Daily Papers of 2026-10-16

  1. MJev-Spatial: System One Spatial Intelligence via Native Attention Matching 414 upvotes, #1 of 2026-10-16
  2. OmniCompass: A Holistic Benchmark for Omni-Modal Video Grounding and Understanding 256 upvotes, #2 of 2026-10-16
  3. Waypoint-OPD: Transferring RL-Induced Policy Shifts via Gated On-Policy Distillation 219 upvotes, #3 of 2026-10-16
  4. MAI-Gym: An Open All-in-One Platform for GUI Agent Evaluation and RL Training across Mobile and Computer Environments 150 upvotes, #4 of 2026-10-16
  5. InfiniSplat v2: Analyzing and Improving Feed-Forward 3DGS from Any Views 133 upvotes, #5 of 2026-10-16
  6. OpenGCUA: Toward a Fully Open Stack for Generalist Computer Use 103 upvotes, #6 of 2026-10-16
  7. Self-Evolving Agent Harness 99 upvotes, #7 of 2026-10-16
  8. Speculative Forcing for Efficient Autoregressive Video Generation 94 upvotes, #8 of 2026-10-16
  9. FDE-Gym: The Last Mile of AI Delivery 77 upvotes, #9 of 2026-10-16
  10. RoTA-Bench: Evaluating 3D Spatial Reasoning through Interactive Projection Alignment 76 upvotes, #10 of 2026-10-16
  11. 3DCodeVerse: Exploring 3D Worlds as Code 67 upvotes, #11 of 2026-10-16
  12. VLA-Δ: Learning Physical State Transition for Vision-Language-Action Models 65 upvotes, #12 of 2026-10-16
  13. Mosaic: Multi-Session Agentic Reinforcement Learning 63 upvotes, #13 of 2026-10-16
  14. Learning Where to Split: Dynamic Band Splitting for Audio Source Separation 61 upvotes, #14 of 2026-10-16
  15. Intern Lumina U2: A Multi-Codebook Diffusion Large Language Model for Omni-Visual Understanding and Image Generation 51 upvotes, #15 of 2026-10-16
  16. Orbia: Do Generated Videos Form a Consistent 3D World? 44 upvotes, #16 of 2026-10-16
  17. Each Reward to Its Relevant Tokens: Reward-Specific Token-Level Credit Assignment for Multi-Reward RL 44 upvotes, #16 of 2026-10-16
  18. CommerceGen-Studio: An End-to-End Framework for Commercial-Grade Main-Image Generation 43 upvotes, #18 of 2026-10-16
  19. SCOPE: Semantic Bounded Credit Optimization for Prompt Engineering in Text-to-image Generation 43 upvotes, #18 of 2026-10-16
  20. AutoSFT: An Agentic System for Automated Data Research in Supervised Fine-Tuning 42 upvotes, #20 of 2026-10-16
  21. Terminal-RSI: Recursive Self-Improving Terminal Agents via Recursive Self-Rewrite and Policy-Refine 41 upvotes, #21 of 2026-10-16
  22. ForcingBooster: Adaptive Patch-wise Denoising for Fast Autoregressive Video Diffusion 38 upvotes, #22 of 2026-10-16
  23. DeltaStreamer: Towards Human-Like Agentic Streaming Video Understanding 38 upvotes, #22 of 2026-10-16
  24. WorldCast: Exploring Intelligent Visual Generation in World State Prediction 38 upvotes, #22 of 2026-10-16
  25. Data World: Scaling Data Environments from Scratch 38 upvotes, #22 of 2026-10-16
  26. OmniVerifier-C1: Decomposed Graded Verifier for Rule-Governed Generative World Modeling 35 upvotes, #26 of 2026-10-16
  27. InfoVis: Unified Slide Image Generation and Editing 32 upvotes, #27 of 2026-10-16
  28. Velox: Spatial Memory Makes World Models Efficient 32 upvotes, #27 of 2026-10-16
  29. Behavioral Compositionality of Sparse Autoencoder Representations 31 upvotes, #29 of 2026-10-16
  30. EdgeWise: Bringing Long-Horizon Agents to the Edge via Selective Cloud Delegation 30 upvotes, #30 of 2026-10-16
  31. Routing Policy Optimization for Online Diffusion Reinforcement Learning 29 upvotes, #31 of 2026-10-16
  32. SimVTA: Simulation-Grounded Video-Trajectory Alignment for Driving World Action Models 28 upvotes, #32 of 2026-10-16
  33. MLExpert: Towards a Generalist Agentic Model for Autonomous Machine Learning Engineering 28 upvotes, #32 of 2026-10-16
  34. Can World and Action Modeling Benefit Each Other under Unified Causal Training? 26 upvotes, #34 of 2026-10-16
  35. Intern-Decision: Efficient Multimodal Decision Making with Autoregressive Masked Language Modeling 26 upvotes, #34 of 2026-10-16
  36. Trajectory-Level Speculation with Semantic Verification for Tool-Using Agents 25 upvotes, #36 of 2026-10-16
  37. Virtual Head Attention 23 upvotes, #37 of 2026-10-16
  38. JevAgentHarness: Benchmarking, Deploying and Training Decision Models That Share Control with LLMs 23 upvotes, #37 of 2026-10-16
  39. SWE-at-Scale: Scalable Issue Synthesis from Historical Bug Fixes 20 upvotes, #39 of 2026-10-16
  40. Do Better Visual Representations Always Lead to Better End-to-End Autonomous Driving? 19 upvotes, #40 of 2026-10-16
  41. MFVSR: A Multi-Style and Fine-Grained Benchmark for Video Subtitle Removal 17 upvotes, #41 of 2026-10-16
  42. Beyond Policy Rollouts: Embedding-Space Directional Exploration with Anchored Noise for LLM RLVR 16 upvotes, #42 of 2026-10-16
  43. Athanor: An Agent-Native System for Autonomous Post-Training Research 16 upvotes, #42 of 2026-10-16
  44. FlashMask-3: Efficient and Expressive Mask-Aware Distributed Attention 15 upvotes, #44 of 2026-10-16
  45. Learning Self-Orchestration in Decentralized Agent Swarms 15 upvotes, #44 of 2026-10-16
  46. SUPRA-TEX: 3D Generation with Super-Aligned Texture 15 upvotes, #44 of 2026-10-16
  47. Cowork-Gym: Synthesizing Interactive Work Environments at Scale 13 upvotes, #47 of 2026-10-16
  48. CarryOn: A Dual-Loop Harness for Personalized and Reliable Robot Assistance 13 upvotes, #47 of 2026-10-16
  49. Skill-Decoupled On-Policy Expert Distillation for Unified Vision-Language Models 13 upvotes, #47 of 2026-10-16
  50. LEASH: Uncertainty-Aware Weak-to-Strong Escalation and Oversight for Computer-Use Agents 13 upvotes, #47 of 2026-10-16
  51. CoDA: Co-Evolution of Data Engines and Action Policies 11 upvotes, #51 of 2026-10-16
  52. Learning to Read Out: Unembedding Dynamics in Language Model Pretraining 10 upvotes, #52 of 2026-10-16
  53. Calibrated Typed Decisions for Entity, Relation, and Event Extraction 9 upvotes, #53 of 2026-10-16
  54. Copy-Paste Language Models 9 upvotes, #53 of 2026-10-16
  55. Memory as Weights: Internalizing Long-Term History for Streaming Videos 9 upvotes, #53 of 2026-10-16
  56. LeanPRM: Lean-Grounded Process Supervision for Natural Language Proof Verification 8 upvotes, #56 of 2026-10-16
  57. SERO: Learning from Executable Oracle Continuations for Long-Horizon Agentic Reinforcement Learning 8 upvotes, #56 of 2026-10-16
  58. Mine, Sample, Render: Test-Time Diversity Alignment Of Generative Models 8 upvotes, #56 of 2026-10-16
  59. ANTE: Anticipate, Then Execute -- Learning Internal World Models for Coding Agents from Verified Beliefs 7 upvotes, #59 of 2026-10-16
  60. Gen4P: Video Generators are 4D Perceivers 7 upvotes, #59 of 2026-10-16
  61. CLIFT: Conformal Self-Verification for Web Agent Training and Test-Time Scaling 6 upvotes, #61 of 2026-10-16
  62. Recursive Environment Scaling for General Long-Horizon Agents 6 upvotes, #61 of 2026-10-16
  63. WorldMemBench: A Comprehensive Benchmark for Memory of Video World Models 6 upvotes, #61 of 2026-10-16
  64. PN-Mix: Pair-Norm Guided Mixed-Precision Key Cache Quantization for Efficient Long-Output LLM Inference 6 upvotes, #61 of 2026-10-16
  65. RepForcing: Representation-Space Initialization for Autoregressive Video Distillation 6 upvotes, #61 of 2026-10-16
  66. MMSearch-OPD: Easy Made Hard, Hard Made Learnable for Long-Horizon Multimodal Search Agents 6 upvotes, #61 of 2026-10-16
  67. A team of Gemini 4 Argon agents beats Minecraft 6 upvotes, #61 of 2026-10-16
  68. Scalable Weak-to-Strong Verifier Hardening for Agent Environments 5 upvotes, #68 of 2026-10-16
  69. TimeRSI: Towards Self-Improved Agentic Deployment for Time-Series Foundation Models 5 upvotes, #68 of 2026-10-16
  70. Explaining the Roles of Experts in Mixture-of-Experts LLMs for Personalization 5 upvotes, #68 of 2026-10-16
  71. PathFM: Fast Generative Forecasting of Multivariate Trajectories 5 upvotes, #68 of 2026-10-16
  72. H-DDCM: Hadamard Codebooks for Fast Generative Image and Video Compression 4 upvotes, #72 of 2026-10-16
  73. One Bias Score, Two Factors: How Mitigations Affect Option Bias in Multiple-Choice Evaluation 4 upvotes, #72 of 2026-10-16
  74. When the KV Cache Is Slower Than the GPU: Scaling Remote KV Tiers for Context-Parallel Prefill 4 upvotes, #72 of 2026-10-16
  75. EVA-Cap: Optimizing Audiovisual Video Captioning via Event-Centric Alignment 4 upvotes, #72 of 2026-10-16
  76. Speech event detector: Timestamped detection of non-verbal vocalizations and background sounds for rich speech transcription 3 upvotes, #76 of 2026-10-16
  77. PINT: A Unified Model with Shared Predictive Intent for Robot Manipulation 3 upvotes, #76 of 2026-10-16
  78. AI-Driven, Timing-Aware Power Optimization 3 upvotes, #76 of 2026-10-16
  79. Equal CER, Unequal Damage: Severity-Weighted Character Error Rate for OCR 3 upvotes, #76 of 2026-10-16
  80. VascuSR: Iterative Anatomy-Aware Diffusion for CT Super-Resolution 3 upvotes, #76 of 2026-10-16
  81. Diminishing Returns of Data in Sparse Autoencoders: Functional Behavior does not Guarantee Feature Reproducibility 3 upvotes, #76 of 2026-10-16
  82. CoWorld Reasoner: Agentic Multimodal Reasoning with the Evolving Video World Model-Driven Harness 3 upvotes, #76 of 2026-10-16
  83. SimForcing: Distilling Simulation Motion Priors into Real-Domain Robot World Models 2 upvotes, #83 of 2026-10-16
  84. PAIR: Bridging Perception and Action in Vision-Language-Action Models 1 upvotes, #84 of 2026-10-16

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.