We present SemanticQA, an evaluation suite designed to assess language models (LMs) in semantic phrase processing tasks. The benchmark consolidates exis...876Sample Complexity Bounds for Stochastic Shortest Pat...We study the sample complexity of learning an $ε$-optimal policy in the Stochastic Shortest Path (SSP) problem. We first derive sample complexity bounds...877Sentiment Analysis of German Sign Language Fairy TalesWe present a dataset and a model for sentiment analysis of German sign language (DGS) fairy tales. First, we perform sentiment analysis for three levels...878Sketching the Readout of Large Language Models for S...Data attribution and valuation are critical for understanding data-model synergy for Large Language Models (LLMs), yet existing gradient-based methods s...879Stargazer: A Scalable Model-Fitting Benchmark Enviro...The rise of autonomous AI agents suggests that dynamic benchmark environments with built-in feedback on scientifically grounded tasks are needed to eval...880Stylistic-STORM (ST-STORM) : Perceiving the Semantic...One of the dominant paradigms in self-supervised learning (SSL), illustrated by MoCo or DINO, aims to produce robust representations by capturing featur...881SwanNLP at SemEval-2026 Task 5: An LLM-based Framewo...Recent advances in language models have substantially improved Natural Language Understanding (NLU). Although widely used benchmarks suggest that Large...882Synthetic data in cryptocurrencies using generative...Data plays a fundamental role in consolidating markets, services, and products in the digital financial ecosystem. However, the use of real data, especi...883"Taking Stock at FAccT": Using Participatory Design...As a relatively new forum, ACM FAccT has become a key space for activists and scholars to critically examine emerging AI and ML technologies. It brings...884Target-Oriented Pretraining Data Selection via Neuro...Everyday tasks come with a target, and pretraining models around this target is what turns them into experts. In this paper, we study target-oriented la...885The Harder Path: Last Iterate Convergence for Uncoup...We study the problem of learning in zero-sum matrix games with repeated play and bandit feedback. Specifically, we focus on developing uncoupled algorit...886Using Large Language Models and Knowledge Graphs to...Explaining Machine Learning (ML) results in a transparent and user-friendly manner remains a challenging task of Explainable Artificial Intelligence (XA...887VEFX-Bench: A Holistic Benchmark for Generic Video E...As AI-assisted video creation becomes increasingly practical, instruction-guided video editing has become essential for refining generated or captured f...888VoxMind: An End-to-End Agentic Spoken Dialogue SystemRecent end-to-end spoken dialogue models enable natural interaction. However, as user demands become increasingly complex, models that rely solely on co...889Where Do Vision-Language Models Fail? World Scale An...Image geolocalization has traditionally been addressed through retrieval-based place recognition or geometry-based visual localization pipelines. Recent...890Abstain-R1: Calibrated Abstention and Post-Refusal C...Reinforcement fine-tuning improves the reasoning ability of large language models, but it can also encourage them to answer unanswerable queries by gues...891Beyond Text-Dominance: Understanding Modality Prefer...Native Omni-modal Large Language Models (OLLMs) have shifted from pipeline architectures to unified representation spaces. However, this native integrat...892Crowded in B-Space: Calibrating Shared Directions fo...Merging separately trained LoRA adapters is a practical alternative to joint multi-task training, but it often hurts performance. Existing methods usual...893EasyVideoR1: Easier RL for Video UnderstandingReinforcement learning from verifiable rewards (RLVR) has demonstrated remarkable effectiveness in improving the reasoning capabilities of large languag...894GenericAgent: A Token-Efficient Self-Evolving LLM Ag...Long-horizon large language model (LLM) agents are fundamentally limited by context. As interactions become longer, tool descriptions, retrieved memorie...895MNAFT: modality neuron-aware fine-tuning of multimod...Multimodal large language models (MLLMs) have shown impressive capabilities, yet they often struggle to effectively capture the fine-grained textual inf...896Test-Time Adaptation for EEG Foundation Models: A Sy...Electroencephalography (EEG) foundation models have shown strong potential for learning generalizable representations from large-scale neural data, yet...897The Cognitive Penalty: Ablating System 1 and System...Decentralized Autonomous Organizations (DAOs) are inclined explore Small Language Models (SLMs) as edge-native constitutional firewalls to vet proposals...898The Illusion of Certainty: Decoupling Capability and...On-policy distillation (OPD) is an increasingly important paradigm for post-training language models. However, we identify a pervasive Scaling Law of Mi...899Understanding and Enforcing Weight Disentanglement i...Task arithmetic provides an efficient, training-free way to edit pre-trained models, yet lacks a fundamental theoretical explanation for its success. Th...900Agents Explore but Agents Ignore: LLMs Lack Environm...LLM-based agents are assumed to integrate environmental observations into their reasoning: discovering highly relevant but unexpected information should...901Back to Repair: A Minimal Denoising Network\ for Tim...We introduce JuRe (Just Repair), a minimal denoising network for time series anomaly detection that exposes a central finding: architectural complexity...902Code-Switching Information Retrieval: Benchmarks, An...Code-switching is a pervasive linguistic phenomenon in global communication, yet modern information retrieval systems remain predominantly designed for,...903Coevolving Representations in Joint Image-Feature Di...Joint image-feature generative modeling has recently emerged as an effective strategy for improving diffusion training by coupling low-level VAE latents...904Diverse Dictionary LearningGiven only observational data X = g(Z), where both the latent variables Z and the generating process g are unknown, recovering Z is ill-posed without ad...905EvoMaster: A Foundational Agent Framework for Buildi...The convergence of large language models and agents is catalyzing a new era of scientific discovery: Agentic Science. While the scientific method is inh...906HSG: Hyperbolic Scene GraphScene graph representations enable structured visual understanding by modeling objects and their relationships, and have been widely used for multiview...907LLaTiSA: Towards Difficulty-Stratified Time Series R...Comprehensive understanding of time series remains a significant challenge for Large Language Models (LLMs). Current research is hindered by fragmented...908MoVE: Translating Laughter and Tears via Mixture of...Recent Speech-to-Speech Translation (S2ST) systems achieve strong semantic accuracy yet consistently strip away non-verbal vocalizations (NVs), such as...909Precise Debugging Benchmark: Is Your Model Debugging...Unlike code completion, debugging requires localizing faults and applying targeted edits. We observe that frontier LLMs often regenerate correct but ove...910SkillFlow:Benchmarking Lifelong Skill Discovery and...As the capability frontier of autonomous agents continues to expand, they are increasingly able to complete specialized tasks through plug-and-play exte...911Speculative Decoding for Autoregressive Video Genera...Autoregressive video diffusion is emerging as a promising paradigm for streaming video synthesis, with step distillation serving as the primary means of...912Terminal Wrench: A Dataset of 331 Reward-Hackable En...We release Terminal Wrench, a subset of 331 terminal-agent benchmark environments, copied from the popular open benchmarks that are demonstrably reward-...913The Continuity Layer: Why Intelligence Needs an Arch...The most important architectural problem in AI is not the size of the model but the absence of a layer that carries forward what the model has come to u...914UniMesh: Unifying 3D Mesh Understanding and GenerationRecent advances in 3D vision have led to specialized models for either 3D understanding (e.g., shape classification, segmentation, reconstruction) or 3D...915When Background Matters: Breaking Medical Vision Lan...Vision-Language Models (VLMs) are increasingly used in clinical diagnostics, yet their robustness to adversarial attacks remains largely unexplored, pos...916Agent-World: Scaling Real-World Environment Synthesi...Large language models are increasingly expected to serve as general-purpose agents that interact with external, stateful tool environments. The Model Co...917AI scientists produce results without reasoning scie...Large language model (LLM)-based systems are increasingly deployed to conduct scientific research autonomously, yet whether their reasoning adheres to t...918AJ-Bench: Benchmarking Agent-as-a-Judge for Environm...As reinforcement learning continues to scale the training of large language model-based agents, reliably verifying agent behaviors in complex environmen...919ClawEnvKit: Automatic Environment Generation for Cla...Constructing environments for training and evaluating claw-like agents remains a manual, human-intensive process that does not scale. We argue that what...920Contrastive Attribution in the Wild: An Interpretabi...Interpretability tools are increasingly used to analyze failures of Large Language Models (LLMs), yet prior work largely focuses on short prompts or toy...921Dual-View Training for Instruction-Following Informa...Instruction-following information retrieval (IF-IR) studies retrieval systems that must not only find documents relevant to a query, but also obey expli...922Extending One-Step Image Generation from Class Label...Few-step generation has been a long-standing goal, with recent one-step generation methods exemplified by MeanFlow achieving remarkable results. Existin...923Latent Preference Modeling for Cross-Session Persona...Users often omit essential details in their requests to LLM-based agents, resulting in under-specified inputs for tool use. This poses a fundamental cha...924LLM Safety From Within: Detecting Harmful Content wi...Guard models are widely used to detect harmful content in user prompts and LLM responses. However, state-of-the-art guard models rely solely on terminal...925MARCO: Navigating the Unseen Space of Semantic Corre...Recent advances in semantic correspondence rely on dual-encoder architectures, combining DINOv2 with diffusion backbones. While accurate, these billion-...926MathNet: a Global Multimodal Benchmark for Mathemati...Mathematical problem solving remains a challenging test of reasoning for large language and multimodal models, yet existing benchmarks are limited in si...927Mitigating Multimodal Hallucination via Phase-wise S...Large Vision-Language Models (LVLMs) still struggle with vision hallucination, where generated responses are inconsistent with the visual input. Existin...928MM-JudgeBias: A Benchmark for Evaluating Composition...Multimodal Large Language Models (MLLMs) have been increasingly used as automatic evaluators-a paradigm known as MLLM-as-a-Judge. However, their reliabi...929Modeling Multiple Support Strategies within a Single...Emotional Support Conversation (ESC) aims to assist individuals experiencing distress by generating empathetic and supportive dialogue. While prior work...930Multiplication in Multimodal LLMs: Computation with...Multimodal LLMs can accurately perceive numerical content across modalities yet fail to perform exact multi-digit multiplication when the identical unde...931MultiWorld: Scalable Multi-Agent Multi-View Video Wo...Video world models have achieved remarkable success in simulating environmental dynamics in response to actions by users or agents. They are modeled as...932On the Reliability of Computer Use AgentsComputer-use agents have rapidly improved on real-world tasks such as web navigation, desktop automation, and software interaction, in some cases surpas...933OneVL: One-Step Latent Reasoning and Planning with V...Chain-of-Thought (CoT) reasoning has become a powerful driver of trajectory prediction in VLA-based autonomous driving, yet its autoregressive nature im...934OpenGame: Open Agentic Coding for GamesGame development sits at the intersection of creative design and intricate software engineering, demanding the joint orchestration of game engines, real...935River-LLM: Large Language Model Seamless Exit Based...Large Language Models (LLMs) have demonstrated exceptional performance across diverse domains but are increasingly constrained by high inference latency...936Stratagem: Learning Transferable Reasoning via Traje...Games offer a compelling paradigm for developing general reasoning capabilities in language models, as they naturally demand strategic planning, probabi...937Streaming Structured Inference with Flash-SemiCRFSemi-Markov Conditional Random Fields (semi-CRFs) assign labels to segments of a sequence rather than to individual positions, enabling exact inference...938The Geometric Canary: Predicting Steerability and De...Reliable deployment of language models requires two capabilities that appear distinct but share a common geometric foundation: predicting whether a mode...939Training LLM Agents for Spontaneous, Reward-Free Sel...Most agents today ``self-evolve'' by following rewards and rules defined by humans. However, this process remains fundamentally dependent on external su...940UDM-GRPO: Stable and Efficient Group Relative Policy...Uniform Discrete Diffusion Model (UDM) has recently emerged as a promising paradigm for discrete generative modeling; however, its integration with rein...941WebCompass: Towards Multimodal Web Coding Evaluation...Large language models are rapidly evolving into interactive coding agents capable of end-to-end web coding, yet existing benchmarks evaluate only narrow...942When Can LLMs Learn to Reason with Weak Supervision?Large language models have achieved significant reasoning improvements through reinforcement learning with verifiable rewards (RLVR). Yet as model capab...943A Self-Evolving Framework for Efficient Terminal Age...As model capabilities advance, research has increasingly shifted toward long-horizon, multi-turn terminal-centric agentic tasks, where raw environment f...944Accurate and scalable exchange-correlation with deep...Density Functional Theory (DFT) underpins much of modern computational chemistry and materials science. Yet, the reliability of DFT-derived predictions...945AnyRecon: Arbitrary-View 3D Reconstruction with Vide...Sparse-view 3D reconstruction is essential for modeling scenes from casual captures, but remain challenging for non-generative reconstruction. Existing...946Chat2Workflow: A Benchmark for Generating Executable...At present, executable visual workflows have emerged as a mainstream paradigm in real-world industrial deployments, offering strong reliability and cont...947CityRAG: Stepping Into a City via Spatially-Grounded...We address the problem of generating a 3D-consistent, navigable environment that is spatially grounded: a simulation of a real location. Existing video...948ClawNet: Human-Symbiotic Agent Network for Cross-Use...Current AI agent frameworks have made remarkable progress in automating individual tasks, yet all existing systems serve a single user. Human productivi...949CoInteract: Physically-Consistent Human-Object Inter...Synthesizing human--object interaction (HOI) videos has broad practical value in e-commerce, digital advertising, and virtual marketing. However, curren...950CreativeGame:Toward Mechanic-Aware Creative Game Gen...Large language models can generate plausible game code, but turning this capability into iterative creative improvement remains difficult. In practice,...951DR-Venus: Towards Frontier Edge-Scale Deep Research...Edge-scale deep research agents based on small language models are attractive for real-world deployment due to their advantages in cost, latency, and pr...952Evaluation-driven Scaling for Scientific DiscoveryLanguage models are increasingly used in scientific discovery to generate hypotheses, propose candidate solutions, implement systems, and iteratively re...953Expert Upcycling: Shifting the Compute-Efficient Fro...Mixture-of-Experts (MoE) has become the dominant architecture for scaling large language models: frontier models routinely decouple total parameters fro...954HP-Edit: A Human-Preference Post-Training Framework...Common image editing tasks typically adopt powerful generative diffusion models as the leading paradigm for real-world content editing. Meanwhile, altho...955LoopCTR: Unlocking the Loop Scaling Power for Click-...Scaling Transformer-based click-through rate (CTR) models by stacking more parameters brings growing computational and storage overhead, creating a wide...956Micro Language Models Enable Instant ResponsesEdge devices such as smartwatches and smart glasses cannot continuously run even the smallest 100M-1B parameter language models due to power and compute...957MMCORE: MultiModal COnnection with Representation Al...We present MMCORE, a unified framework designed for multimodal image generation and editing. MMCORE leverages a pre-trained Vision-Language Model (VLM)...958PlayCoder: Making LLM-Generated GUI Code PlayableLarge language models (LLMs) have achieved strong results in code generation, but their ability to generate GUI applications, especially games, remains...959RDP LoRA: Geometry-Driven Identification for Paramet...Fine-tuning Large Language Models (LLMs) remains structurally uncertain despite parameter-efficient methods such as Low-Rank Adaptation (LoRA), as the l...960ReImagine: Rethinking Controllable High-Quality Huma...Human video generation remains challenging due to the difficulty of jointly modeling human appearance, motion, and camera viewpoint under limited multi-...961SAVOIR: Learning Social Savoir-Faire via Shapley-bas...Social intelligence, the ability to navigate complex interpersonal interactions, presents a fundamental challenge for language agents. Training such age...962Sessa: Selective State Space AttentionModern sequence modeling is dominated by two families: Transformers, whose self-attention can access arbitrary elements of the visible sequence, and str...963ShadowPEFT: Shadow Network for Parameter-Efficient F...Parameter-efficient fine-tuning (PEFT) reduces the training cost of full-parameter fine-tuning for large language models (LLMs) by training only a small...964SmartPhotoCrafter: Unified Reasoning, Generation and...Traditional photographic image editing typically requires users to possess sufficient aesthetic understanding to provide appropriate instructions for ad...965Tadabur: A Large-Scale Quran Audio DatasetDespite growing interest in Quranic data research, existing Quran datasets remain limited in both scale and diversity. To address this gap, we present T...966TEMPO: Scaling Test-time Training for Large Reasonin...Test-time training (TTT) adapts model parameters on unlabeled test instances during inference time, which continuously extends capabilities beyond the r...967Tstars-Tryon 1.0: Robust and Realistic Virtual Try-O...Recent advances in image generation and editing have opened new opportunities for virtual try-on. However, existing methods still struggle to meet compl...968UniT: Toward a Unified Physical Language for Human-t...Scaling humanoid foundation models is bottlenecked by the scarcity of robotic data. While massive egocentric human data offers a scalable alternative, b...969Visual Reasoning through Tool-supervised Reinforceme...In this paper, we investigate the problem of how to effectively master tool-use to solve complex visual reasoning tasks for Multimodal Large Language Mo...970What Makes an LLM a Good Optimizer? A Trajectory Ana...Recent work has demonstrated the promise of orchestrating large language models (LLMs) within evolutionary and agentic optimization systems. However, th...971Chasing the Public Score: User Pressure and Evaluati...Frontier coding agents are increasingly used in workflows where users supervise progress primarily through repeated improvement of a public score, namel...972Co-Evolving LLM Decision and Skill Bank Agents for L...Long horizon interactive environments are a testbed for evaluating agents skill usage abilities. These environments demand multi step reasoning, the cha...973COMPASS: COntinual Multilingual PEFT with Adaptive S...Large language models (LLMs) often exhibit performance disparities across languages, with naive multilingual fine-tuning frequently degrading performanc...974Convergent Evolution: How Different Language Models...Language models trained on natural text learn to represent numbers using periodic features with dominant periods at T=2, 5, 10. In this paper, we identi...975Cortex 2.0: Grounding World Models in Real-World Ind...Industrial robotic manipulation demands reliable long-horizon execution across embodiments, tasks, and changing object distributions. While Vision-Langu...976DeVI: Physics-based Dexterous Human-Object Interacti...Recent advances in video generative models enable the synthesis of realistic human-object interaction videos across a wide range of scenarios and object...977Exploring Spatial Intelligence from a Generative Per...Spatial intelligence is essential for multimodal large language models, yet current benchmarks largely assess it only from an understanding perspective....978Hybrid Policy Distillation for LLMsKnowledge distillation (KD) is a powerful paradigm for compressing large language models (LLMs), whose effectiveness depends on intertwined choices of d...979Image Generators are Generalist Vision LearnersRecent works show that image and video generators exhibit zero-shot visual understanding behaviors, in a way reminiscent of how LLMs develop emergent ca...980LLaDA2.0-Uni: Unifying Multimodal Understanding and...We present LLaDA2.0-Uni, a unified discrete diffusion large language model (dLLM) that supports multimodal understanding and generation within a nativel...981MedSkillAudit: A Domain-Specific Audit Framework for...Background: Agent skills are increasingly deployed as modular, reusable capability units in AI agent systems. Medical research agent skills require safe...982Near-Future Policy OptimizationReinforcement learning with verifiable rewards (RLVR) has become a core post-training recipe. Introducing suitable off-policy trajectories into on-polic...983SkillLearnBench: Benchmarking Continual Learning Met...Skills have become the de facto way to enable LLM agents to perform complex real-world tasks with customized instructions, workflows, and tools, but how...984SWE-chat: Coding Agent Interactions From Real Users...AI coding agents are being adopted at scale, yet we lack empirical evidence on how people actually use them and how much of their output is useful in pr...985Temporally Extended Mixture-of-Experts ModelsMixture-of-Experts models, now popular for scaling capacity at fixed inference speed, switch experts at nearly every token. Once a model outgrows availa...986WebGen-R1: Incentivizing Large Language Models to Ge...While Large Language Models (LLMs) excel at function-level code generation, project-level tasks such as generating functional and visually aesthetic mul...987Context Unrolling in Omni ModelsWe present Omni, a unified multimodal model natively trained on diverse modalities, including text, images, videos, 3D geometry, and hidden representati...988DiffNR: Diffusion-Enhanced Neural Representation Opt...Neural representations (NRs), such as neural fields and 3D Gaussians, effectively model volumetric data in computed tomography (CT) but suffer from seve...989Encoder-Free Human Motion Understanding via Structur...The world knowledge and reasoning capabilities of text-based large language models (LLMs) are advancing rapidly, yet current approaches to human motion...990Explainable Disentangled Representation Learning for...Learning robust representations of authorial style is crucial for authorship attribution and AI-generated text detection. However, existing methods ofte...991Probing Visual Planning in Image Editing ModelsVisual planning represents a crucial facet of human intelligence, especially in tasks that require complex spatial reasoning and navigation. Yet, in mac...992Seeing Fast and Slow: Learning the Flow of Time in V...How can we tell whether a video has been sped up or slowed down? How can we generate videos at different speeds? Although videos have been central to mo...993StyleID: A Perception-Aware Dataset and Metric for S...Creative face stylization aims to render portraits in diverse visual idioms such as cartoons, sketches, and paintings while retaining recognizable ident...994TingIS: Real-time Risk Event Discovery from Noisy Cu...Real-time detection and mitigation of technical anomalies are critical for large-scale cloud-native services, where even minutes of downtime can result...995Trust but Verify: Introducing DAVinCI -- A Framework...Large Language Models (LLMs) have demonstrated remarkable fluency and versatility across a wide range of NLP tasks, yet they remain prone to factual ina...996UniGenDet: A Unified Generative-Discriminative Frame...In recent years, significant progress has been made in both image generation and generated image detection. Despite their rapid, yet largely independent...997Vista4D: Video Reshooting with 4D Point CloudsWe present Vista4D, a robust and flexible video reshooting framework that grounds the input video and target cameras in a 4D point cloud. Specifically,...998VLAA-GUI: Knowing When to Stop, Recover, and Search,...Autonomous GUI agents face two fundamental challenges: early stopping, where agents prematurely declare success without verifiable evidence, and repetit...999WorldMark: A Unified Benchmark Suite for Interactive...Interactive video generation models such as Genie, YUME, HY-World, and Matrix-Game are advancing rapidly, yet every model is evaluated on its own benchm...1000Agentic World Modeling: Foundations, Capabilities, L...As AI systems move from generating text to accomplishing goals through sustained interaction, the ability to model environment dynamics becomes a centra...1001AgentSearchBench: A Benchmark for AI Agent Search in...The rapid growth of AI agent ecosystems is transforming how complex tasks are delegated and executed, creating a new challenge of identifying suitable a...1002dWorldEval: Scalable Robotic Policy Evaluation via D...Evaluating robotics policies across thousands of environments and thousands of tasks is infeasible with existing approaches. This motivates the need for...1003FlowAnchor: Stabilizing the Editing Signal for Inver...We propose FlowAnchor, a training-free framework for stable and efficient inversion-free, flow-based video editing. Inversion-free editing methods have...1004Learning Evidence Highlighting for Frozen LLMsLarge Language Models (LLMs) can reason well, yet often miss decisive evidence when it is buried in long, noisy contexts. We introduce HiLight, an Evide...1005Enhanced Privacy and Communication Efficiency in Non...Federated learning (FL) is a distributed machine learning method where multiple devices collaboratively train a model under the management of a central...1006OceanPile: A Large-Scale Multimodal Ocean Corpus for...The vast and underexplored ocean plays a critical role in regulating global climate and supporting marine biodiversity, yet artificial intelligence has...1007Diffusion Model as a Generalist Segmentation LearnerDiffusion models are primarily trained for image synthesis, yet their denoising trajectories encode rich, spatially aligned visual priors. In this paper...1008Large Language Models Explore by Latent DistillingGenerating diverse responses is crucial for test-time scaling of large language models (LLMs), yet standard stochastic sampling mostly yields surface-le...1009Meta-CoT: Enhancing Granularity and Generalization i...Unified multi-modal understanding/generative models have shown improved image editing performance by incorporating fine-grained understanding into their...1010Programming with Data: Test-Driven Data Engineering...Reliably transferring specialized human knowledge from text into large language models remains a fundamental challenge in artificial intelligence. Fine-...1011X2SAM: Any Segmentation in Images and VideosMultimodal Large Language Models (MLLMs) have demonstrated strong image-level visual understanding and reasoning, yet their pixel-level perception acros...1012AutoResearchBench: Benchmarking AI Agents on Complex...Autonomous scientific research is significantly advanced thanks to the development of AI agents. One key step in this process is finding the right scien...1013DV-World: Benchmarking Data Visualization Agents in...Real-world data visualization (DV) requires native environmental grounding, cross-platform evolution, and proactive intent alignment. Yet, existing benc...1014How Fast Should a Model Commit to Supervision? Train...Adapting reasoning models to new tasks during post-training with only output-level supervision stalls under reinforcement learning from verifiable rewar...1015Praxy Voice: Voice-Prompt Recovery + BUPS for Commer...Commercial TTS systems produce near-native Indic audio, but the best open-source bases (Chatterbox, Indic Parler-TTS, IndicF5) trail them on measured ph...1016Prior-Aligned Data Cleaning for Tabular Foundation M...Tabular Foundation Models (TFMs) achieve state-of-the-art zero-shot accuracy on small tabular datasets by meta-learning over synthetic data-generating p...1017PSP: An Interpretable Per-Dimension Accent Benchmark...Standard text-to-speech (TTS) evaluation measures intelligibility (WER, CER) and overall naturalness (MOS, UTMOS) but does not quantify accent. A synthe...1018RADIO-ViPE: Online Tightly Coupled Multi-Modal Fusio...We present RADIO-ViPE (Reduce All Domains Into One -- Video Pose Engine), an online semantic SLAM system that enables geometry-aware open-vocabulary gro...1019Recursive Multi-Agent SystemsRecursive or looped language models have recently emerged as a new scaling axis by iteratively refining the same model computation over latent states to...1020Refinement via Regeneration: Enlarging Modification...Unified multimodal models (UMMs) integrate visual understanding and generation within a single framework. For text-to-image (T2I) tasks, this unified ca...1021A Note on How to Remove the $\ln\ln T$ Term from the...In Orabona and Pál [2016], we introduced the shifted KT potentials, to remove the $\ln \ln T$ factor in the parameter-free learning with expert bound. I...1022ComboStoc: Combinatorial Stochasticity for Diffusion...In this paper, we study an under-explored but important factor of diffusion generative models, i.e., the combinatorial complexity. Data samples are gene...1023Compliance versus Sensibility: On the Reasoning Cont...Large Language Models (LLMs) are known to acquire reasoning capabilities through shared inference patterns in pre-training data, which are further elici...1024Efficient Training on Multiple Consumer GPUs with Ro...Fine-tuning Large Language Models (LLMs) on consumer-grade GPUs is highly cost-effective, yet constrained by limited GPU memory and slow PCIe interconne...1025GLM-5V-Turbo: Toward a Native Foundation Model for M...We present GLM-5V-Turbo, a step toward native foundation models for multimodal agents. As foundation models are increasingly deployed in real environmen...1026HiL-Bench (Human-in-Loop Benchmark): Do Agents Know...Frontier coding agents solve complex tasks when given complete context but collapse when specifications are incomplete or ambiguous. The bottleneck is n...1027Hyper Input Convex Neural Networks for Shape Constra...We introduce Hyper Input Convex Neural Networks (HyCNNs), a novel neural network architecture designed for learning convex functions. HyCNNs combine the...1028Learning Rate Transfer in Normalized TransformersThe Normalized Transformer, or nGPT (arXiv:2410.01131) achieves impressive training speedups and does not require weight decay or learning rate warmup....1029Length Value Model: Scalable Value Pretraining for T...Token serves as the fundamental unit of computation in modern autoregressive models, and generation length directly influences both inference cost and r...1030Linear Models, Variable Selection, Artificial Intell...Variable selection in linear regression models has been a problem since hypothesis testing began. Which variables to include or exclude from a model is...1031Step-level Optimization for Efficient Computer-use A...Computer-use agents provide a promising path toward general software automation because they can interact directly with arbitrary graphical user interfa...1032The Last Human-Written Paper: Agent-Native Research...Scientific publication compresses a branching, iterative research process into a linear narrative, discarding the majority of what was discovered along...1033Turning the TIDE: Cross-Architecture Distillation fo...Diffusion large language models (dLLMs) offer parallel decoding and bidirectional context, but state-of-the-art dLLMs require billions of parameters for...10343D-ReGen: A Unified 3D Geometry Regeneration FrameworkWe consider the problem of regenerating 3D objects from 2D images and initial 3D shapes. Most 3D generators operate in a one-shot fashion, converting te...1035A Dirac-Frenkel-Onsager principle: Instantaneous res...Dirac-Frenkel instantaneous residual minimization evolves nonlinear parametrizations of PDE solutions in time, but ill-conditioning can render the param...1036A Novel Computational Framework for Causal Inference...Causal inference is essential for data-driven decision-making, as it aims to uncover causal relationships from observational data. However, identifying...1037A unified perspective on fine-tuning and sampling wi...We study the problem of training diffusion and flow generative models to sample from target distributions defined by an exponential tilting of a base de...1038Action Motifs: Self-Supervised Hierarchical Represen...Effective human behavior modeling requires a representation of the human body movement that capitalizes on its compositionality. We propose a hierarchic...1039AEGIS: A Holistic Benchmark for Evaluating Forensic...We introduce AEGIS, A holistic benchmark for Evaluating forensic analysis of AI-Generated academic ImageS. Compared to existing benchmarks, AEGIS featur...1040An adaptive wavelet-based PINN for problems with loc...In recent years, physics-informed neural networks (PINNs) have gained significant attention for solving differential equations, although they suffer fro...1041Assessing Pancreatic Ductal Adenocarcinoma Vascular...Surgical resection remains the only potentially curative treatment for pancreatic ductal adenocarcinoma (PDAC), and eligibility depends on accurate asse...1042Bayesian X-Learner: Calibrated Posterior Inference f...Conditional Average Treatment Effect (CATE) estimation in practice demands three properties simultaneously: heterogeneous effects $τ(x)$, calibrated unc...1043Beyond Gaussian Bottlenecks: Topologically Aligned E...Modern visual world modeling systems increasingly rely on high-capacity architectures and large-scale data to produce plausible motion, yet they often f...1044Beyond Pixel Fidelity: Minimizing Perceptual Distort...Night Photography Rendering (NPR) poses a significant challenge due to the extreme contrast between dark and illuminated areas in scenes, stemming from...1045Claw-Eval-Live: A Live Agent Benchmark for Evolving...LLM agents are expected to complete end-to-end units of work across software tools, business services, and local workspaces. Yet many agent benchmarks f...1046Computing Equilibrium beyond Unilateral DeviationMost familiar equilibrium concepts, such as Nash and correlated equilibrium, guarantee only that no single player can improve their utility by deviating...1047Continuous-tone Simple Points: An $\ell_0$-Norm of C...Topological features play an essential role in ensuring geometric plausibility and structural consistency in image analysis tasks such as segmentation a...1048Crab: A Semantics-Aware Checkpoint/Restore Runtime f...Autonomous agents act through sandboxed containers and microVMs whose state spans filesystems, processes, and runtime artifacts. Checkpoint and restore...1049Decoupled Descent: Exact Test Error Tracking Via App...In modern parametric model training, full-batch gradient descent (and its variants) suffers due to progressively stronger biasing towards the exact real...1050DEFault++: Automated Fault Detection, Categorization...Transformer models are widely deployed in critical AI applications, yet faults in their attention mechanisms, projections, and other internal components...1051Defending Quantum Classifiers against Adversarial Pe...Machine learning models can learn from data samples to carry out various tasks efficiently. When data samples are adversarially manipulated, such as by...1052Do Sparse Autoencoders Capture Concept Manifolds?Sparse autoencoders (SAEs) are widely used to extract interpretable features from neural network representations, often under the implicit assumption th...1053Ease of dependency distance minimization in star-lik...The syntactic structure of a sentence can be represented as a tree where edges indicate syntactic dependencies between words. When that structure is a s...1054EDU-CIRCUIT-HW: Evaluating Multimodal Large Language...Multimodal Large Language Models (MLLMs) hold significant promise for revolutionizing traditional education and reducing teachers' workload. However, ac...1055Efficient Multivector Retrieval with Token-Aware Clu...Multivector retrieval models achieve state-of-the-art effectiveness through fine-grained token-level representations, but their deployment incurs substa...1056Explainable Load Forecasting with Covariate-Informed...Time Series Foundation Models (TSFMs) have recently emerged as general-purpose forecasting models and show considerable potential for applications in en...1057Exploration Hacking: Can LLMs Learn to Resist RL Tra...Reinforcement learning (RL) has become essential to the post-training of large language models (LLMs) for reasoning, agentic capabilities and alignment....1058FlashRT: Towards Computationally and Memory Efficien...Long-context large language models (LLMs)-for example, Gemini-3.1-Pro and Qwen-3.5-are widely used to empower many real-world applications, such as retr...1059FlexiTac: A Low-Cost, Open-Source, Scalable Tactile...We present FlexiTac, a low-cost, open-source, and scalable piezoresistive tactile sensing solution designed for robotic end-effectors. FlexiTac is a pra...1060FoReco and FoRecoML: A Unified Toolbox for Forecast...Forecast reconciliation has become key to improving the accuracy and coherence of forecasts for linearly constrained multiple time series, such as hiera...1061Generalizable Sparse-View 3D Reconstruction from Unc...Reconstructing 3D scenes from sparse, unposed images remains challenging under real-world conditions with varying illumination and transient occlusions....1062Global Optimality for Constrained Exploration via Pe...Efficient exploration is a central problem in reinforcement learning and is often formalized as maximizing the entropy of the state-action occupancy mea...1063HERMES++: Toward a Unified Driving World Model for 3...Driving world models serve as a pivotal technology for autonomous driving by simulating environmental dynamics. However, existing approaches predominant...1064Heterogeneous Scientific Foundation Model CollaborationAgentic large language model systems have demonstrated strong capabilities. However, their reliance on language as the universal interface fundamentally...1065Information-geometric adaptive sampling for graph di...Standard diffusion models for graph generation typically rely on uniform time-stepping, an approach that overlooks the non-homogeneous dynamics of distr...1066Instruction-Guided Poetry Generation in Arabic and I...Poetry has long been a central art form for Arabic speakers, serving as a powerful medium of expression and cultural identity. While modern Arabic speak...1067InteractWeb-Bench: Can Multimodal Agent Escape Blind...With the advancement of multimodal large language models (MLLMs) and coding agents, the website development has shifted from manual programming to agent...1068Intern-Atlas: A Methodological Evolution Graph as Re...Existing research infrastructure is fundamentally document-centric, providing citation links between papers but lacking explicit representations of meth...1069Kernelized Advantage Estimation: From Nonparametric...Recent advances in large language models (LLMs) have increasingly relied on reinforcement learning (RL) to improve their reasoning capabilities. Three a...1070LaST-R1: Reinforcing Action via Adaptive Physical La...Vision-Language-Action (VLA) models have increasingly incorporated reasoning mechanisms for complex robotic manipulation. However, existing approaches s...1071Latent Adversarial Detection: Adaptive Probing of LL...Multi-turn prompt injection follows a known attack path -- trust-building, pivoting, escalation but text-level defenses miss covert attacks where indivi...1072Latent-GRPO: Group Relative Policy Optimization for...Latent reasoning offers a more efficient alternative to explicit reasoning by compressing intermediate reasoning into continuous representations and sub...1073Linear-Core Surrogates: Smooth Loss Functions with L...The choice of loss function in classification involves a fundamental trade-off: smooth losses (like Cross-Entropy) enable fast optimization rates but yi...1074Linking spatial biology and clinical histology via H...Integrating molecular, morphological, and clinical data is essential for basic and translational biomedical research, yet systematic frameworks for join...1075LLM as Clinical Graph Structure Refiner: Enhancing R...Electroencephalogram (EEG) signals are vital for automated seizure detection, but their inherent noise makes robust representation learning challenging....1076Mapping the Methodological Space of Classroom Intera...Research on classroom interaction has long been divided between large-scale observation and in-depth ethnographic work. We propose a framework mapping t...1077Mapping the Phase Diagram of the Vicsek Model with M...In this study, we use machine learning to classify and interpolate the phase structure of the Vicsek flocking model across the three-dimensional paramet...1078Measuring research data reuse in scholarly publicati...Numerous metascience studies and other initiatives have begun to monitor the prevalence of open science practices when it is more important to understan...1079Mind the Gap: Structure-Aware Consistency in Prefere...Preference learning has become the foundation of aligning Large Language Models (LLMs) with human intent. Popular methods, such as Direct Preference Opt...1080MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Mo...Recent progress in multimodal large language models (MLLMs) has brought AI capabilities from static offline data processing to real-time streaming inter...1081MM-StanceDet: Retrieval-Augmented Multi-modal Multi-...Multimodal Stance Detection (MSD) is crucial for understanding public discourse, yet effectively fusing text and image, especially with conflicting sign...1082MoCapAnything V2: End-to-End Motion Capture for Arbi...Recent methods for arbitrary-skeleton motion capture from monocular video follow a factorized pipeline, where a Video-to-Pose network predicts joint pos...1083Models Recall What They Violate: Constraint Adherenc...When researchers iteratively refine ideas with large language models, do the models preserve fidelity to the original objective? We introduce DriftBench...1084Normativity and Productivism: Ableist Intelligence?...Sign languages, of any geographical or accentual variation, understandably face continuous scrutiny under the ever present popularity of verbal dictatio...1085OmniRobotHome: A Multi-Camera Platform for Real-Time...Human-robot collaboration has been studied primarily in dyadic or sequential settings. However, real homes require multiadic collaboration, where multip...1086On the Proper Treatment of Units in Surprisal TheorySurprisal theory links human processing effort to the predictability of an upcoming linguistic unit, but empirical work often leaves the notion of a uni...1087Optimized Deferral for Imbalanced SettingsLearning algorithms can be significantly improved by routing complex or uncertain inputs to specialized experts, balancing accuracy with computational c...1088PhyCo: Learning Controllable Physical Priors for Gen...Modern video diffusion models excel at appearance synthesis but still struggle with physical consistency: objects drift, collisions lack realistic rebou...1089Prediction-powered Inference by Mixture of ExpertsThe rapidly expanding artificial intelligence (AI) industry has produced diverse yet powerful prediction tools, each with its own network architecture,...1090PRISM: Pre-alignment via Black-box On-policy Distill...The standard post-training recipe for large multimodal models (LMMs) applies supervised fine-tuning (SFT) on curated demonstrations followed by reinforc...1091Reliable Answers for Recurring Questions: Boosting T...Large language models (LLMs) have revolutionized Text-to-SQL generation, allowing users to query structured data using natural language with growing eas...1092Repetition over Diversity: High-Signal Data Filterin...Recent research has shown that filtering massive English web corpora into high-quality subsets significantly improves training efficiency. However, for...1093Representation Fréchet Loss for Visual GenerationWe show that Fréchet Distance (FD), long considered impractical as a training objective, can in fact be effectively optimized in the representation spac...1094RouteProfile: Elucidating the Design Space of LLM Pr...As the large language model (LLM) ecosystem expands, individual models exhibit varying capabilities across queries, benchmarks, and domains, motivating...1095Sequential Inference for Gaussian Processes: A Signa...The proliferation of capable and efficient machine learning (ML) models marks one of the strongest methodological shifts in signal processing (SP) in it...1096Skills-Coach: A Self-Evolving Skill Optimizer via Tr...We introduce Skills-Coach, a novel automated framework designed to significantly enhance the self-evolution of skills within Large Language Model (LLM)-...1097Splitting Argumentation Frameworks with Collective A...This work proposes novel splitting techniques for argumentation formalisms that incorporate supports between defeasible elements. We base our studies on...1098Stable Behavior, Limited Variation: Persona Validity...Large Language Models (LLMs) are increasingly used as proxies for human perception in urban analysis, yet it remains unclear whether persona prompting p...1099Stop Holding Your Breath: CT-Informed Gaussian Splat...Bronchoscopic navigation relies on registering endoscopic video to a preoperative CT scan, but respiratory motion deforms the airway by 5-20 mm, creatin...1100Strait: Perceiving Priority and Interference in ML I...Machine learning (ML) inference serving systems host deep neural network (DNN) models and schedule incoming inference requests across deployed GPUs. How...1101Synthetic Computers at Scale for Long-Horizon Produc...Realistic long-horizon productivity work is strongly conditioned on user-specific computer environments, where much of the work context is stored and or...1102The Bernstein-von Mises theorem for Bayesian one-pas...Bayesian online learning provides a coherent framework for sequential inference. However, its theoretical understanding remains limited, particularly in...1103TopBench: A Benchmark for Implicit Prediction and Re...Large Language Models (LLMs) have advanced Table Question Answering, where most queries can be answered by extracting information or simple aggregation....1104Universal statistical laws governing culinary designCooking is a cultural expression of human creativity that transcends geography and time through the orchestration of ingredients and techniques, much li...1105Visual Generation in the New Era: An Evolution from...Recent visual generation models have made major progress in photorealism, typography, instruction following, and interactive editing, yet they still str...1106WindowsWorld: A Process-Centric Benchmark of Autonom...While GUI agents have shown impressive capabilities in common computer-use tasks such as OSWorld, current benchmarks mainly focus on isolated and single...1107World2Minecraft: Occupancy-Driven Simulated Scenes C...Embodied intelligence requires high-fidelity simulation environments to support perception and decision-making, yet existing platforms often suffer from...1108Adaptive Querying with AI Persona PriorsWe study adaptive querying for learning user-dependent quantities of interest, such as responses to held-out items and psychometric indicators, within t...1109Augmented Lagrangian Multiplier Network for State-wi...Safety is a primary challenge in real-world reinforcement learning (RL). Formulating safety requirements as state-wise constraints has become a prominen...1110Batch Normalization for Neural Networks on Complex D...Riemannian neural networks have proven effective in solving a variety of machine learning tasks. The key to their success lies in the development of pri...1111Beyond Benchmarks: MathArena as an Evaluation Platfo...Large language models (LLMs) are becoming increasingly capable mathematical collaborators, but static benchmarks are no longer sufficient for evaluating...1112Beyond SFT-to-RL: Pre-alignment via Black-Box On-Pol...The standard post-training recipe for large multimodal models (LMMs) applies supervised fine-tuning (SFT) on curated demonstrations followed by reinforc...1113BlenderRAG: High-Fidelity 3D Object Generation via R...Automatic generation of executable Blender code from natural language remains challenging, with state-of-the-art LLMs producing frequent syntactic error...1114Can Coding Agents Reproduce Findings in Computationa...Large language models are increasingly deployed as autonomous coding agents and have achieved remarkably strong performance on software engineering benc...1115Characterizing the Expressivity of Local Attention i...The transformer is the most popular neural architecture for language modeling. The cornerstone of the transformer is its global attention mechanism, whi...1116Code World Model Preparedness ReportThis report documents the preparedness assessment of Code World Model (CWM), a model for code generation and reasoning about code from Meta. We conducte...1117Concentration and Calibration in Predictive Bayesian...Predictive Bayesian inference (PBI) represents a model-and prior-agnostic approach to standard Bayesian inference which allows users to quantify uncerta...1118Decentralized Proximal Stochastic Gradient Langevin...We propose Decentralized Proximal Stochastic Gradient Langevin Dynamics (DE-PSGLD), a decentralized Markov chain Monte Carlo (MCMC) algorithm for sampli...1119Directed Social Regard: Surfacing Targeted Advocacy,...The language in online platforms, influence operations, and political rhetoric frequently directs a mix of pro-social sentiment (e.g., advocacy, helpful...1120EASE: Federated Multimodal Unlearning via Entangleme...Federated Multimodal Learning (FML) trains multimodal models across decentralized clients while keeping their image-text pairs private. However, joint e...1121EGREFINE: An Execution-Grounded Optimization Framewo...Text-to-SQL enables non-expert users to query databases in natural language, yet real-world schemas often suffer from ambiguous, abbreviated, or inconsi...1122Empowering Heterogeneous Graph Foundation Models via...While Graph Foundation Models (GFMs) have achieved remarkable success in homogeneous graphs, extending them to multi-domain heterogeneous graphs (MDHGs)...1123Exploring the Limits of End-to-End Feature-Affinity...Single-point supervised infrared small target detection (IRSTD) drastically reduces dense annotation costs. Current state-of-the-art (SOTA) methods achi...1124FinSafetyBench: Evaluating LLM Safety in Real-World...Large language models (LLMs) are increasingly applied in financial scenarios. However, they may produce harmful outputs, including facilitating illegal...1125Foundation AI Models for Aerosol Optical Depth Estim...Aerosol Optical Depth (AOD) retrieval is essential for Earth observation, supporting applications from air quality monitoring to climate studies. Conven...1126Generating Statistical Charts with Validation-Driven...Generating diverse, readable statistical charts from tabular data remains challenging for LLMs, as many failures become apparent after rendering and are...1127GeoContra: From Fluent GIS Code to Verifiable Spatia...Reliable spatial analysis in GIScience requires preserving coordinate semantics, topology, units, and geographic plausibility. Current LLM-based GIS sys...1128GMGaze: MoE-Based Context-Aware Gaze Estimation with...Gaze estimation methods commonly use facial appearances to predict the direction of a person gaze. However, previous studies show three major challenges...1129Gradient Regularized Newton Boosting Trees with Glob...Gradient Boosting Decision Trees (GBDTs) dominate tabular machine learning, with modern implementations like XGBoost, LightGBM, and CatBoost being based...1130Healthcare AI GYM for Medical AgentsClinical reasoning demands multi-step interactions -- gathering patient history, ordering tests, interpreting results, and making safe treatment decisio...1131Hierarchical Abstract Tree for Cross-Document Retrie...Retrieval-augmented generation (RAG) enhances large language models with external knowledge, and tree-based RAG organizes documents into hierarchical in...1132H-RAG at SemEval-2026 Task 8: Hierarchical Parent-Ch...We present H-RAG, our submission to SemEval-2026 Task 8 (MTRAGEval), addressing both Task A (Retrieval) and Task C (Generation with Retrieved Passages)....1133HyCOP: Hybrid Composition Operators for Interpretabl...We introduce HyCOP, a modular framework that learns parametric PDE solution operators by composing simple modules (advection, diffusion, learned closure...1134InpaintSLat: Inpainting Structured 3D Latents via In...We present a training-free approach for controllable 3D inpainting based on initial noise optimization. In the structured 3D latent diffusion framework,...1135LASE: Language-Adversarial Speaker Encoding for Indi...A speaker encoder used in multilingual voice cloning should treat the same speaker identically regardless of which script the audio was uttered in. Off-...1136Learning Coarse-to-Fine Osteoarthritis Representatio...Knee osteoarthritis (OA) assessment involves a natural but often underused label hierarchy: a coarse binary OA decision and a fine-grained Kellgren--Law...1137Learning How and What to Memorize: Cognition-Inspire...Large language model (LLM) agents require long-term user memory for consistent personalization, but limited context windows hinder tracking evolving pre...1138Learning the Helmholtz equation operator with DeepON...This paper deals with solving the 2D Helmholtz equation on non-parametric domains, leveraging a physics-informed neural operator network based on the De...1139Let ViT Speak: Generative Language-Image Pre-trainingIn this paper, we present \textbf{Gen}erative \textbf{L}anguage-\textbf{I}mage \textbf{P}re-training (GenLIP), a minimalist generative pretraining frame...1140Make Your LVLM KV Cache More LightweightKey-Value (KV) cache has become a de facto component of modern Large Vision-Language Models (LVLMs) for inference. While it enhances decoding efficiency...1141Map2World: Segment Map Conditioned Text to 3D World...3D world generation is essential for applications such as immersive content creation or autonomous driving simulation. Recent advances in 3D world gener...1142M-CaStLe: Uncovering Local Causal Structures in Mult...Causal graph discovery for space-time systems is challenging in high-dimensional gridded data, which often has many more grid cells than temporal observ...1143Meritocratic Fairness in Budgeted Combinatorial Mult...We propose a new framework for meritocratic fairness in budgeted combinatorial multi-armed bandits with full-bandit feedback (BCMAB-FBF). Unlike semi-ba...1144ML-Bench&Guard: Policy-Grounded Multilingual Saf...As Large Language Models (LLMs) are increasingly deployed in cross-linguistic contexts, ensuring safety in diverse regulatory and cultural environments...1145Modeling Subjective Urban Perception with Human GazeUrban perception describes how people subjectively evaluate urban environments, shaping how cities are experienced and understood. Existing computationa...1146NonZero: Interaction-Guided Exploration for Multi-Ag...Monte Carlo Tree Search (MCTS) scales poorly in cooperative multi-agent domains because expansion must consider an exponentially large set of joint acti...1147Observable Performance Does Not Fully Reflect System...In biomechanical systems, observable performance is often used as a proxy for underlying system organization. However, this assumption implicitly presum...1148Optimal Spatio-Temporal Decoupling for Bayesian Conf...Online Conformal Prediction (CP) struggles to balance temporal adaptability and structural stability. Feedback-driven methods (e.g., Adaptive Conformal...1149Persistent Visual Memory: Sustaining Perception for...While autoregressive Large Vision-Language Models (LVLMs) demonstrate remarkable proficiency in multimodal tasks, they face a 'Visual Signal Dilution' p...1150PhysEdit: Physically-Consistent Region-Aware Image E...Image editing instructions are heterogeneous: a color swap, an object insertion, and a physical-action edit all demand different spatial coverage and di...1151Position: agentic AI orchestration should be Bayes-c...LLMs excel at predictive tasks and complex reasoning tasks, but many high-value deployments rely on decisions under uncertainty, for example, which tool...1152Posterior Augmented Flow MatchingFlow matching (FM) trains a time-dependent vector field that transports samples from a simple prior to a complex data distribution. However, for high-di...1153Quantum Gradient-Based Approach for Edge and Corner...Edge detection refers to identifying points in a digital image where intensity changes sharply, indicating object boundaries or structural features. Cor...1154Quantum Interval Bound Propagation for Certified Tra...Quantum machine learning is a promising field for efficiently learning features of a dataset to perform a specified task, such as classification. Interv...1155Randomized Subspace Nesterov Accelerated GradientRandomized-subspace methods reduce the cost of first-order optimization by using only low-dimensional projected-gradient information, a feature that is...1156Recovering Hidden Reward in Diffusion-Based PoliciesThis paper introduces EnergyFlow, a framework that unifies generative action modeling with inverse reinforcement learning by parameterizing a scalar ene...1157Recursive Maximum Likelihood Estimation for Interact...We study recursive maximum likelihood estimation for stochastic interacting particle systems based on continuous observation of a single particle. In th...1158Reinforcement Learning with Markov Risk Measures and...For a risk-averse finite-horizon Markov Decision Problem, we introduce a special class of Markov coherent risk measures, called mini-batch measures. We...1159ResRL: Boosting LLM Reasoning via Negative Sample Pr...Reinforcement Learning with Verifiable Rewards (RLVR) enhances reasoning of Large Language Models (LLMs) but usually exhibits limited generation diversi...1160RunAgent: Interpreting Natural-Language Plans with C...Humans solve problems by executing targeted plans, yet large language models (LLMs) remain unreliable for structured workflow execution. We propose RunA...1161SAVGO: Learning State-Action Value Geometry with Cos...While representation and similarity learning have improved the sample efficiency of Reinforcement Learning (RL), they are rarely used to shape policy up...1162SC-Taxo: Hierarchical Taxonomy Generation under Sema...Scientific literature is expanding at an unprecedented pace, making it increasingly challenging to efficiently organize and access domain knowledge. A h...1163Static and Dynamic Graph Alignment Network for Tempo...Temporal Video Grounding (TVG) aims to localize temporal moments in an untrimmed video that semantically correspond to given natural language queries. R...1164Temporal Data Requirement for Predicting Unplanned H...With the proliferation of Electronic Health Records (EHRs), a critical challenge in building predictive models is determining the optimal historical dat...1165Themis: Training Robust Multilingual Code Reward Mod...Reward models (RMs) have become an indispensable fixture of the language model (LM) post-training playbook, enabling policy alignment and test-time scal...1166To Call or Not to Call: A Framework to Assess and Op...Agentic AI architectures augment LLMs with external tools, unlocking strong capabilities. However, tool use is not always beneficial; some calls may be...1167Towards Customized Multimodal Role-PlayUnified multimodal understanding and generation models enable richer human-AI interaction. Yet jointly customizing a character's persona, dialogue style...1168Towards Improving Speaker Distance Estimation throug...The Room Acoustics and Speaker Distance Estimation (SDE) Challenge at ICASSP 2025 explores the effectiveness of augmented room impulse response (RIR) da...1169Uniform-Correct Policy Optimization: Breaking RLVR's...Reinforcement Learning with Verifiable Rewards (RLVR) has achieved substantial gains in single-attempt accuracy (Pass@1) on reasoning tasks, yet often s...1170UniVidX: A Unified Multimodal Framework for Versatil...Recent progress has shown that video diffusion models (VDMs) can be repurposed for diverse multimodal graphics tasks. However, existing methods often tr...1171Unpaired Image Deraining Using Reward-Guided Self-Re...Unsupervised deraining has attracted attention for its ability to learn the real-world distribution of rain without paired supervision. However, the lac...1172Unsupervised Denoising of Real Clinical Low Dose Liv...With the development of deep learning, medical image processing has been widely used to assist clinical research. This paper focuses on the denoising pr...1173When LLMs Stop Following Steps: A Diagnostic Study o...Large language models (LLMs) often achieve strong performance on reasoning benchmarks, but final-answer accuracy alone does not show whether they faithf...1174When RAG Chatbots Expose Their Backend: An Anonymize...Background: Patient-facing medical chatbots based on retrieval-augmented generation (RAG) are increasingly promoted to deliver accessible, grounded heal...1175Agentic AI Systems Should Be Designed as Marginal To...This position paper argues that agentic AI systems should be designed and evaluated as marginal token allocation economies rather than as text generator...1176Chain of Evidence: Pixel-Level Visual Attribution fo...Iterative Retrieval-Augmented Generation (iRAG) has emerged as a powerful paradigm for answering complex multi-hop questions by progressively retrieving...1177ESARBench: A Benchmark for Agentic UAV Embodied Sear...The rapid advancement of Multimodal Large Language Models (MLLMs) has empowered Unmanned Aerial Vehicle (UAV) with exceptional capabilities in spatial r...1178Hallucinations Undermine Trust; Metacognition is a W...Despite significant strides in factual reliability, errors -- often termed hallucinations -- remain a major concern for generative AI, especially as LLM...1179Prescriptive Scaling Laws for Data Constrained TrainingTraining compute is increasingly outpacing the availability of high-quality data. This shifts the central challenge from optimal compute allocation to e...1180SplAttN: Bridging 2D and 3D with Gaussian Soft Splat...Although multi-modal learning has advanced point cloud completion, the theoretical mechanisms remain unclear. Recent works attribute success to the conn...1181TT4D: A Pipeline and Dataset for Table Tennis 4D Rec...We present TT4D, a large-scale, high-fidelity table tennis dataset. It provides 140+ hours of reconstructed singles and doubles gameplay from monocular...1182Turning Drift into Constraint: Robust Reasoning Alig...This paper identifies a critical yet underexplored challenge in reasoning alignment from multiple multi-modal large language models (MLLMs): In non-stat...1183Beyond Semantic Similarity: Rethinking Retrieval for...Modern retrieval systems, whether lexical or semantic, expose a corpus through a fixed similarity interface that compresses access into a single top-k r...1184Counting as a minimal probe of language model reliab...Large language models perform strongly on benchmarks in mathematical reasoning, coding and document analysis, suggesting a broad ability to follow instr...1185From Context to Skills: Can Language Models Learn fr...Many real-world tasks require language models (LMs) to reason over complex contexts that exceed their parametric knowledge. This calls for context learn...1186Linear-Time Global Visual Modeling without Explicit...Existing research largely attributes the global sequence modeling capability of Transformers to the explicit computation of attention weights, a process...1187Motion-Aware Caching for Efficient Autoregressive Vi...Autoregressive video generation paradigms offer theoretical promise for long video synthesis, yet their practical deployment is hindered by the computat...1188TCDA: Thread-Constrained Discourse-Aware Modeling fo...Conversational Aspect-based Sentiment Quadruple Analysis (DiaASQ) needs to capture the complex interrelationships in multiple rounds of dialogues. Exist...1189A Hybrid Approach for Closing the Sim2real Appearanc...Video game engines have been an important source for generating large volumes of visual synthetic datasets for training and evaluating computer vision a...1190AcademiClaw: When Students Set Challenges for AI AgentsBenchmarks within the OpenClaw ecosystem have thus far evaluated exclusively assistant-level tasks, leaving the academic-level capabilities of OpenClaw...1191ARIS: Autonomous Research via Adversarial Multi-Agen...This report describes ARIS (Auto-Research-in-sleep), an open-source research harness for autonomous research, including its architecture, assurance mech...1192Generative Modeling with Orbit-Space Particle Flow M...We present Orbit-Space Geometric Probability Paths (OGPP), a particle-native flow-matching framework for generative modeling of particle systems. OGPP i...1193HeavySkill: Heavy Thinking as the Inner Skill in Age...Recent advances in agentic harness with orchestration frameworks that coordinate multiple agents with memory, skills, and tool use have achieved remarka...1194KinDER: A Physical Reasoning Benchmark for Robot Lea...Robotic systems that interact with the physical world must reason about kinematic and dynamic constraints imposed by their own embodiment, their environ...1195MolmoAct2: Action Reasoning Models for Real-world De...Vision-Language-Action (VLA) models aim to provide a single generalist controller for robots, but today's systems fall short on the criteria that matter...1196Perceptual Flow Network for Visually Grounded ReasoningDespite the success of Large-Vision Language Models (LVLMs), general optimization objectives (e.g., standard MLE) fail to constrain visual trajectories,...1197PhysicianBench: Evaluating LLM Agents in Real-World...We introduce PhysicianBench, a benchmark for evaluating LLM agents on physician tasks grounded in real clinical setting within electronic health record...1198Reinforcement Learning for LLM-based Multi-Agent Sys...As large language model (LLM) agents evolve from isolated tool users into coordinated teams, reinforcement learning (RL) must optimize not only individu...1199SVGS: Enhancing Gaussian Splatting Using Primitives...Gaussian Splatting demonstrates impressive results in multi-view reconstruction based on Gaussian explicit representations. However, the current Gaussia...1200T^2PO: Uncertainty-Guided Exploration Control for St...Recent progress in multi-turn reinforcement learning (RL) has significantly improved reasoning LLMs' performances on complex interactive tasks. Despite...1201The TTS-STT Flywheel: Synthetic Entity-Dense Audio C...Niche-domain Indic ASR -- digit strings, currency amounts, addresses, brand names, English/Indic codemix -- is under-served by both open-source SOTA and...1202Video Generation with Predictive LatentsVideo Variational Autoencoder (VAE) enables latent video generative modeling by mapping the visual world into compact spatiotemporal latent spaces, impr...1203A Benchmark for Interactive World Models with a Unif...Achieving Artificial General Intelligence (AGI) requires agents that learn and interact adaptively, with interactive world models providing scalable env...1204APEX: Large-scale Multi-task Aesthetic-Informed Popu...Music popularity prediction has attracted growing research interest, with relevance to artists, platforms, and recommendation systems. However, the expl...1205Audio-Visual Intelligence in Large Foundation ModelsAudio-Visual Intelligence (AVI) has emerged as a central frontier in artificial intelligence, bridging auditory and visual modalities to enable machines...1206Awaking Spatial Intelligence in Unified Multimodal U...We present JoyAI-Image, a unified multimodal foundation model for visual understanding, text-to-image generation, and instruction-guided image editing....1207CASCADE: Case-Based Continual Adaptation for Large L...Large language models (LLMs) have become a central foundation of modern artificial intelligence, yet their lifecycle remains constrained by a rigid sepa...1208OpenSeeker-v2: Pushing the Limits of Search Agents w...Deep search capabilities have become an indispensable competency for frontier Large Language Model (LLM) agents, yet their development remains dominated...1209Parameter-Efficient Multi-View Proficiency Estimatio...Estimating how well a person performs an action, rather than which action is performed, is central to coaching, rehabilitation, and talent identificatio...1210PatRe: A Full-Stage Office Action and Rebuttal Gener...Patent examination is a complex, multi-stage process requiring both technical expertise and legal reasoning, increasingly challenged by rising applicati...1211Rethinking Reasoning-Intensive Retrieval: Evaluating...Reasoning-intensive retrieval aims to surface evidence that supports downstream reasoning rather than merely matching topical similarity. This capabilit...1212RLDX-1 Technical ReportWhile Vision-Language-Action models (VLAs) have shown remarkable progress toward human-like generalist robotic policies through the versatile intelligen...1213Stream-R1: Reliability-Perplexity Aware Reward Disti...Distillation-based acceleration has become foundational for making autoregressive streaming video diffusion models practical, with distribution matching...1214SymptomAI: Towards a Conversational AI Agent for Eve...Language models excel at diagnostic assessments on currated medical case-studies and vignettes, performing on par with, or better than, clinical profess...1215The Scaling Properties of Implicit Deductive Reasoni...We investigate the scaling properties of implicit deductive reasoning over Horn clauses in depth-bounded Transformers. By systematically decorrelating p...1216Traceprop: ML Data ProvenanceThe first unified system connecting raw source files through preprocessing and training to individual predictions, with gradient-based attribution and provenance-guided machine unlearning. Sub-1% lineage overhead, 266x faster than TRAK on CPU, exceeds retrain-from-scratch unlearning gold standard.1217Workspace-Bench 1.0: Benchmarking AI Agents on Works...Workspace learning requires AI agents to identify, reason over, exploit, and update explicit and implicit dependencies among heterogeneous files in a wo...1218A Foundation Model for Zero-Shot Logical Rule InductionInductive Logic Programming (ILP) learns interpretable logical rules from data. Existing methods are transductive: their learned parameters are bound to...1219Beyond Retrieval: A Multitask Benchmark and Model fo...Code search has usually been evaluated as first-stage retrieval, even though production systems rely on broader pipelines with reranking and developer-s...1220CreativityBench: Evaluating Agent Creative Reasoning...Recent advances in large language models have led to strong performance on reasoning and environment-interaction tasks, yet their ability for creative p...1221D-OPSD: On-Policy Self-Distillation for Continuously...The landscape of high-performance image generation models is currently shifting from the inefficient multi-step ones to the efficient few-step counterpa...1222Generative Quantum-inspired Kolmogorov-Arnold Eigens...High-performance computing (HPC) is increasingly important for scalable quantum chemistry workflows that couple classical generative models, quantum cir...1223KernelBench-X: A Comprehensive Benchmark for Evaluat...LLM-based Triton kernel generation has attracted significant interest, yet a fundamental empirical question remains unanswered: where does this capabili...1224Lightning Unified Video Editing via In-Context Spars...Video editing has evolved toward In-Context Learning (ICL) paradigms, yet the resulting quadratic attention costs create a critical computational bottle...1225OpenSearch-VL: An Open Recipe for Frontier Multimoda...Deep search has become a crucial capability for frontier multimodal agents, enabling models to solve complex questions through active search, evidence v...1226PhysForge: Generating Physics-Grounded 3D Assets for...Synthesizing physics-grounded 3D assets is a critical bottleneck for interactive virtual worlds and embodied AI. Existing methods predominantly focus on...1227RaguTeam at SemEval-2026 Task 8: Meno and Friends in...We present our winning system for Task~B (generation with reference passages) in SemEval-2026 Task~8: MTRAGEval. Our method is a heterogeneous ensemble...1228ReflectDrive-2: Reinforcement-Learning-Aligned Self-...We introduce ReflectDrive-2, a masked discrete diffusion planner with separate action expert for autonomous driving that represents plans as discrete tr...1229RemoteZero: Geospatial Reasoning with Zero Human Ann...Geospatial reasoning requires models to resolve complex spatial semantics and user intent into precise target locations for Earth observation. Recent pr...1230StableI2I: Spotting Unintended Changes in Image-to-I...In most real-world image-to-image (I2I) scenarios, existing evaluations primarily focus on instruction following and the perceptual quality or aesthetic...1231Stream-T1: Test-Time Scaling for Streaming Video Gen...While Test-Time Scaling (TTS) offers a promising direction to enhance video generation without the surging costs of training, current test-time video ge...1232SWE-WebDevBench: Evaluating Coding Agent Application...The emergence of 'vibe coding' platforms, where users describe applications in natural language and AI agents autonomously generate full-stack software,...1233TabEmbed: Benchmarking and Learning Generalist Embed...Foundation models have established unified representations for natural language processing, yet this paradigm remains largely unexplored for tabular dat...1234The First Token Knows: Single-Decode Confidence for...Self-consistency detects hallucinations by generating multiple sampled answers to a question and measuring agreement, but this requires repeated decodin...1235When to Think, When to Speak: Learning Disclosure Po...In single-stream autoregressive interfaces, the same tokens both update the model state and constitute an irreversible public commitment. This coupling...12364DThinker: Thinking with 4D Imagery for Dynamic Spat...Dynamic spatial reasoning from monocular video is essential for bridging visual intelligence and the physical world, yet remains challenging for vision-...1237A^2RD: Agentic Autoregressive Diffusion for Long Vid...Synthesizing consistent and coherent long video remains a fundamental challenge. Existing methods suffer from semantic drift and narrative collapse over...1238A^2TGPO: Agentic Turn-Group Policy Optimization with...Reinforcement learning for agentic large language models (LLMs) typically relies on a sparse, trajectory-level outcome reward, making it difficult to ev...1239AI Co-Mathematician: Accelerating Mathematicians wit...We introduce the AI co-mathematician, a workbench for mathematicians to interactively leverage AI agents to pursue open-ended research. The AI co-mathem...1240Are We Making Progress in Multimodal Domain Generali...Despite the growing popularity of Multimodal Domain Generalization (MMDG) for enhancing model robustness, it remains unclear whether reported performanc...1241Auto Research with Specialist Agents Develops Effect...We study auto research as a closed empirical loop driven by external measurement. Each submitted trial carries a hypothesis, an executable code edit, an...1242BioTool: A Comprehensive Tool-Calling Dataset for En...Despite the success of large language models (LLMs) on general-purpose tasks, their performance in highly specialized domains such as biomedicine remain...1243Can RL Teach Long-Horizon Reasoning to LLMs? Express...Reinforcement learning (RL) has been applied to improve large language model (LLM) reasoning, yet the systematic study of how training scales with task...1244Continuous Latent Diffusion Language ModelLarge language models have achieved remarkable success under the autoregressive paradigm, yet high-quality text generation need not be tied to a fixed l...1245Continuous-Time Distribution Matching for Few-Step D...Step distillation has become a leading technique for accelerating diffusion models, among which Distribution Matching Distillation (DMD) and Consistency...1246CPCANet: Deep Unfolding Common Principal Component A...Domain Generalization (DG) aims to learn representations that remain robust under out-of-distribution (OOD) shifts and generalize effectively to unseen...1247EMO: Pretraining Mixture of Experts for Emergent Mod...Large language models are typically deployed as monolithic systems, requiring the full model even when applications need only a narrow subset of capabil...1248From Storage to Experience: A Survey on the Evolutio...Large Language Model (LLM)-based agents have fundamentally reshaped artificial intelligence by integrating external tools and planning capabilities. Whi...1249GeoStack: A Framework for Quasi-Abelian Knowledge Co...We address the challenge of knowledge composition in Vision-Language Models (VLMs), where accumulating expertise across multiple domains or tasks typica...1250HumanNet: Scaling Human-centric Video Learning to On...Progress in embodied intelligence increasingly depends on scalable data infrastructure. While vision and language have scaled with internet corpora, lea...1251IntentGrasp: A Comprehensive Benchmark for Intent Un...Accurately understanding the intent behind speech, conversation, and writing is crucial to the development of helpful Large Language Model (LLM) assista...1252Listwise Policy Optimization: Group-based RLVR as Ta...Reinforcement learning with verifiable rewards (RLVR) has become a standard approach for large language models (LLMs) post-training to incentivize reaso...1253LiVeAction: a Lightweight, Versatile, and Asymmetric...Modern sensors generate rich, high-fidelity data, yet applications operating on wearable or remote sensing devices remain constrained by bandwidth and p...1254Long Context Pre-Training with Lighthouse AttentionTraining causal transformers at extreme sequence lengths is bottlenecked by the quadratic time and memory of scaled dot-product attention (SDPA). In thi...1255MARBLE: Multi-Aspect Reward Balance for Diffusion RLReinforcement learning fine-tuning has become the dominant approach for aligning diffusion models with human preferences. However, assessing images is i...1256MDN: Parallelizing Stepwise Momentum for Delta Linea...Linear Attention (LA) offers a promising paradigm for scaling large language models (LLMs) to long sequences by avoiding the quadratic complexity of sel...1257MiA-Signature: Approximating Global Activation for L...A growing body of work in cognitive science suggests that reportable conscious access is associated with global ignition over distributed memory systems...1258Nonsense Helps: Prompt Space Perturbation Broadens R...Reinforcement learning with verifiable rewards, particularly Group Relative Policy Optimization (GRPO), has significantly advanced the reasoning capabil...1259PianoCoRe: Combined and Refined Piano MIDI DatasetSymbolic music datasets with matched scores and performances are essential for many music information retrieval (MIR) tasks. Yet, existing resources oft...1260Skill1: Unified Evolution of Skill-Augmented Agents...A persistent skill library allows language model agents to reuse successful strategies across tasks. Maintaining such a library requires three coupled c...1261SkillOS: Learning Skill Curation for Self-Evolving A...LLM-based agents are increasingly deployed to handle streaming tasks, yet they often remain one-off problem solvers that fail to learn from past interac...1262Sparkle: Realizing Lively Instruction-Guided Video B...In recent years, open-source efforts like Senorita-2M have propelled video editing toward natural language instruction. However, current publicly availa...1263Steering Visual Generation in Unified Multimodal Mod...Unified multimodal models are envisioned to bridge the gap between understanding and generation. Yet, to achieve competitive performance, state-of-the-a...1264StraTA: Incentivizing Agentic Reinforcement Learning...Large language models (LLMs) are increasingly used as interactive agents, but optimizing them for long-horizon decision making remains difficult because...1265SwiftI2V: Efficient High-Resolution Image-to-Video G...High-resolution image-to-video (I2V) generation aims to synthesize realistic temporal dynamics while preserving fine-grained appearance details of the i...1266The Granularity Axis: A Micro-to-Macro Latent Direct...Large language models (LLMs) are routinely prompted to take on social roles ranging from individuals to institutions, yet it remains unclear whether the...1267Think, then Score: Decoupled Reasoning and Scoring f...Recent advances in generative video models are increasingly driven by post-training and test-time scaling, both of which critically depend on the qualit...1268TIDE: Every Layer Knows the Token Beneath the ContextWe revisit a universally accepted but under-examined design choice in every modern LLM: a token index is looked up once at the input embedding layer and...1269UniPool: A Globally Shared Expert Pool for Mixture-o...Modern Mixture-of-Experts (MoE) architectures allocate expert capacity through a rigid per-layer rule: each transformer layer owns a separate expert set...1270UniPrefill: Universal Long-Context Prefill Accelerat...As large language models (LLMs) continue to advance rapidly, they are becoming increasingly capable while simultaneously demanding ever-longer context l...1271UniSD: Towards a Unified Self-Distillation Framework...Self-distillation (SD) offers a promising path for adapting large language models (LLMs) without relying on stronger external teachers. However, SD in a...1272When No Benchmark Exists: Validating Comparative LLM...Many deployments must compare candidate language models for safety before a labeled benchmark exists for the relevant language, sector, or regulatory re...1273When to Trust Imagination: Adaptive Action Execution...World Action Models (WAMs) have recently emerged as a promising paradigm for robotic manipulation by jointly predicting future visual observations and f...1274XL-SafetyBench: A Country-Grounded Cross-Cultural Be...Current LLM safety benchmarks are predominantly English-centric and often rely on translation, failing to capture country-specific harms. Moreover, they...1275Anisotropic Modality AlignTraining multimodal large language models has long been limited by the scarcity of high-quality paired multimodal data. Recent studies show that the sha...1276Fast Byte Latent TransformerRecent byte-level language models (LMs) match the performance of token-level models without relying on subword vocabularies, yet their utility is limite...1277Flow-OPD: On-Policy Distillation for Flow Matching M...Existing Flow Matching (FM) text-to-image models suffer from two critical bottlenecks under multi-task alignment: the reward sparsity induced by scalar-...1278HyperEyes: Dual-Grained Efficiency-Aware Reinforceme...Existing multimodal search agents process target entities sequentially, issuing one tool call per entity and accumulating redundant interaction rounds w...1279InterLV-Search: Benchmarking Interleaved Multimodal...Existing benchmarks for multimodal agentic search evaluate multimodal search and visual browsing, but visual evidence is either confined to the input or...1280LLMs Improving LLMs: Agentic Discovery for Test-Time...Test-time scaling (TTS) has become an effective approach for improving large language model performance by allocating additional computation during infe...1281MISA: Mixture of Indexer Sparse Attention for Long-C...DeepSeek Sparse Attention (DSA) sets the state of the art for fine-grained inference-time sparse attention by introducing a learned token-wise indexer t...1282Normalizing Trajectory ModelsDiffusion-based models decompose sampling into many small Gaussian denoising steps -- an assumption that breaks down when generation is compressed to a...1283Scaling Continual Learning to 300+ Tasks with Bi-Lev...Continual learning, especially class-incremental learning (CIL), on the basis of a pre-trained model (PTM) has garnered substantial research interest in...1284STARFlow2: Bridging Language Models and Normalizing...Deep generative models have advanced rapidly across text and vision, motivating unified multimodal systems that can understand, reason over, and generat...1285TextLDM: Language Modeling with Continuous Latent Di...Diffusion Transformers (DiT) trained with flow matching in a VAE latent space have unified visual generation across images and videos. A natural next st...1286What Matters for Diffusion-Friendly Latent Manifold?...Tokenizers are a crucial component of latent diffusion models, as they define the latent space in which diffusion models operate. However, existing toke...1287Who Prices Cognitive Labor in the Age of Agents? Com...A natural intuition about the economics of AI agents is that, because agents can be replicated at very low marginal cost, agent labor may be supplied hi...1288Seeing the Needle in the Haystack: Towards Weakly-Su...Log anomaly detection is a critical task for system operations and security assurance. However, in networked systems at scale, log data are generated at...1289Forcing-KV: Hybrid KV Cache Compression for Efficien...Autoregressive (AR) video diffusion models adopt a streaming generation framework, enabling long-horizon video generation with real-time responsiveness,...1290WildClawBench: A Benchmark for Real-World, Long-Hori...Large language and vision-language models increasingly power agents that act on a user's behalf through command-line interface (CLI) harnesses. However,...1291OmniHumanoid: Streaming Cross-Embodiment Video Gener...Cross-embodiment video generation aims to transfer motions across different humanoid embodiments, such as human-to-robot and robot-to-robot, enabling sc...1292Achieving Gold-Medal-Level Olympiad Reasoning via Si...Recent progress in reasoning models has substantially advanced long-horizon mathematical and scientific problem solving, with several systems now reachi...1293BOOKMARKS: Efficient Active Storyline Memory for Rol...Memory systems are critical for role-playing agents (RPAs) to maintain long-horizon consistency. However, existing RPA memory methods (e.g., profiling)...1294CurveBench: A Benchmark for Exact Topological Reason...We introduce CurveBench, a benchmark for hierarchical topological reasoning from visual input. CurveBench consists of 756 images of pairwise non-interse...1295Beyond Individual Intelligence: Surveying Collaborat...LLM-based autonomous agents have demonstrated strong capabilities in reasoning, planning, and tool use, yet remain limited when tasks require sustained...1296Does Synthetic Layered Design Data Benefit Layered D...Recent advances in image generation have made it easy to produce high-quality images. However, these outputs are inherently flattened, entangling foregr...1297EverAnimate: Minute-Scale Human Animation via Latent...We propose EverAnimate, an efficient post-training method for long-horizon animated video generation that preserves visual quality and character identit...1298FrontierSmith: Synthesizing Open-Ended Coding Proble...Many real-world coding challenges are open-ended and admit no known optimal solution. Yet, recent progress in LLM coding has focused on well-defined tas...1299InsightTok: Improving Text and Face Fidelity in Disc...Text and faces are among the most perceptually salient and practically important patterns in visual generation, yet they remain challenging for autoregr...1300IntentVLA: Short-Horizon Intent Modeling for Aliased...Robot imitation data are often multimodal: similar visual-language observations may be followed by different action chunks because human demonstrators a...1301MemEye: A Visual-Centric Evaluation Framework for Mu...Long-term agent memory is increasingly multimodal, yet existing evaluations rarely test whether agents preserve the visual evidence needed for later rea...1302MemLens: Benchmarking Multimodal Long-Term Memory in...Memory is essential for large vision-language models (LVLMs) to handle long, multimodal interactions, with two method directions providing this capabili...1303MMSkills: Towards Multimodal Skills for General Visu...Reusable skills have become a core substrate for improving agent capabilities, yet most existing skill packages encode reusable behavior primarily as te...1304Orchard: An Open-Source Agentic Modeling FrameworkAgentic modeling aims to transform LLMs into autonomous agents capable of solving complex tasks through planning, reasoning, tool use, and multi-turn in...1305PhyMotion: Structured 3D Motion Reward for Physics-G...Generating realistic human motion is a central yet unsolved challenge in video generation. While reinforcement learning (RL)-based post-training has dri...1306RAVEN: Real-time Autoregressive Video Extrapolation...Causal autoregressive video diffusion models support real-time streaming generation by extrapolating future chunks from previously generated content. Di...1307ReactiveGWM: Steering NPC in Reactive Game World ModelsCurrent game world models simulate environments from a subjective, player-centric perspective. However, by treating the Non-Player Character (NPC) merel...1308Self-Distilled Agentic Reinforcement LearningReinforcement learning (RL) has emerged as a central paradigm for post-training LLM agents, yet its trajectory-level reward signal provides only coarse...1309ViMU: Benchmarking Video Metaphorical UnderstandingAny new medium, once it emerges, is used for more than the transmission of overt content alone. The information it carries typically operates on two lev...1310Warp-as-History: Generalizable Camera-Controlled Vid...Camera-controlled video generation has made substantial progress, enabling generated videos to follow prescribed viewpoint trajectories. However, existi...1311DexJoCo: A Benchmark and Toolkit for Task-Oriented D...Achieving human-level manipulation requires dexterous robotic hands capable of complex object interactions. Advancing such capabilities further demands...1312FashionChameleon: Towards Real-Time and Interactive...Human-centric video customization, particularly at the garment level, has shown significant commercial value. However, existing approaches cannot suppor...1313Flash-GRPO: Efficient Alignment for Video Diffusion...Group Relative Policy Optimization has emerged as essential for aligning video diffusion models with human preferences, but faces a critical computation...1314Nudging Beyond the Comfort Zone: Efficient Strategy-...Reinforcement learning with verifiable rewards (RLVR) has emerged as a scalable paradigm for improving the reasoning capabilities of large language mode...1315PAGER: Bridging the Semantic-Execution Gap in Point-...Large vision-language models have significantly advanced GUI agents, enabling executable interaction across web, mobile, and desktop interfaces. Yet the...1316Unlocking Dense Metric Depth Estimation in VLMsVision-Language Models (VLMs) excel at 2D tasks such as grounding and captioning, yet remain limited in 3D understanding. A key limitation is their text...1317WorldAct: Activating Monolithic 3D Worlds into Inter...Recent 3D world modeling systems based on generative scene synthesis, such as Marble, can create coherent and explorable 3D environments, yet their outp...1318MemForest: An Efficient Agent Memory System with Hie...Memory is a fundamental component for enabling long-context LLM agents, supporting persistent state across interactions through a continuous serve-and-u...1319Soap2Soap: Long Cinematic Video Remaking via Multi-A...We study series-level cinematic remaking, a long-horizon video-to-video generation problem that localizes full episodes or films via stylization or acto...1320ParaVT: Taming the Tool Prior Paradox for Parallel T...Training large multimodal models (LMMs) via reinforcement learning (RL) to natively invoke video-processing tools (e.g., cropping) has become a promisin...1321Broadening Access to Transportation Safety Data with...Transportation safety analysis requires integrating crash records, roadway attributes, and geospatial data through GIS-based workflows, but access remai...1322Decoupling Communication from Policy: Robust MARL un...Communication enables coordination in multi-agent reinforcement learning (MARL), but many real-world applications, e.g., search-and-rescue with drone sw...1323How Far Will They Go? Red-Teaming Online Influence w...As large language model (LLM)-based agents increasingly participate in online discourse, red-teaming their capacity to support political influence campa...1324RankJudge: A Multi-Turn LLM-as-a-Judge Synthetic Ben...As interactive LLM-based applications are created and refined, model developers need to evaluate the quality of generated text along many possible axes....1325Reflective Prompt Tuning through Language Model Func...Large language models (LLMs) have become increasingly capable of following instructions and complex reasoning, making prompting a flexible interface for...1326ZeroUnlearn: Few-Shot Knowledge Unlearning in Large...Large language models inevitably retain sensitive information, defined as inputs that may induce harmful generations, due to training on massive web cor...1327Agentic CLEAR: Automating Multi-Level Evaluation of...Agentic systems are becoming more capable: agents define strategies, take actions, and interact with different environments. This autonomy poses serious...1328Learning A Unified Risk Map for Autonomous Driving i...Occlusion-aware prediction remains a critical challenge in autonomous driving due to the inherent uncertainty of unobserved regions. Existing approaches...1329MotiMotion: Motion-Controlled Video Generation with...Current motion-controlled image-to-video generation models rigidly follow user-provided trajectories that are often sparse, imprecise, and causally inco...1330Representation over Routing: Overcoming Surrogate Ha...Temporal credit assignment in reinforcement learning has long been a central challenge. Inspired by the multi-timescale encoding of the dopamine system...1331ThriftAttention: Selective Mixed Precision for Long-...Efficient attention algorithms are critical to mitigate the quadratic cost of attention in long-context workloads. Prior work utilises block-scaled quan...1332Understanding Data Temporality Impact on Large Langu...Large language models (LLMs) are typically trained on shuffled corpora, yielding models whose knowledge is frozen at train time and whose temporal groun...1333Uniform Diffusion Models Revisited: Leave-One-Out De...Discrete diffusion models are often trained through clean-data prediction, but the prediction can be used in different ways to define the reverse dynami...1334AutoResearch AI: Towards AI-Powered Research Automat...Scientific research is being reshaped by AI systems that move beyond isolated assistance toward longer-horizon workflows spanning literature grounding,...1335Breaking the Chains of Probability: Neutrosophic Log...Large Language Models (LLMs) are predominantly governed by probabilistic frameworks in which the sum of outcome probabilities is constrained to unity. T...1336Coloring the Noise: Adversarial Sobolev Alignment fo...Generative priors in Image Super-Resolution (SR) often compromise faithful restoration, we attribute this limitation to a fundamental spectral misalignm...1337Convex Low-resource Accent-Robust Language Detection...Globalization and multiculturalism continue to produce increasingly diverse speech varieties. Yet current spoken dialogue systems frequently fail on und...1338CoSPlay: Cooperative Self-Play at Test-Time with Sel...Recently, Reinforcement Learning with Verifiable Rewards (RLVR) and Test-Time Scaling (TTS) have advanced LLM code generation through executable verific...1339CRONOS: Benchmarking Counterfactual Physical Consist...Video prediction is increasingly viewed as a path toward generalizable world models, yet it remains unclear whether these systems learn underlying causa...1340EvalVerse: Pipeline-Aware and Expert-Calibrated Benc...The rapid evolution of generative video foundation models has propelled the field toward professional-grade cinematic synthesis. To achieve such demandi...1341FastKernels: Benchmarking GPU Kernel Generation in P...LLM-based agents for GPU kernel generation are advancing rapidly, yet their progress is fundamentally constrained by the benchmarks they optimize agains...1342Foundation Protocol: A Coordination Layer for Agenti...Autonomous agents are moving from tools into a layer of social infrastructure: they browse, purchase, deploy software, manage systems, and increasingly...1343HorizonStream: Long-Horizon Attention for Streaming...Online 3D reconstruction requires estimating camera pose and scene geometry under strict causal and bounded-memory constraints. Existing methods often s...1344QUEST: Training Frontier Deep Research Agents with F...Deep research agents extend the role of search engines from retrieving keyword-matched pages to synthesizing knowledge, fundamentally changing how human...1345SkillEvolBench: Benchmarking the Evolution from Epis...Large language model (LLM) agents accumulate rich episodic trajectories while solving real-world tasks, but it remains unclear whether such experience c...1346Towards Evaluation Engineering: An Empirical Study o...Evaluation harnesses are software systems that orchestrate model evaluation by managing model invocation, data loading, metric computation, and result r...1347AgentFugue: Agent Scaling for Long-Horizon Tasks thr...Recent progress on long-horizon agentic tasks has been driven largely by scaling up individual agents through stronger models, better tools, and more ef...1348Benchmarking Composed Image Retrieval for Applied Ea...Remote sensing composed image retrieval (RSCIR) enables search in large satellite image archives using composed queries that combine a reference image w...1349ECHO: Terminal Agents Learn World Models for FreeCLI agents are the closest thing language models have to an embodied setting: the model emits commands, the terminal executes them, and the returned str...1350SAM: State-Adaptive Memory for Long-Horizon Reasonin...Long-horizon agentic reasoning requires large language models to act over long interaction histories containing thoughts, tool calls, observations, and...1351SEAL: Synergistic Co-Evolution of Agents and Learnin...Large Language Model (LLM) agents are increasingly improved through interaction, yet most self-evolution methods adapt either the policy or the learning...1352CONF-KV: Confidence-Aware KV Cache Eviction with Mix...Long-horizon LLM inference turns the key--value (KV) cache into the dominant GPU memory consumer and makes per-token attention increasingly expensive. M...1353DarkForest: Less Talk, Higher Accuracy for Multi-Age...Multi-agent LLM systems improve reasoning by combining outputs from multiple agents, but interaction-heavy methods can introduce error propagation and h...1354Directional Alignment Mitigates Reward Hacking in Re...Reward hacking arises when a model improves a proxy reward by exploiting shortcuts rather than solving the intended task. We study this failure mode thr...1355Faithfulness Metrics Don't Measure Faithfulness: A M...Chains of thought (CoTs) have become central in interpreting and auditing behaviors of large language models. Yet growing evidence suggests that these t...1356Geometry-Aware Image Flow MatchingRecent advances in generative models highlight the power of geometry-aware modeling in manifold-constrained settings. Yet, for natural images, the field...1357Injecting Image Guidance into Text-Conditioned Diffu...Text-to-image diffusion models like Stable Diffusion generate high-quality images from text, but lack a way to inject visual guidance (e.g. sketches, st...1358Learning High-Frequency Continuous Action Chunks in...Modern robotic policies increasingly rely on action chunking to execute complex tasks in the physical world. While action chunking improves temporal con...1359Macaron-A2UI: A Model for Generative UI in Personal...As personal agents evolve to handle complex, user-centric tasks, static plain-text chat is rapidly becoming a bottleneck. Generative UI emerges as the n...1360Multi-view Consistent 3D Gaussian Head Avatars 'with...High-fidelity 3D Gaussian head avatar generation is critical for applications such as AR/VR, telepresence, and digital humans. Existing methods depend o...1361SimuWoB: Simulating Real-World Mobile Apps for Fast...Mobile GUI agents powered by large language models have progressed rapidly, creating urgent needs for realistic and comprehensive evaluation. Existing b...1362STREAM: A Data-Centric Framework for Mining High-Val...Large language models for vertical domains are bottlenecked by the scarcity of complex, domain-specific task-oriented dialogues. Existing data acquisiti...1363Your Embedding Model is SMARTer Than You ThinkMultimodal retrieval relies heavily on single-vector retrievers, which compress rich, sequential token sequences into one single global representation....1364Advancing Creative Physical Intelligence in Large Mu...Large multimodal models (LMMs) have rapidly advanced in perception and reasoning; however, it remains unclear whether these capabilities generalize to d...1365AgentHijack: Benchmarking Computer Use Agent Robustn...Autonomous computer use agents that powered by multimodal large language models (MLLMs) are emerging as capable assistants for completing complex digita...1366Anticipate and Learn: Unleashing Idle-Time Compute i...While AI agents demonstrate remarkable capabilities in reasoning and tool use, they remain fundamentally reactive: they compute responses only after exp...1367Can LLMs Introspect? A Reality CheckCan large language models detect and report their own internal states? A number of studies have argued that the answer to this question is yes. We argue...1368Channel-wise Vector QuantizationWe present Channel-wise Vector Quantization (CVQ), a novel image tokenization paradigm that replaces patch-wise tokens with channel-wise tokens. Unlike...1369Claw-Anything: Benchmarking Always-On Personal Assis...Large language model agents are increasingly envisioned as always-on personal assistants with access to anything relevant in the user's digital world. Y...1370CollectionLoRA: Collecting 50 Effects in 1 LoRA via...Customized image editing aims to equip pre-trained diffusion models with specific visual effects using limited paired data, typically via Low-Rank Adapt...1371Confidence and Calibration of Activation Oracles for...Activation oracles aim to make the activations of other models legible to humans and yield promising results compared to white-box interpretability tech...1372ControlLight: Towards Controllable, Consistent, and...Existing deep learning-based low-light enhancement methods are typically trained on limited datasets with single enhancement targets, which restricts th...1373CroCo: Cross-Lingual Contrastive Preference Tuning o...Prior work establishes that controlled contrastiveness between self-generated responses from large language models, set via reward scores, improves down...1374CUA-Gym: Scaling Verifiable Training Environments an...Reinforcement learning with verifiable rewards (RLVR) has driven breakthroughs in domains such as math, tool-use, and software engineering, yet its exte...1375D^2-Monitor: Dynamic Safety Monitoring for Diffusion...Despite the emergence of diffusion large language models (D-LLMs) as an alternative to autoregressive large language models (AR-LLMs), safety monitoring...1376Does Seeing More Mean Knowing More? Mono-Anchored Ad...Visual reasoning through reinforcement learning with verifiable rewards (RLVR) has achieved remarkable progress. However, when dealing with multi-source...1377DVAO: Dynamic Variance-adaptive Advantage Optimizati...Reinforcement Learning has become a standard paradigm for aligning Large Language Models with human intent and task requirements. While Group Relative P...1378Everything at Every Scale: Scale-Invariant Diffusion...Creating images from noise is image generation; reconstructing fine details from coarse inputs is super-resolution. Despite their practical differences,...1379Fast-dDrive: Efficient Block-Diffusion VLM for Auton...End-to-end autonomous driving via Vision-Language-Action (VLA) models demands a precarious balance between high-fidelity trajectory planning and efficie...1380From Model Scaling to System Scaling: Scaling the Ha...This paper studies the next major bottleneck in agentic AI as system scaling, not only model scaling: the design of auditable, persistent, modular, and...1381Geometry-Aware Representation Denoising for Robust M...Multi-view 3D reconstruction has achieved remarkable progress with the advent of feed-forward 3D reconstruction models. However, these models are typica...1382Helix4D: Complex 4D Mesh GenerationCurrent video-to-4D methods struggle with complex topology changes, transparent materials, thin structures, and inner surfaces. We present Helix4D, a dy...1383InstructSAM: Segment Any Instance with Any InstructionsIn this paper, we introduce InstructSAM, a unified and streamlined framework designed for multi-instance segmentation under arbitrary instructions. We f...1384Language Models Need SleepTransformer-based large language models are increasingly used for long-horizon tasks; however, their attention mechanism scales poorly with context leng...1385LLaVA-OneVision-2: Towards Next-Generation Perceptua...We introduce LLaVA-OneVision-2 (LLaVA-OV-2), the most capable vision-language model in the LLaVA-OneVision series to date, achieving superior performanc...1386LongAV-Compass: Towards Unified Evaluation of Minute...Audio-visual generation is rapidly advancing from short clips to minute-long content, while existing evaluation protocols remain largely confined to sho...1387MetaphorVU: Towards Metaphorical Video UnderstandingMetaphorical videos are prevalent across various real-world scenarios to convey complex ideas, and understanding them typically requires high-order cogn...1388MobileGym: A Verifiable and Highly Parallel Simulati...We present MobileGym, a browser-hosted, lightweight, fully controllable environment for everyday mobile use, targeting interaction fidelity without repl...1389NSF-SciFy: Mining the NSF Awards Database for Scient...We introduce NSF-SciFy, a comprehensive dataset of scientific claims and investigation proposals extracted from National Science Foundation award abstra...1390On-Policy Adversarial Flow Distillation for Autoregr...Autoregressive video generators are attractive for streaming, long-horizon, and interactive applications, but distilling strong black-box teachers into...1391Pantheon360: Taming Digital Twin Generation via 3D-A...Generating complete digital twins from videos requires precise camera control, global scene coverage, and strict spatial-temporal consistency constraint...1392Personalize-then-Store: Benchmarking and Learning Pe...Existing large language model (LLM) based memory systems apply universal, static policies that overlook a fundamental reality: the contexts that are wor...1393Reinforcing Few-step Generators via Reward-Tilted Di...Recent advances in few-step diffusion distillation have enabled efficient image generation, yet aligning these models with human preferences remains cha...1394Rethinking VLM Representation for VLA InitializationVision-Language-Action (VLA) models widely adopt pretrained Vision-Language Models (VLMs) as policy backbones, yet it remains unclear what kind of pretr...1395ScientistOne: Towards Human-Level Autonomous Researc...Autonomous research agents produce competitive solutions and professional-looking manuscripts, yet their outputs contain verifiability failures undetect...1396SemBridge: Language Transfer in Sparse Encoders via...Sparse encoders offer high-precision retrieval by representing term importance within a vocabulary space, yet their English-centric structures pose a cr...1397Squeezing Capacity from Multimodal Large Language Mo...Subject-driven image generation aims to synthesize new images that preserve the identity of the given subject while following textual instructions. Exis...1398Toward Native Multimodal Modeling: A RoadmapMultimodal modeling represents a vital step from modality-agnostic reasoning toward world modeling. While early approaches predominantly rely on late-fu...1399Triplet-Block Diffusion RWKVCausal Transformer language models suffer from strictly sequential decoding and a quadratic per-step attention cost. While linear-time causal models and...1400TriSplat: Simulation-Ready Feed-Forward 3D Scene Rec...Sparse-view 3D reconstruction is increasingly addressed with feed-forward splatting networks that predict explicit primitives directly from images. Yet...1401Unified Panoramic Geometry Estimation via Multi-View...Geometry estimation from perspective images has greatly advanced, maturing to the point where off-the-shelf foundation models are able to reconstruct 3D...1402WBench: A Comprehensive Multi-turn Benchmark for Int...Interactive world models are advancing rapidly, yet existing benchmarks cover only part of the required competencies, leaving no unified standard for sy...1403Your Agents Are Aging Too: Agent Lifespan Engineerin...Long-lived AI agents are increasingly deployed as persistent operational systems, yet they are still evaluated like freshly initialized models. Day-one...1404AgensFlow: A Coordination-Policy Substrate for Multi...Multi-agent systems built on large language models (LLMs) require many coordination choices that are difficult to fix a priori: which skill protocol to...1405Alignment Tampering: How Reinforcement Learning from...Reinforcement Learning from Human Feedback (RLHF) is the standard method to align Large Language Models (LLMs) with human preferences. In this work, we...1406Balancing Fidelity and Diversity in Diffusion Models...We characterize the pre-softmax attention matrix QK^top in transformers as an associative memory matrix encoding pairwise associations between input fea...1407Beyond Final Answers: Auditing Trajectory-Level Hall...Large Language Models (LLMs) are increasingly deployed as autonomous agents that reason, use tools, and act over multiple steps. Yet most hallucination...1408Cross-scale Aligned Supervision for Training GANsModern GANs often introduce adversarial supervision on intermediate generator outputs and interpret the resulting multi-stage synthesis as coarse-to-fin...1409Efficient Agentic Reinforcement Learning with On-Pol...Agentic reinforcement learning (RL) has proven effective for training LLM-based agents with external tool-use capabilities. However, we identify that ag...1410Gemini Embedding 2: A Native Multimodal Embedding Mo...We introduce Gemini Embedding 2, a native multimodal embedding model that allows embedding video, audio, image, and text modalities in a unified represe...1411GE-Sim 2.0: A Roadmap Towards Comprehensive Closed-l...We introduce GE-Sim 2.0 (Genie Envisioner World Simulator 2.0), a closed-loop video world simulator for robotic manipulation. Building on the action-con...1412GradSentry: Gradient Spectral Entropy for Backdoor S...Fine-tuning Large Language Models with untrusted data exposes models to backdoor attacks, where poisoned samples cause targeted misbehavior. Existing sa...1413Guiding LLM Post-training Data Engineering with Mode...Model internals encode rich information about how a large language model (LLM) processes its training data; however, post-training data engineering larg...1414How and What to Imagine? Visual Thinking in Unified...Cross-view spatial reasoning remains a weak spot for vision-language models (VLMs): they often reason in language and lose the fine-grained geometry nee...1415Is Position Bias in Dense Retrievers Built In-or Lea...Dense retrievers exhibit positional bias, favoring documents whose query-relevant information appears near the beginning and degrading retrieval perform...1416JLT: Clean-Latent Prediction in Latent Diffusion Tra...Flow matching with clean-data prediction has shown that regressing the clean point can exploit low-dimensional structure more effectively than predictin...1417Learning to Act under Noise: Enhancing Agent Robustn...Recent advances in large language models (LLMs) have facilitated the widespread deployment of LLMs as interactive agents capable of reasoning, planning,...1418Less is More: Early Stopping Rollout for On-Policy D...On-policy distillation has recently emerged as a promising alternative to standard sequence-level imitation, training a student by scoring its own rollo...1419LocateAnything: Fast and High-Quality Vision-Languag...Vision-language models (VLMs) commonly formulate visual grounding and detection as a coordinate-token generation problem, serializing each 2D box into m...1420Lost in Sampling: Assessing Lexical Reachability in...Modern Large Language Models (LLMs) are often criticized for producing repetitive and homogeneous text, despite possessing vast latent vocabularies. Whi...1421MobileMoE: Scaling On-Device Mixture of ExpertsMixture-of-Experts (MoE) has become the de facto architecture for hundred-billion-parameter language models, yet its advantages at sub-billion scales fo...1422MRT: Masked Region Transformer for Layered Image Gen...Layered image generation and editing is a fundamental capability that enables layer-wise reuse, editing, and composition of generated visual content, an...1423MUSE-Autoskill: Self-Evolving Agents via Skill Creat...Large language model (LLM) agents rely on reusable skills to solve complex tasks. However, existing skill creation approaches treat skills as isolated a...1424Negligible in Size, Significant in Effect: On Scale...Normalization layers in modern large language models (LLMs) consist of a deterministic normalization operation and a learnable scale vector. While the n...1425Not All Disagreement Is Learnable: Token Teachabilit...On-policy distillation (OPD) trains a student on its own rollouts with token-level teacher supervision. Recent selective OPD methods exploit the non-uni...1426PANDO: Efficient Multimodal AI Agents via Online Ski...Recent advances in multimodal web agents often rely on increased inference-time computation, including rollout search, verifier passes, offline skill di...1427PEAM: Parametric Embodied Agent Memory through Contr...We present PEAM, a Parametric Embodied Agent Memory framework in Minecraft that transforms agent memory from inference-time retrieval into parameter-res...1428PRISM: Position-encoded Regressive Inverse Spectral...The inverse problem of multilayer thin-film optical coatings design represents a complex combinatorial-continuous optimization challenge. We present PRI...1429QUACK: Questioning, Understanding, and Auditing Comm...Social deduction games have become a popular testbed for probing reasoning, deception, coordination, and belief modeling in Large Language Model (LLM) a...1430Recursive Flow MatchingGenerative models have emerged as a powerful paradigm for solving physics systems and modeling complex spatiotemporal dynamics. However, achieving high...1431RT-Lynx: Putting the GEMM Sparsity In a Right Way fo...Diffusion Transformers (DiT) achieve strong performance in image generation but incur substantial inference costs. While prior work has reduced this cos...1432Share More, Search Less: Collaborative Parallel Thin...Test-Time Scaling (TTS) enhances the reasoning capabilities of large language models by allocating additional inference compute to explore the solution...1433SkillGrad: Optimizing Agent Skills Like Gradient Des...Agent skills provide a lightweight way to adapt LLM agents to specialized domains by storing reusable procedural knowledge in structured files. However,...1434SpatialBench: Is Your Spatial Foundation Model an Al...While spatial foundation models have demonstrated impressive performance on standard datasets, a critical question remains: are they truly all-round pla...1435The MiniMax-M2 Series: Mini Activations Unleashing M...We introduce the MiniMax-M2 series, a family of Mixture-of-Experts language models built around the principle that mini activations can unleash maximum...1436Verus-SpecGym: An Agentic Environment for Evaluating...AI coding agents are increasingly used to write real-world software, but ensuring that their outputs are correct remains a fundamental challenge. Formal...1437VitaBench 2.0: Evaluating Personalized and Proactive...Large language models (LLMs) have evolved into interactive agents that collaborate with users in real-world tasks. Effective collaboration in such setti...1438Agent Explorative Policy Optimization for Multimodal...Vision-language models with extended reasoning succeed on complex problems, but many real-world problems require external tools that internal reasoning...1439AI Research Agents Narrow Scientific ExplorationAI research agents can now generate research ideas, design experiments, run code, and draft papers, raising the possibility of large-scale AI-assisted s...1440AlphaTransit: Learning to Design City-scale Transit...Designing a transit network requires many sequential route extension decisions, but their quality is often visible only after the full network is assemb...1441AsyncTool: Evaluating the Asynchronous Function Call...Large language model (LLM)-based agents have shown strong capabilities in using external tools to solve complex tasks. However, existing evaluations oft...1442AutoScientists: Self-Organizing Agent Teams for Long...Scientific research proceeds through iterative cycles of hypothesis generation, experiment design, execution, and revision. AI agents can automate parts...1443Category-Level 3D Correspondence in Camera Space via...Understanding 3D objects from images is fundamental to robotics and AR/VR applications. While recent work has made progress in category-level pose estim...1444Comprehensive Benchmarking of Long-Form Speech Gener...Recent advances in speech generation have enabled high-fidelity synthesis, yet systematic evaluation of models under long-context conditions remains lar...1445CubePart: An Open-Vocabulary Part-Controllable 3D Ge...Interactive 3D assets used in games and simulation are typically decomposed into specific semantic parts to support animation, physics, and scripted beh...1446DEMON: Diffusion Engine for Musical Orchestrated NoiseWe present DEMON, a real-time diffusion engine that makes the denoising process playable as a live musical instrument: a control surface both broad (man...1447DenoiseRL: Bootstrapping Reasoning Models to Recover...Reinforcement learning has become a central paradigm for advancing reasoning in large language models, yet most existing methods still depend on stronge...1448Efficient and Scalable Provenance Tracking for LLM-G...Large language models (LLMs) for code completion and generation are increasingly used in software development, yet they may reproduce training examples...1449ESC-Skills: Discovering and Self-Evolving Skills for...Existing emotional support conversation (ESC) systems mainly rely on end-to-end response generation or coarse strategy supervision, offering limited int...1450FRAPPE: Full Input, Residual Output Autoencoding wit...Media compression standards have reached a plateau in terms of the rate-distortion-complexity trade-off, limiting the ability to offload expensive AI pe...1451From Pixels to Words -- Towards Native One-Vision Mo...Current vision-language models (VLMs) typically stitch together separate image encoders and language decoders via multi-stage alignment, a modular frame...1452Gamma-World: Generative Multi-Agent World Modeling B...World models for interactive video generation have largely focused on single-agent settings, where future observations are generated from a single contr...1453GEM: Generative Supervision Helps Embodied IntelligenceEmbodied Vision-Language Models (VLMs) have demonstrated impressive performance and generalization in robotics, particularly within Vision-Language-Acti...1454GUI-CIDER: Mid-training GUI Agents via Causal Intern...Despite the rapid progress of multimodal large language models in building Graphical User Interface (GUI) agents, their real-world task completion is fu...1455HRBench: Benchmarking and Understanding Thinking-Mod...Hybrid-reasoning large language models (LLMs) expose explicit controls over reasoning effort, allowing users or systems to trade off answer quality agai...1456Joint Training of Multi-Token Prediction in Reinforc...Reinforcement Learning from Verifiable Rewards (RLVR) has emerged as the standard paradigm for improving reasoning capability of large language models,...1457LACUNA: Safe Agents as Recursive Program HolesLLM agents increasingly act by writing code, yet a split persists between the runtime that drives the agent and the code the model writes. The runtime o...1458Learn from Weaknesses: Automated Domain Specializati...Computer-use agents (CUAs) have recently made substantial progress, but deploying a separate large expert for each software domain remains expensive. Sm...1459LiveBrowseComp: Are Search Agents Searching, or Just...Are LLM-based search agents genuinely searching, or using the web to verify what they already know? We study this question on BrowseComp with three diag...1460Long Live The Balance: Information Bottleneck Driven...Recent advances in online reinforcement learning (RL) for large language models (LLMs) have demonstrated promising performance in complex reasoning task...1461MemTrace: Tracing and Attributing Errors in Large La...Memory is essential for enabling large language models to support long-horizon reasoning, yet existing memory systems remain unreliable and difficult to...1462Models That Know How Evaluations Are Designed Score...The validity of AI safety evaluations depends on models behaving consistently across controlled and deployment settings. Prior work has identified test-...1463OmniVerifier-M1: Multimodal Meta-Verifier with Expli...Visual outcomes are increasingly central to multimodal large language models, making reliable and fine-grained verification essential for scaling genera...1464OR-Space: A Full-Lifecycle Workspace Benchmark for I...Large language model (LLM) agents are increasingly used to assist with operations research (OR) modeling, yet existing OR-oriented benchmarks often redu...1465OSP-Next: Efficient High-Quality Video Generation wi...Diffusion Transformers achieve strong video generation quality, but the quadratic cost of full attention limits efficiency. We introduce OSP-Next, an ef...1466Parallax: Parameterized Local Linear Attention for L...Large Language Models (LLMs) have become the central paradigm in artificial intelligence, yet the core computational primitive of attention has remained...1467PEFT-Arena: Understanding Parameter-Efficient Finetu...Parameter-efficient finetuning (PEFT) has become the standard approach for adapting large language models, yet evaluations largely emphasize downstream...1468PRISM: A Multi-Dimensional Benchmark for Evaluating...The rapid growth in submissions to machine learning venues has strained the scientific peer-review system and intensified interest in LLM-based automate...1469ProRL: Effective Reinforcement Learning for Proactiv...Proactive Recommender Systems (PRSs) aim to guide user preference shift toward target items by generating paths of intermediate recommendations. Reinfor...1470ResearchMath-14K: Scaling Research-Level Mathematics...The frontier of mathematics is defined by problems whose solutions are not yet known, yet it remains unclear whether language models can meaningfully en...1471Rethinking Memory as Continuously Evolving ConnectivityExisting memory-augmented LLM agents often treat memory as a static repository with pre-defined representations and fixed retrieval pipelines, which is...1472RUBRIC-ARROW: Alternating Pointwise Rubric Reward Mo...Pointwise reward modeling offers critical signals for LLM post-training, yet struggles with absolute scoring in subjective, non-verifiable settings. Rub...1473Self-Improving Language Models with Bidirectional Ev...Search has been proposed as an effective method for self-improving language models and agentic systems, both for post-training sample generation and for...1474Skill0.5: Joint Skill Internalization and Utilizatio...Equipping large language models with explicit skills has emerged as a promising paradigm for enabling autonomous agents to solve complex tasks. Agent sk...1475SmartDirector: Keyframe-Conditioned Cinematic Video...The narrative quality of a video fundamentally determines its perceptual value. Although existing video generation methods can produce visually appealin...1476The Fragility of Chain-of-Thought Monitoring Across...Chain-of-thought (CoT) monitoring has been proposed as a promising safety mechanism for detecting misaligned behavior in large language models. However,...1477VibeSearchBench: Benchmarking Long-horizon Proactive...LLM-based agents score well on search benchmarks, yet real users consistently find results unsatisfying, revealing a persistent evaluation-experience ga...1478When Confidence Misleads: Suffix Anchoring and Ancho...Diffusion language models decode text by iteratively denoising masked token sequences, making the choice of which positions to decode a central inferenc...1479AdaState: Self-Evolving Anchors for Streaming Video...Autoregressive video diffusion models generate streaming video by producing frames sequentially, conditioning each chunk on previously generated content...1480AgentDoG 1.5: A Lightweight and Scalable Alignment F...Modern open-world agents such as OpenClaw exhibit powerful cross-environment execution capabilities yet introduce broad new safety risk sources. Meanwhi...1481CausaLab: A Scalable Environment for Interactive Cau...We introduce CausaLab, a scalable environment for evaluating interactive causal discovery by LLM agents. Unlike prior evaluations, CausaLab evaluates bo...1482ChildVox: A Speech, Audio, and Large Audio-Language...We present ChildVox, a novel benchmark for characterizing the diverse acoustic signals through which children communicate. Specifically, ChildVox follow...1483CoHyDE: Iterative Co-Training of LLM Rewriter & Dens...Tool retrieval over large API catalogs is a core bottleneck for LLM agents: user queries arrive in colloquial, often underspecified language, while the...1484Colored Noise Diffusion SamplingDiffusion models achieve state-of-the-art image synthesis, with their generative trajectories fundamentally exhibiting a spectral bias, resolving low-fr...1485DynaFLIP: Rethinking Robotics Perception via Tri-Mod...Robot manipulation critically depends on perception that preserves the action-relevant aspects of a scene. Yet most robot learning pipelines are built u...1486EarlyTom: Early Token Compression Completes Fast Vid...Video large language models (Video-LLMs) have demonstrated strong capabilities in video understanding tasks. However, their practical deployment is stil...1487Exploring Autonomous Agentic Data Engineering for Mo...Large Language Models (LLMs) have demonstrated strong performance on general tasks, while often struggling to adapt to specialized domains without high-...1488GDSD: Reinforcement Learning as Guided Denoiser Self...Reinforcement learning (RL) can be used to improve the policy (denoiser) of diffusion large language models (dLLMs), while being hindered by the intract...1489GenClaw: Code-Driven Agentic Image GenerationImage generation models have evolved from text-conditioned pixel synthesis toward multimodal agents endowed with visual comprehension and tool invocatio...1490Geometry Matters: 3D Foundation Priors for Learning...Foundation features from self-supervised vision models and text-to-image diffusion models have proven effective for semantic correspondence estimation....1491GrepSeek: Training Search Agents for Direct Corpus I...Large Language Model (LLM) search agents have shown strong promise for knowledge-intensive language tasks through multiple rounds of reasoning and infor...1492How LoRA Remembers? A Parametric Memory Law for LLM...Large Language Models (LLMs) must continuously learn and update knowledge to remain effective in dynamic real-world environments. While Low-Rank Adaptat...1493LaRA: Layer-wise Representation Analysis for Detecti...Reinforcement learning (RL) post-training has shown to improve reasoning in large language models (LLMs). However, there has been little exploration on...1494LiteCoder-Terminal: Scaling Long-Horizon Terminal En...Mastering terminal environments requires language agents capable of multi-step planning, feedback-grounded execution, and dynamic state adaptation. Howe...1495LoMo: Local Modality Substitution for Deeper Vision-...Vision-Language Models (VLMs) have achieved substantial progress across a wide range of understanding and reasoning tasks, driven by large-scale image-t...1496LongDS-Bench: On the Failure of Long-Horizon Agentic...Real-world data analysis is inherently iterative, yet existing benchmarks mostly evaluate isolated or short interactive tasks, leaving agents' ability t...1497MAAT: Multi-phase Adapter-Aware Targeted UnlearningMachine unlearning evaluation is structurally skewed: Why-type questions, which probe causal and relational knowledge, comprise less than 0.06% of Count...1498minWM: A Full-Stack Open-Source Framework for Real-T...Recent video diffusion foundation models have achieved remarkable progress in high-quality video generation, yet turning them into real-time interactive...1499Native Audio-Visual Alignment for GenerationJoint audio-video generation aims to synthesize temporally synchronized and semantically coherent visual-acoustic content. However, existing open-source...1500NeuROK: Generative 4D Neural Object KinematicsData-driven approaches have revolutionized 3D vision, enabling transformers to effectively reconstruct and generate static 3D objects. However, generati...1501OmniRetrieval: Unified Retrieval across Heterogeneou...Real-world information needs require access to structurally diverse knowledge sources, from unstructured text and relational tables to knowledge graphs...1502One Click per Cell Type Suffices: Training-free Grou...Cell instance segmentation models trained on cell-specific datasets suffer severe performance drops on out-of-distribution cell types, while interactive...1503OpenSkillEval: Automatically Auditing the Open Skill...Skills, i.e., structured workflow instructions distilled for large language models (LLMs), are becoming an increasingly important mechanism for improvin...1504PhyGenHOI: Physically-Aware 4D Generation of Dynamic...We address the task of generating physically accurate and visually faithful 4D Human-Object Interaction (HOI). Given a static 3D human and target object...1505Qwen-VLA: Unifying Vision-Language-Action Modeling a...Embodied intelligence is often studied through specialized models for individual tasks such as manipulation or navigation, resulting in fragmented capab...1506Recovering Policy-Induced Errors: Benchmarking and T...While GUI agents have advanced rapidly, they often lack the robustness to recover from their own errors, hindering real-world deployment. To bridge this...1507REPOT: Recoverable Program-of-Thought via Checkpoint...One-shot Program-of-Thought (PoT) emits a Python program that prints a primitive-action plan; a single invalid action silently invalidates the trajector...1508SAAS: Self-Aware Reinforcement Learning for Over-Sea...Agentic search enables LLMs to solve complex multi-hop questions through iterative reasoning and external search. Despite the effectiveness, these syste...1509SANA-Streaming: Real-time Streaming Video Editing wi...Real-time streaming video-to-video editing (V2V) is critical for interactive applications such as live broadcasting and gaming, yet it remains a formida...1510Seeing Isn't Knowing: Do VLMs Know When Not to Answe...Spatial reasoning is a fundamental capability for vision-language models (VLMs) deployed in real-world environments. However, visual observations are in...1511SoundnessBench: Can Your AI Scientist Really Tell Go...Autonomous AI research agents aim to accelerate scientific discovery by automating the research pipeline, from hypothesis generation to peer review. How...1512The Good, the Bad, and the Ugly of Markov Boundary f...Under standard graphical assumptions, the Markov boundary of a target variable is the smallest set of features that renders every other feature redundan...1513Thinking Before Constraining: A Unified Decoding Fra...Natural generation allows Large Language Models (LLMs) to produce free-form responses with rich reasoning, yet the lack of structure makes outputs diffi...1514Token-Level Generalization in LoRA Adapter Backdoors...We show that LoRA adapters, the dominant distribution format for fine-tuned LLMs, can be reliably backdoored through training data poisoning while prese...1515Towards Verifiable Multimodal Deep Research: A Multi...Large Language Models (LLMs) have advanced autonomous agents from deep search, which retrieves concise factual answers, to deep research, which synthesi...1516UI-KOBE: Knowledge-Oriented Behavior Exploration for...Recent advances in mobile GUI agents have shown strong potential for automating mobile tasks, but most effective systems still depend on large vision-la...1517UniSteer: Text-Guided Flow Matching in Activation Sp...Activation-based control steers large language models (LLMs) by intervening on their internal representations during inference, and has emerged as an ef...1518Verifiable Rewards Beyond Math and Code: Lightweight...Applying reinforcement learning to improve factual accuracy in knowledge-intensive question answering faces a reward design dilemma. Response-level rewa...1519VisualThink-VLA: Visual Intermediate Reasoning for E...Recent work has begun to equip vision-language-action (VLA) policies with explicit intermediate reasoning. In embodied control, however, textual chain-o...1520VLM3: Vision Language Models Are Native 3D LearnersVision Language Models (VLMs) enable a unified model to solve various vision tasks through prompting. They have shown promising performance in semantic...1521When Cloud Agents Meet Device Agents: Lessons from H...The design space of agentic AI inference spans two extremes: frontier large language models (LLMs), typically hosted in the cloud and offering strong pe...1522When Should Models Change Their Minds? Contextual Be...Long-horizon interactions require language models to manage accumulating information: when to update their state, when to preserve their state, and what...1523Why Far Looks Up: Probing Spatial Representation in...Vision-language models (VLMs) achieve strong performance on spatial reasoning benchmarks, yet it remains unclear whether this reflects structured 3D und...1524Why Larger Models Learn More: Effects of Capacity, I...Larger models learn tasks smaller models do not. What drives this phenomenon? We develop a simple phenomenological argument that power-law scaling alrea...1525WorldMemArena: Evaluating Multimodal Agent Memory Th...Multimodal large language models are increasingly deployed as long-horizon agents, where memory must do more than recall: it must track an evolving worl...1526Xetrieval: Mechanistically Explaining Dense RetrievalExplaining why dense retrievers assign high relevance scores remains challenging because retrieval decisions are made through opaque high-dimensional em...1527YoCausal: How Far is Video Generation from World Mod...As video diffusion models (VDMs) advance toward world models, a key question arises: do they truly understand causality, or merely overfit to statistica...1528A Tight Theory of Error Feedback Algorithms in Distr...Communication costs are a major bottleneck in distributed learning and first-order optimization. A common approach to alleviate this issue is to compres...1529Approximation and learning of anisotropic and mixed...This paper studies how efficiently deep ReLU neural networks can approximate and learn smooth functions. When the error is measured in $L^p([0,1]^d)$ no...1530Are Full Rollouts Necessary for On-Policy Distillation?On-policy distillation (OPD) provides dense teacher feedback along rollouts generated by the student and has emerged as a promising post-training paradi...1531Assign and Add: A Mechanistic Study of Compositional...Large language models are able to compose skills in order to perform complex tasks, many of which might not have been seen during training. The details...1532Automated Prediction of Postoperative Pancreatic Fis...Postoperative pancreatic fistula (POPF) is a serious complication after pancreatic resection, increasing morbidity, hospital stay, and healthcare costs....1533BenHalluEval: A Multi-Task Hallucination Evaluation...Despite Bengali being the sixth most spoken language in the world, no prior work has systematically evaluated hallucination in large language models (LL...1534Beyond Additive Decompositions: Interpretability Thr...Interpretable machine learning requires models that are accurate and structurally faithful to the data.Existing explainability methods rely heavily on a...1535Chem-PerturBridge: a harmonized compendium of small...Large perturbation models require training data encompassing chemical, cellular, and assay diversity. Current transcriptomic resources for small-molecul...1536Choosing the Lens: Strategic Perspective Activation...The same arguments often need to be evaluated under different external regimes. An agent with influence over the regime has a strategic lever that stand...1537CoFiDA-M: Concept-Aware Feature Modulation for Cross...Models for AI-based skin cancer screening suffer a severe performance drop when shifting from expert dermoscopic (source) images to consumer-grade clini...1538COLLEAGUE.SKILL: Automated AI Skill Generation via E...LLM agents are increasingly expected not only to complete isolated tasks, but also to carry bounded representations of human expertise, judgment, and in...1539Consolidating Rewarded Perturbations for LLM Post-Tr...Post-training of language models is commonly framed as a sample-score-update loop implemented by gradient descent. A recent line of work, exemplified by...1540Convergence of Two-Timescale Markovian Stochastic Ap...This work studies the convergence of two-timescale stochastic approximations (SA), a class of iterative algorithms that update two sets of parameters in...1541Correcting Split Selection in Online Decision Trees...Bagging-based ensembles, most notably Adaptive Random Forests, are among the strongest performers for learning from data streams. A common denominator a...1542Count AnythingObject counting remains fragmented across domain-specific datasets and task formulations, despite rapid progress in generalist vision models. Existing c...1543DecMem: Towards Minute-Long Consistent World Generat...Recent advances in video generative models have promoted rapid progress in controllable world models. However, maintaining fine-grained spatio-temporal...1544Disagreeing Rationales: Rethinking Classification an...Human disagreement is ubiquitous and well-known in labeling. However, variation in explanations, captured through token-level human rationales, remains...1545Discovering Thermodynamically Admissible Dissipation...Constitutive laws for inelastic materials must satisfy strict thermodynamic admissibility requirements, yet current data-driven approaches sacrifice int...1546dMoE: dLLMs with Learnable Block ExpertsDiffusion Large Language Models (dLLMs) have recently emerged as a promising alternative to autoregressive models, offering competitive performance whil...1547DRIFT: Decoupled Rollouts and Importance-Weighted Fi...Large language models are increasingly deployed in multi-turn interactive settings where users or environments can iteratively provide lightweight feedb...1548Effective Biological Representation Learning by Mask...RNA sequencing produces rich and diverse datasets of gene expression, offering compelling insights into cellular state and function that have many appli...1549EGOSTREAM: A Diagnostic Benchmark for Streaming Epis...Continuous episodic memory is a core capability for autonomous agents operating in dynamic, real-world environments, yet current streaming video benchma...1550Emergent Languages in Populations of Language Model...Monitoring autonomous language model agents currently relies mostly on surface behavior. But what happens when agent populations invent new languages wi...1551Entropic Projection Alignment: Estimating, Explainin...We propose a unified framework for addressing three key challenges of distribution shift: (1) estimating a model's performance on an unlabeled target do...1552Evaluating Factual Density in Multi-Source RAG: A St...Retrieval-Augmented Generation (RAG) is the current industry standard for grounding AI in real-world facts. Traditional retrieval methods rely on keywor...1553Feature-Optimized Vision for Adaptive 3D Scene Recon...Three-dimensional scene reconstruction depends on local image evidence that is both visually discriminative and geometrically useful. Fixed feature thre...1554Fraud Type Decomposition and the Observation-Mechani...Fraud detection in payment networks relies on labels generated through heterogeneous and imperfect observation processes, yet existing approaches treat...1555From Prompt Injection to Persistent Control: Defendi...LLM agents are evolving from conversational chatbots to operational tools in real-world workspaces. In local agentic harnesses, an LLM can read and writ...1556Function2Scene: 3D Indoor Scene Layout from Function...Most text-driven 3D indoor scene synthesis methods generate rooms from object-centric prompts, asking what furniture should be placed rather than how th...1557Functional Attention: From Pairwise Affinities to Fu...Learning mappings between infinite-dimensional function spaces, or operator learning, is essential for many machine learning applications. Although tran...1558GGT-100K: Generative Ground Truth for Generalizable...Real-world image restoration (IR) is bottlenecked by the scarcity of high-quality paired training data. Synthetic datasets are abundant but often fail t...1559Giving Sensors a Voice: Multimodal JEPA for Semantic...Transformer-based architectures have advanced sequence modeling in language and vision, yet general-purpose representation learning for heterogeneous mu...1560Guidance Contrastive Token Credit Assignment for Dis...Group-advantage-based reinforcement learning methods, such as GRPO and DAPO, have demonstrated strong performance across diverse domains, including math...1561Hide-and-Seek in Trajectories: Discovering Failure S...Vision-Language-Action (VLA) models enable robots to follow natural language instructions and generalize across diverse tasks, but they remain vulnerabl...1562How can embedding models bind concepts?Humans easily determine which color belongs to which shape in multi-object scenes, an ability known as concept binding. Vision-language embedding models...1563If LLMs Have Human-Like Attributes, Then So Does Age...Much research has been carried out on large language models (LLMs) and LLM-powered agentic workflows. However, many works within the field state emergen...1564iVGR: Internalizing Visually Grounded Reasoning for...While visually grounded Chain-of-Thought (CoT) has emerged as a promising paradigm to enhance fine-grained perception in multimodal large language model...1565Joint Multi-Camera LiDAR Extrinsic Calibration via L...Most learning-based camera-LiDAR calibration methods treat each camera-LiDAR pair independently, ignoring the rigid geometric coupling in multi-camera p...1566KLIP: localized distribution shift detection via KL-...Diffusion models have shown promising performance as data-driven priors for computational imaging, as well as some capacity to detect out-of-distributio...1567Language Models Learn Constructional Semantics, Not...Grasping the semantics of rare constructions (form-meaning pairings) has been shown to be a challenging problem that has currently only been solved by t...1568Learning Global Motion with Compact Gaussians for Fe...Dynamic scene reconstruction from monocular video remains a fundamental challenge in computer vision. Existing feed-forward methods predict 3D Gaussians...1569Light Interaction: Training-Free Inference Accelerat...Interactive video world models generate video chunk by chunk in response to user-controlled camera movements, enabling applications such as real-time ga...1570Linear Scaling Video VLMs for Long Video UnderstandingVideo vision-language models (VLMs) are increasingly used in long-horizon and streaming settings, yet most video encoders still rely on spatiotemporal s...1571Log-Ratio Propagation on the Simplex: A Theory of Ce...Compositional data must be analysed through log-ratios: scale invariance, the defining axiom of the field, leaves no alternative. The centred log-ratio...1572LongTraceRL: Learning Long-Context Reasoning from Se...Long-context reasoning remains a central challenge for large language models, which often fail to locate and integrate key information in extensive dist...1573Lumos-Nexus: Efficient Frequency Bridging with Homog...Connector-based video unified models have demonstrated strong capability in instruction-grounded video synthesis, but integrating a large high-fidelity...1574Mellum2 Technical ReportWe present Mellum 2, an open-weight 12B-parameter Mixture-of-Experts (MoE) language model with 2.5B active parameters per token. Mellum 2 is a general-p...1575Memory by Design: Probabilistic Sequence LayersWe introduce the design-model framework: a way to derive efficient recurrent sequence maps from explicit assumptions about memory. A design model writes...1576Model-Agnostic Signal Discovery with Machine Learnin...Searches for new phenomena in complex scientific data are predominantly model-dependent, optimized for specific hypotheses, and therefore limited in the...1577Non-Asymptotic Convergence of Stochastic Iterative A...We survey Lyapunov-based techniques for the finite-time analysis of stochastic iterative algorithms, also known as stochastic approximation (SA) algorit...1578nuReasoning: A Reasoning-Centric Dataset and Benchma...Reasoning is essential for autonomous driving (AD) in long-tail scenarios, where vehicles must apply commonsense knowledge, understand spatial relations...1579On public and private binary classification with met...We consider the problem of binary classification in a framework where the predictor $X$ takes values in an arbitrary separable metric space $\mathcal X$...1580On the Relationship Between Activation Outliers and...Sparse autoencoders (SAEs) decompose neural network activations into interpretable features, but many learned features never activate, a problem called...1581PEEK: Picking Essential frames via Efficient Knowled...Video-language models can process only a limited number of frames, making frame selection a key bottleneck for efficient video captioning. Most captioni...1582Positional versus Symbolic Attention Heads: Learning...Transformer-based language models are widespread in today's society. As such, understanding the mechanisms by which they solve structured tasks and pred...1583Preference-Aware Rubric Learning for Personalized Ev...As Large Language Models (LLMs) evolve from general-purpose assistants to user-centric agents, personalization has become central to aligning model beha...1584RayDer: Scalable Self-Supervised Novel View Synthesi...Self-supervised novel view synthesis (NVS) remains challenging to scale, despite the abundance of video data, largely due to the brittleness of training...1585Recognizing Co-Speech Gestures in-the-WildWhile humans naturally gesture during speech, only a sparse subset of these movements are visually depictive and semantically linked to specific spoken...1586Reliable Multilingual Orthopedic Decision Support fr...Multilingual orthopedic decision support remains challenging in low-resource healthcare settings, where clinical narratives contain specialized terminol...1587Representation Forcing for Bottleneck-Free Unified M...Unified multimodal models (UMMs) aim to handle perception and generation in a single model. Yet existing UMMs still rely on a frozen, separately pretrai...1588SCOPE: Self-Play via Co-Evolving Policies for Open-E...Self-play can train language models without external supervision. However, existing methods require rule-checkable answers, leaving open-ended tasks dep...1589Semantic Triplet Restoration: A Novel Protocol for H...Table question answering requires models to recover semantic relations encoded implicitly by two-dimensional layout, merged cells, and hierarchical head...1590Separating Secrets from Placeholders: A Hybrid CNN-C...Credential leakage in public source code repositories poses a critical security threat, with over 23.8 million secrets exposed in 2024 alone. Existing d...1591Skill Reuse as Compression in Agentic RLLarge language model agents trained with reinforcement learning (RL) often learn brittle, task-specific shortcuts. We hypothesize that agents generalize...1592SMART: SMPLest-X Mesh Adaptation and RAFT Tracking f...We present our approach to the FIFA Skeletal Tracking Challenge 2026, which requires estimating 3D world-space poses of soccer players from broadcast vi...1593SOCO: Benchmarking Semantic Object Correspondence in...Measuring structured object understanding in vision foundation models remains challenging due to inconsistent evaluation protocols and limited part-leve...1594SPECTRA: Synthetic IR Test Collections with Relevanc...Scalable information retrieval testing needs corpora that are large enough to stress index construction, ranking latency, query routing, and evaluation...1595Stateful Online Monitoring Catches Distributed Agent...Language models can find thousands of severe software vulnerabilities, and agents are increasingly being misused for cyberattacks. To avoid detection, a...1596SurGe: Improved Surface Geometry in Point MapsRecent feedforward 3D reconstruction methods predict point maps and estimate global 3D geometry remarkably well. However, their predictions still exhibi...1597SwanVoice: Expressive Long-Form Zero-Shot Speech Syn...Zero-shot text-to-speech (TTS) has improved substantially for single-speaker synthesis, yet expressive long-form multi-speaker dialogue remains difficul...1598Task-Focused Memorization for Multimodal AgentsLong-term memory is essential for multimodal agents to build coherent experience, accumulate world knowledge, and achieve continual learning. However, c...1599The Dynamic-Probabilistic Consistency Gap in Chaotic...Dynamical systems reconstruction (DSR) aims to learn surrogate models that capture the dynamics underlying time-series data. Reliably deploying these su...1600The Flip Side of RLHF: On-Policy Feedback for Reward...Building strong reward models (RMs) for language model alignment is bottlenecked by the cost and difficulty of acquiring diverse and reliable preference...1601Towards Streaming Synchronized Spatial Audio Generat...Real-time and accurate spatial audio generation is pivotal for delivering an immersive experience. However, existing spatial audio synthesis technologie...1602Trust-Region Behavior Blending for On-Policy Distill...On-policy distillation (OPD) trains a student on prefixes sampled from its own policy while matching a stronger teacher. This addresses the prefix misma...1603TunerDiT: Training-free Progressive Steering of Diff...Text-to-video (T2V) generation faces challenging questions when generating videos with long horizons containing multiple events. Inspired by the intrins...1604UniAudio-Token: Empowering Semantic Speech Tokenizer...Semantic speech tokenizers have become a widely used interface for Audio-LLMs, owing to their compact single-codebook design and strong linguistic align...1605Value Functions as Supermartingale CertificatesCertification methods for stochastic systems provide sufficient proof rules, based on real-valued supermartingale certificates, to determine the almost-...1606Vision-Language Models Suppress Female Representatio...Alignment teaches vision-language models (VLMs) to avoid expressing demographic biases, and when gender is clearly visible they largely succeed. Far les...1607What Am I Missing? Question-Answering as Hidden Stat...Test-time reasoning has become a significant field of study since the introduction of chain-of-thought reasoning in large language models (LLMs). Howeve...1608What Gets Unmasked First? Trajectory Analysis of Dif...We present the first systematic study of masked diffusion language models (MDLMs) for graph-to-text generation. We analyze MDLM generation trajectories...1609When Are Multimodal Predictions Biologically Support...Multimodal models in oncology can produce accurate predictions, but accurate prediction does not reveal whether the model has learned biology that is sh...1610Why Linear Recurrent Memory Works in Partially Obser...The family of linear recurrent neural networks has shown strong performance as recurrent memory units in partially observable reinforcement learning. We...1611ReAct - Reasoning + ActingEngineering breakdown of the ReAct paper (Yao et al., 2022) - the foundation of every AI agent built today. Plain English, production viability rating, implementation notes.
