All papers
Every paper that carries a page of its own here. The Papers page leads with what is trending and lets you filter the recent catalog; this is the plain index of the rest.
Model-Based Online Decision Making via Generative Trajectory Planning2026ProAS: A Progress-Aware Advantage Shaping Framework for Multi-Turn Agentic Reinforcement Learning2026Jeva: System One Models from First Principles2026AesCode: Aesthetic Code Generation with Decoupled Cross-Modal Rewards2026FlashMask-3: Efficient and Expressive Mask-Aware Distributed Attention2026Depth Memory2026Correction Training for Diffusion Models2026TokenEconomizer: Experience-Driven Black-Box Cost Control for Coding Agents2026Self-Evolving Coding Agents via Harness-Policy Optimization2026Execution Before Memory: Diagnosing Memory Failures in Video World Models2026DLite: History-free Speculative Decoding for Agentic AI2026DiscoveryRSI: Recursively Learning to Discover Valuable Hypotheses2026DeLoop: Repeated Actions as Self-Supervision for Long-Horizon Agents2026DeltaStreamer: Towards Human-Like Agentic Streaming Video Understanding2026MorphAct: Robot Policies Morphing with the Worlds They Act In2026Doc2Harness: Benchmarking LLMs on Long-Horizon Harness Engineering of Coding Agents2026Player-1: Learning to Play from Synthetic Experience2026Continuum Attention2026RT-J: Large-Scale Pretraining of Relational Transformers for Context-Efficient Predictions2026ReTerm: Recovery Training for Self-Improving Terminal Agents2026Co-Evolving Acting and Forecasting for Language Agents2026Amortizing Sequential Experimental Design in Reasoning LLMs2026MAI-Gym: An Open All-in-One Platform for GUI Agent Evaluation and RL Training across Mobile and Computer Environments2026Sci$Π$vot: Benchmarking Principal-Investigator Judgment at Pivotal Moments of Scientific AutoResearch Agents2026JevAgentHarness: Benchmarking, Deploying and Training Decision Models That Share Control with LLMs2026Can Visual Generation Expand What Unified Multimodal Models Know?2026Data World: Scaling Data Environments from Scratch2026RSIFlow: Adaptive Self-Improvement Workflows with an Evolving Meta-Agent2026Memory as Weights: Internalizing Long-Term History for Streaming Videos2026Skill-Decoupled On-Policy Expert Distillation for Unified Vision-Language Models2026AVC-Reward: Event-Centric Audio-Video Consistency Evaluation20263DCodeVerse: Exploring 3D Worlds as Code2026GeoState: Learning View-Decoupled Geometry States for 3D World Modeling2026GroundAnything: Reconciling Parallel Decoding with Precise Visual Grounding at Flash Speed2026GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives2026Harness-Delta Attribution: Dissecting Gains From Harness Evolution2026R$^2$Orch: Recursive Self-Improvement for Agent Orchestration via Self-Play within a Skill-Internalization Loop2026Are We Overthinking One-Step Generation?2026Generative Representation Learning for Scalable Latent Tokenization2026TimeRSI: Towards Self-Improved Agentic Deployment for Time-Series Foundation Models2026Spectral-Spatial Interpretation2026Each Reward to Its Relevant Tokens: Reward-Specific Token-Level Credit Assignment for Multi-Reward RL2026DuET: Note-Frame Dual Encoding for Piano Performance-to-Score Transcription2026When the KV Cache Is Slower Than the GPU: Scaling Remote KV Tiers for Context-Parallel Prefill2026REKI: Registering Knowledge in VLAs2026On the SPOT: Enhancing Speculative Decoding for Diffusion Language Models via Commit-Order Prediction2026Towards Looped Models Done Right, Part I: Topology, Input Injection, Recurrent-State Design2026A Language-Informed Loss for Visual Generation2026OR-License: Evaluating LLM Agents as Real-World Operations Research Practitioners2026ECHO: Efficient Causal Self-Recursion Beyond Auxiliary Multi-Token Prediction2026Epi: Towards a Persistent, Self-Improving Agent for Autonomous Research2026CoworkEnv: Synthesizing Training Environments for Long-Horizon Cowork Agents2026CoEvo-RSI: When Harnesses Evolve, Which Models Should Follow?2026Legal Guidance in the Wild: How Users Seek Legal Help in Real-World LLM Conversations2026AdaExploreBench: Benchmarking LLM Agents for Proactive Exploration and Test-time Adaptation in Evolving Environments2026Token-Set Choice Confounds POPE: A Systematic Audit of Yes/No Extraction in VLM Hallucination Evaluation2026HarnessRL: Co-Evolving Policies with a Just-in-Time Training Harness for Open-Ended RL2026SpatialScout: Can VLM Agents Actively Explore and Map 3D Scenes?2026TokenGate: Adapting LLMs for Fast and Memory-Efficient Prefill with Learned Token Pruning2026HarnessPro: Recursive Harness Self-Improvement with Pareto Optimization for Performance-Cost Trade-offs2026SpecMRL: Redesign Matryoshka Representation Learning within Parametric Spectrum2026MemT-1.0: Towards Memory-Native Evolution2026Agent World Model in the Wild: Computer-Use Agent Environment Synthesis2026Competing Learning in Hybrid Architectures2026OpenVLHarness: Real World Visual Agentic Tool-Use with Multimodal Memory2026Vision as Proprioception: Estimating Humanoid Robot Pose from Video2026World Model Interpretability: From Decoding Representations to Understanding Latent Worlds2026QD-Rec: Open-Ended Evolution of Recommendations via Quality-Diversity2026EdgeKernelBench: Unified LLM-Driven Kernel Generation for Edge Backends2026Has GPT-6-Astra Achieved Embodied Intelligence?2026TILDE: Teacher-in-the-Loop Distillation from Embodied Foundation Models2026BiDex-100: A Comprehensive Benchmark for Bimanual Dexterous Manipulation Across 100+ Simulation Tasks2026Omni-Woodpecker: Hallucination Correction via Evidence-Prior Rebalancing for Omni LLMs2026On the Fragility of Tasks under Composition: Implicit Budgeting in Language Models2026Which Agent's Skill Should Change? Group-Relative Credit Assignment for Evolving Multi-Agent Teams2026MMedAgent-Plus: Scaling Multimodal Medical Agents across Modalities, Tasks and Workflows2026Ask to Distill: Learning to Synthesize Data for On-Policy Distillation2026Judging the Justification: Decision Certificate Supervision for Generative Reward Models2026From RL Papers to Framework Code: Inclusion and Review in Seven Open-Source Training Frameworks2026Large Flow Language Models2026Cache Engineering for Agents2026MoveBench: A Rubric-Based Benchmark for Evaluating Large Language Models in Sports and Health Coaching2026Kronecker Delta Rule for Long Context Linear Attention2026AnyGoal: On World Model Generalization Beyond Expert Rollouts2026Where is Physical RSI capability from? An Observational Scaling Law of LLM-as-policy2026HABIT-Bench: Evaluating Habit Induction from Longitudinal Weak Evidence in Agent Memory2026RLSplat: Resource-Linear Budget Control for Feed-forward 3D Gaussian Splatting2026Scaling Parameter-Efficient Fine-Tuning for Brain Encoding2026Specter: Fast MoE Inference via Quantization-Enabled Multi-Step Speculative Offloading2025Automating ML for Science: Can Frontier Agents Climb Scientific Hills in the Wild?2026Harness Scaling for Generalizable Agentic Reinforcement Learning2026Retrieval Augmented Generation Compositionality2026Evolving with Open Worlds: Towards Recursive Self-Improvement for Future Forecasting2026Your Benchmark Is an Empirical Measure Over Difficulty2026ToM2Vec: Continuous Theory-of-Mind Tokens for Multi-Agent Coordination2026CulturalRiddles: A Multicultural Benchmark of Riddles2026Building Auditable AI Text Detectors with Sparse Autoencoders2026Looped Uniform Diffusion Language Models2026VX-One: Typed, Calibrated Decision Heads for Real-Time Perimeter Triage2026Think Harder Where It Matters: Item-Level Adaptive Recursion for Sequential Recommendation2026MMEnterprise-Bench: Evaluating Enterprise Agents over Evolving Multimodal Records2026Inherited, Not Invented: Tracing the Provenance of Wait in Reasoning Models2026Joint or Per-Option? Compact Encoders for Closed-Set Decisions from 17 Million to 1 Billion Parameters2026Beyond Watching and Searching: Benchmarking Agents for Video-centered Deep Research Reports2026Slot-JEPA: Object-Structured Prediction Interfaces for Latent World Models2026S-Agent 2.0: Spatial Reasoning with a General Coding Agent2026Brain2Vec: A Self-Supervised 3D Foundation Model for Structural Brain MRI2026Code2Video-Compass: A Comprehensive Benchmark for General-Purpose Code-Based Video Generation across T2V, I2V, and V2V2026Stratified Sparse Attention: Reducing the Sparse-to-Dense Gap at Extreme Sparsity2026DecisionPivot: A Diagnostic Benchmark for Decision Models in LLM Systems2026Sage-A1.5: From Long-Horizon Execution to Scientific Exploration with Capability Composition2026SceneRig: An Agentic System for Simulation-Ready 3D Scene Reconstruction from Single Images2026Evolving Agent Interfaces2026What Git Does Not Record: Provenance Legibility in Two AI-Assisted Research Repositories2026Scaling in Games: Continued Pre-Training For Embodied Agents In Diverse Virtual Worlds2026Hidden Messages Between Diffusion Language Model Agents: Causal Evidence for Latent Communication2026DiffResearch: Accelerating Deep Research with Diffusion Language Models2026LaLa-World: Scaling Latent World Models with Natural Language Actions2026HyperDrone: Simulation-to-Reality Transfer at Millions of Frames per Second2026LLMSeg: Training-Free Text-Prompted Image Segmentation with B-Splines2026Toward Recursive Self-Improvement in Physical AI2026Multimodal Embodied Agents: A Unified Capability-centric Perspective from Computer-Use to Robot-Use2026SplatLab: Autonomous Scientific Discovery for 3DGS Research2026LLM2Cert: LLM Synthesis of Certificates with SMT/Lean Verification2026Is Complex Agentic Memory Always Needed? Naive RAG Is Stronger Than We Thought2026Different Tasks, Different Timesteps: Temporal Task-Aware Quantization for Diffusion Language Models2026RIDAR: Safeguarding Research Integrity in Autonomous Research Agents via Automated Detection and Runtime Intervention2026PaperFormalizer: Author-in-the-Loop Autoformalization of Research Papers in Lean2026Can Test-Time Memory Evolution Outperform Test-Time Training for Scientific Discovery?2026$Ψ$-0.6: Adaptive Streams of Thought in a Unified Vision-Language World Model2026Principia: Test-Time Hypotheses to Principle Construction for Scientific Discovery2026RSIHub: A Composable Framework for Agent-Evolution Methods2026Trial by Trace: Benchmarking AI Agent Harnesses for Reliability and Failure Attribution2026FrontierCS 2: A Benchmark for Computer Science R&D2026Extreme Compression of Per-Layer Embeddings via Cross-Layer Latent Vector Quantization2026On the Token Collapse Phenomenon in Textual Graph Soft Prompting2026Rethinking Sequence Compression with Test-time Regression2026Nested Sparsity for Scalable Pretraining and Continual Learning2026Every Weight Is a Path of Shared Discrete Decisions2026Sensorimotor Alignment: A Platonic Proxy for Embodied Policy Success2026LoCA: Enabling Multi-Modality in Geospatial Location Encoders2026DraMaze: A Multi-Agent Pipeline with Future-Aware Memory for Diverse Microdrama Script Generation2026CRM-Agent-Bench: Benchmarking Tool-Using Agents in Stateful Enterprise Environments2026SceneBloom: Agentic Indoor Environment Generation with Realistic and Diverse Layouts for Embodied Learning2026Evolving to Match Capabilities: Improving Harness Self-Evolution via Skill Adaptation2026Novel or Not? PatEx-Gym: Agentic Prior-Art Search for Novelty Assessment in Patent Examination2026Mycelium: A Meta-Study of Hallucination Detection Across Provenance, Typology, and Granularities2026Trust Mechanisms for Offline Generative AI at the Edge: A Connectivity-Regime Framework and an Empirical Study on a Raspberry Pi 52026EnvLunch: The Environment as a Free Lunch for Learning to Interact2026Build, Don't Play: Self-Improving LLM Research Agents Develop a Hearthstone Battlegrounds AI That Rivals Top Players2026ReGE: Stronger Agents through Experience Synthesis2026OmniReasoning: Pushing the Limits of Audio-Visual Joint Reasoning2026OpenEBFM: Energy-based Foundation Models Are Scalable Architecture2026The Lexical Imprint of Pre-training Data2026Eat Your Precision: Precision Objective Mitigates Reward Hacking in Rubric-based RL2026Trajectory2Gym: Decision-Grounded Compositional Interaction Synthesis for Tool-Use Agents2026Continual Learning Exam: Benchmarking Long-Horizon RSI and Experiential Learning in LLM Agents2026Memory Tax: Dissecting the Net Value of Agent Memory2026Identically Trained Language Models Break on Different Jailbreaks2026Humans in Space: Can MLLMs Construct Human-Centric Spatial Understanding via Active Exocentric Exploration?2026$Δτ$-Bench: Pushing Agents Toward a Moving Capability2026What Effort Level Should I Choose? On the Economics of Reasoning Effort in LLM Agents2026What Outcome RL Buys: Finishing, Reliability, Rarely Reach2026DTCA: Denoising Token Credit Assignment for Reinforcement Learning in Diffusion Language Models2026Did the Model Use That? Counterfactual Training Boosts Self-Attribution in Language Models2026Skill$^{\infty}$: Recursive Self-Improvement for Agent Skill Optimization2026S3AC: Stabilized, Selective, and Scalable Actor-Critic for LLM Reasoning2026Matched Capabilities, Different Behaviors: A Controlled Study of MCP and CLI Interfaces for Language Agents2026ParetoRSI-Gym: Can AI Agents Improve Language Models Across Multiple Domains?2026Spatial Audio Generation and Editing with Executable Plans and Self-Distilling Transfusion2026AgentProofBench: Benchmarking Reusable Verifiers for Web Agents2026Intrinsic Goals for Autonomous Agents: Hierarchical Exploration Captures Future Volitional Action in Drosophila2026Benchmarking Obstacle-Conditioned Safety in Household Robot Navigation2026SearchHorizonBench: A Benchmark for Search Value in AI Agents2026PhysEvolve: Verifier-grounded Self-evolution for Physical Reasoning with Executable Worlds2026Co-Design Bench: Multimodal Collaborative Coding Evaluation for Agentic Models2026Trustworthy Recursive Self-Improvement in the Era of Large Models: Foundation, Risks, Evaluation and Open Challenges2026Omni-InteractScope: Benchmarking Proactive and Agentic Interaction in Continuous Omnimodal Streams2026OmniShowcase: Omnidirectional Subject-Consistent Video Generation from Multi-View Images2026Work Agents: Systems, Capabilities, and Optimization2026How Much Does the Hint Matter? Scope Conditions for Hindsight Self-Distillation in Financial Reasoning2026Lipi-Ghor: A Large-Scale, Diarized, Multi-Domain Bengali Speech Corpus and a Scalable Pipeline for Building It2026Cross-modal Steering Vectors Exist Across Both Image and Text Generation2026Zooming without Zooming v2: From Seeing Details to Connecting Evidence for Generalizable Multimodal Perception2026Your Policy Is Secretly a Harness: The Decision Geometry of Agent Interfaces2026EmbeddingGemma 2: Unified and Efficient Multimodal Representations2026Can LLMs Uncover Hidden Physical Laws through Free Exploration?2026The Trinity: Capability-Driven Recursive Self-Improvement for Multimodal Agents2026Beyond the Estimate: Wiener-Informed Decoding for Easier Training of Pixel-Space Flow Models2026FrameForge: Harnessing Priors of Video Foundation Models for Unified Image Generation2026Look Only When Needed: Certificate-Guided Visual Routing for Efficient Diffusion VLMs2026I-HITL: Inalienable Human-in-the-Loop2026WorkFactory: Turning Professional Intents into Executable Work for Scalable Agent Training2026MLE-World: Predicting Machine Learning Experiment Outcomes with a Learned World Model2026Channel Selection for Mixture-of-Experts Decoding Needs No Calibration Data2026Timestral2026Instruction Capacity: Behavior and Degradation of Language Models under Controlled Load2026GTFormer: A Syntax-Free Graph Network Alternative to Transformers for 8192$\times$ Length Extrapolation in Linear Time2026Proving and Improving the Platonic Representation Hypothesis with JEPAs2026DeFT-RAG: Decomposing Multi-Hop Questions into Self-Contained Sub-Questions with Reusable Answers2026