Daily Papers of 2026-10-22

  1. StreamEval: Evaluating Interactive Streaming Video Generation Models 865 upvotes, #1 of 2026-10-22
  2. Spectral-Spatial Interpretation 758 upvotes, #2 of 2026-10-22
  3. Right Screen, Wrong Transition: World Models as Verifiers for GUI Agents 189 upvotes, #3 of 2026-10-22
  4. Looking Ahead: Modeling Future Evolution in Autoregressive Video Generation 174 upvotes, #4 of 2026-10-22
  5. SpatialScout: Can VLM Agents Actively Explore and Map 3D Scenes? 170 upvotes, #5 of 2026-10-22
  6. DONIC: Generalist Dexterous Bimanual Tracking Control from Human Demonstrations 162 upvotes, #6 of 2026-10-22
  7. Cache Engineering for Agents 161 upvotes, #7 of 2026-10-22
  8. Agent World Model in the Wild: Computer-Use Agent Environment Synthesis 101 upvotes, #8 of 2026-10-22
  9. Scaling in Games: Continued Pre-Training For Embodied Agents In Diverse Virtual Worlds 86 upvotes, #9 of 2026-10-22
  10. PrismaCoder: Pushing the Boundaries of Open Multimodal Code Intelligence through Decomposed Rubric Synthesis 72 upvotes, #10 of 2026-10-22
  11. Retrieval Augmented Generation Compositionality 55 upvotes, #11 of 2026-10-22
  12. TraceForge: Distilling with Your Own Thinking 50 upvotes, #12 of 2026-10-22
  13. Socratic Skill Learning: Interleaving Teaching and Practice for Transferable Agent Skills 49 upvotes, #13 of 2026-10-22
  14. AgenticRoboRL: Bringing Coding-Agent Post-Training to Agentic Robotics 49 upvotes, #13 of 2026-10-22
  15. CulturalRiddles: A Multicultural Benchmark of Riddles 46 upvotes, #15 of 2026-10-22
  16. Building AI Research Teams for Autonomous Discovery of Linear Attention Architectures 45 upvotes, #16 of 2026-10-22
  17. Memory-Centric Continual Learning for AI Agents 43 upvotes, #17 of 2026-10-22
  18. Think Harder Where It Matters: Item-Level Adaptive Recursion for Sequential Recommendation 41 upvotes, #18 of 2026-10-22
  19. Qwen-Audio-3.1-ASR Technical Report 37 upvotes, #19 of 2026-10-22
  20. CompactSplat: Spatially Adaptive Gaussian Distribution for Feedforward Scene Reconstruction 34 upvotes, #20 of 2026-10-22
  21. From VLM Planners to Robotic Agents: Self-Evolving Harnesses for History-Dependent Manipulation 34 upvotes, #20 of 2026-10-22
  22. ALPS: Visual State Encoders for World Models 34 upvotes, #20 of 2026-10-22
  23. DeLoop: Repeated Actions as Self-Supervision for Long-Horizon Agents 33 upvotes, #23 of 2026-10-22
  24. Harness Scaling for Generalizable Agentic Reinforcement Learning 33 upvotes, #23 of 2026-10-22
  25. VerifierBench: Benchmarking the Generation of Executable Verifiers for LLM Agents 32 upvotes, #25 of 2026-10-22
  26. Automating ML for Science: Can Frontier Agents Climb Scientific Hills in the Wild? 30 upvotes, #26 of 2026-10-22
  27. RealMe: Immersive Volumetric Video from Minimal Captures 28 upvotes, #27 of 2026-10-22
  28. Real In, Real Out: What If We Only Use Real Data for Scene Text Editing? 27 upvotes, #28 of 2026-10-22
  29. Sparsity Should Be ALLOTted by Output: Training-Free Allocation of Block-Sparse Attention for Video Diffusion Transformers 25 upvotes, #29 of 2026-10-22
  30. Agentic World Action Models 25 upvotes, #29 of 2026-10-22
  31. PACE: Periodic and Channel-Aligned Encoding of Four-Bit Recurrent States 24 upvotes, #31 of 2026-10-22
  32. Multimodal Embodied Agents: A Unified Capability-centric Perspective from Computer-Use to Robot-Use 24 upvotes, #31 of 2026-10-22
  33. Entertain, then Cancel: How Thinking Makes Reasoning Models Read Some and Or Literally 23 upvotes, #33 of 2026-10-22
  34. Language Models Learn to Improve Language Models 23 upvotes, #33 of 2026-10-22
  35. Which Agent's Skill Should Change? Group-Relative Credit Assignment for Evolving Multi-Agent Teams 22 upvotes, #35 of 2026-10-22
  36. WorldStream: Streaming Interactive 3D Worlds with Hierarchical Planning and Persistent State 20 upvotes, #36 of 2026-10-22
  37. SpatialBrain: Spatial Intelligence Growing from Experience 20 upvotes, #36 of 2026-10-22
  38. Evolving with Open Worlds: Towards Recursive Self-Improvement for Future Forecasting 19 upvotes, #38 of 2026-10-22
  39. Toward Part-Aware Choral Transcription with singing voice assignment 18 upvotes, #39 of 2026-10-22
  40. Hidden Messages Between Diffusion Language Model Agents: Causal Evidence for Latent Communication 18 upvotes, #39 of 2026-10-22
  41. ACE-Brain-0.6: Scaling Spatial Grounding for Physical Foundation Models 17 upvotes, #41 of 2026-10-22
  42. StyleStream 2.0: Fast and Controllable Streaming Voice Style Conversion 17 upvotes, #41 of 2026-10-22
  43. Answering in the User's Variety: Calibrated Accommodation Pairs for Post-Training 16 upvotes, #43 of 2026-10-22
  44. ToM2Vec: Continuous Theory-of-Mind Tokens for Multi-Agent Coordination 16 upvotes, #43 of 2026-10-22
  45. RubatoCodec: Budgeted Adaptive Speech Tokenization with Streaming Decoding 15 upvotes, #45 of 2026-10-22
  46. From RL Papers to Framework Code: Inclusion and Review in Seven Open-Source Training Frameworks 14 upvotes, #46 of 2026-10-22
  47. DiffResearch: Accelerating Deep Research with Diffusion Language Models 13 upvotes, #47 of 2026-10-22
  48. Quant-Arena: Benchmarking LLM Agents for Quantitative Trading Research 13 upvotes, #47 of 2026-10-22
  49. VX-One: Typed, Calibrated Decision Heads for Real-Time Perimeter Triage - An empirical study of unknown-attack honesty, cascade gray zones, and calibration under drift 12 upvotes, #49 of 2026-10-22
  50. Looped Uniform Diffusion Language Models 11 upvotes, #50 of 2026-10-22
  51. Stratified Sparse Attention: Reducing the Sparse-to-Dense Gap at Extreme Sparsity 7 upvotes, #51 of 2026-10-22
  52. Design Anything, Design at Scale 6 upvotes, #52 of 2026-10-22
  53. Exploring The Empirical Privacy Landscape of Synthetic Data 5 upvotes, #53 of 2026-10-22
  54. Bridged Self-Distillation for On-Policy Learning 4 upvotes, #54 of 2026-10-22
  55. MMedAgent-Plus: Scaling Multimodal Medical Agents across Modalities, Tasks and Workflows 4 upvotes, #54 of 2026-10-22
  56. FrontierCS 2: A Benchmark for Computer Science R&D 4 upvotes, #54 of 2026-10-22
  57. Same Simulation, Different Evidence: Where Retrieved Tool Output Costs LLM Agents Accuracy 3 upvotes, #57 of 2026-10-22
  58. RoboCasa-Shift: Stage-Aware Robustness Evaluation of Vision-Language-Action Policies in Long-Horizon Household Manipulation 3 upvotes, #57 of 2026-10-22

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.