Daily Papers of 2026-10-19

  1. BizBench: Unveiling Hidden Challenges in Deploying Long-Horizon Agents for Real-World Business 496 upvotes, #1 of 2026-10-19
  2. Has GPT-6-Astra Achieved Embodied Intelligence? 481 upvotes, #2 of 2026-10-19
  3. Into the Unknown: Benchmarking AI agents on open scientific research problems 159 upvotes, #3 of 2026-10-19
  4. Learning to Think and Explore in 3D World 141 upvotes, #4 of 2026-10-19
  5. Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents 125 upvotes, #5 of 2026-10-19
  6. HexaShell: A Thin Harness for Exploring the Limits of VLM Agents as Robot Coding Agents 123 upvotes, #6 of 2026-10-19
  7. LEV: Thinking Fast and Deep with Looped Decision Models 105 upvotes, #7 of 2026-10-19
  8. MOSS-Image Technical Report 104 upvotes, #8 of 2026-10-19
  9. Scene-Designer: Learning Design Intelligence for Agentic 3D Scene Generation 63 upvotes, #9 of 2026-10-19
  10. LiveMACE: Process-Aware Evaluation of LLM Agent Capabilities in Evolving Markets 61 upvotes, #10 of 2026-10-19
  11. OpenMobile-2: Building Versatile Mobile Agents with Scalable Environments and App-Native Tools 61 upvotes, #10 of 2026-10-19
  12. Tree-Aware Posterior Distillation for Long-Horizon Agentic Reinforcement Learning 57 upvotes, #12 of 2026-10-19
  13. From Research Goals to Scientific Closure: Evaluating Agents Through Evidence-Supported Physics Research 54 upvotes, #13 of 2026-10-19
  14. Lexical Reference Frames: Interpreting LM Residual Streams in Native Semantics 54 upvotes, #13 of 2026-10-19
  15. Xiaomi-Lens-0 Technical Report 49 upvotes, #15 of 2026-10-19
  16. BridgeWAM: Bridging Video and Action Experts via Latent Bridge Queries 48 upvotes, #16 of 2026-10-19
  17. PremiseCheck: Efficient Fact-Checking of Human and AI Peer Reviews 47 upvotes, #17 of 2026-10-19
  18. RAP: A Dataset for Paper-to-Poster Generation via Declarative Markup from Human-Crafted Exemplars 39 upvotes, #18 of 2026-10-19
  19. Can Visual Generation Expand What Unified Multimodal Models Know? 38 upvotes, #19 of 2026-10-19
  20. DLite: History-free Speculative Decoding for Agentic AI 36 upvotes, #20 of 2026-10-19
  21. Execution Before Memory: Diagnosing Memory Failures in Video World Models 35 upvotes, #21 of 2026-10-19
  22. MTEdit: Packed Trajectories Beat Rollouts for Multi-Turn Image Editing 35 upvotes, #21 of 2026-10-19
  23. Consistent Distribution Matching for Data-Free Diffusion Distillation 24 upvotes, #23 of 2026-10-19
  24. Lean4AgenticRL: Advancing Agentic RL with Verifiable Multi-level Process Rewards 24 upvotes, #23 of 2026-10-19
  25. OmniXRef: A Unified Framework for Multimodal Multi-Reference Video Editing 23 upvotes, #25 of 2026-10-19
  26. OfficeTown: Generative Organizations for Evolving Agent Intelligence 23 upvotes, #25 of 2026-10-19
  27. Embeddings Are Overpriced? A Frozen 16 MB Vocabulary for Small and Ternary Language Models 21 upvotes, #27 of 2026-10-19
  28. OPMD: On-Policy Mechanism Distillation Beyond Output Supervision 20 upvotes, #28 of 2026-10-19
  29. DiscoveryRSI: Recursively Learning to Discover Valuable Hypotheses 20 upvotes, #28 of 2026-10-19
  30. Belief-Trajectory Energy: Measuring the Path to a Prediction 19 upvotes, #30 of 2026-10-19
  31. Dusting the Phantom: Gaussian Allocation for Fading Dynamic Content 19 upvotes, #30 of 2026-10-19
  32. Stuck on Shuffle: On Attribute-Object Binding in Vision-Language Models 18 upvotes, #32 of 2026-10-19
  33. Towards Looped Models Done Right, Part I: Topology, Input Injection, Recurrent-State Design 18 upvotes, #32 of 2026-10-19
  34. Dream2Discern: Learning from Hard Latent Variants for Runtime Failure Prediction in Multi-Agent Systems 16 upvotes, #34 of 2026-10-19
  35. DuoDiscover: Scaling Autonomous Discovery Beyond the Evaluation Bottleneck 16 upvotes, #34 of 2026-10-19
  36. MorphAct: Robot Policies Morphing with the Worlds They Act In 16 upvotes, #34 of 2026-10-19
  37. SciΠvot: Benchmarking Principal-Investigator Judgment at Pivotal Moments of Scientific AutoResearch Agents 16 upvotes, #34 of 2026-10-19
  38. AVC-Reward: Event-Centric Audio-Video Consistency Evaluation 16 upvotes, #34 of 2026-10-19
  39. OS-Helix: Unifying GUI Interaction and Code Execution in Native Computer-Using Agents 13 upvotes, #39 of 2026-10-19
  40. Librarian Agents: Training LLMs to Curate Long-Context Workspaces that Anticipate Future Queries 13 upvotes, #39 of 2026-10-19
  41. PRISM: PRIvileged Sibling-guided Multi-Agent Topology Design 11 upvotes, #41 of 2026-10-19
  42. FC-SWE: Failure-Conditioned RL for Long-Horizon Software Engineering Agents 10 upvotes, #42 of 2026-10-19
  43. BRACE: Adapting Whole-Body References for Force and Terrain Aware Humanoid Motion Tracking 9 upvotes, #43 of 2026-10-19
  44. Failure Recovery via Belief Updates and World Modeling for Long-Horizon Agents 8 upvotes, #44 of 2026-10-19
  45. DAGFormer: Global Topology Prediction for Transformer Cross-Layer Connections 8 upvotes, #44 of 2026-10-19
  46. Sparse Autoencoders Are a Shared Coordinate System for Symbolic Matching 8 upvotes, #44 of 2026-10-19
  47. GlobeBench: Evaluating Text-Based World Model Capabilities for Scalable Environment Simulations 8 upvotes, #44 of 2026-10-19
  48. EvoWeb: A Benchmark for Coding Agents on Evolving Web Applications 7 upvotes, #48 of 2026-10-19
  49. Recognition Is Not Grounding: Benchmarking Multilingual Financial Evidence Extraction 6 upvotes, #49 of 2026-10-19
  50. TokenEconomizer: Experience-Driven Black-Box Cost Control for Coding Agents 6 upvotes, #49 of 2026-10-19
  51. Game2Real-Bench: Do Toy Games Reveal and Transfer to Real-World Visual Reasoning? 5 upvotes, #51 of 2026-10-19
  52. ReaDy: Recurrent States as Dynamic Context for Speculative Drafting in Hybrid Linear Attention Large Language Models 5 upvotes, #51 of 2026-10-19
  53. Persistent Planar Memory for Video World Models 5 upvotes, #51 of 2026-10-19
  54. Robust Explanatory Analysis for Structured Optimization Networks (REASON) in Learning 4 upvotes, #54 of 2026-10-19
  55. Skip the Compressor: Coding Agents Fix More Issues When They Pick Whole Files 4 upvotes, #54 of 2026-10-19
  56. Resisting Harmful Instructions through Training-Free Weight Editing 4 upvotes, #54 of 2026-10-19
  57. Selection Is Not Learning: Choosing Model Futures 4 upvotes, #54 of 2026-10-19
  58. What It Is, Not What It Says: Securing Agentic Browsing Interactions with Web Semantics 4 upvotes, #54 of 2026-10-19
  59. StopWatch: Goal Grounding for Object-goal Navigation 4 upvotes, #54 of 2026-10-19
  60. DisaggKD: Efficient Knowledge Distillation through Teacher-Student Disaggregation and a Fused Loss 3 upvotes, #60 of 2026-10-19
  61. DuET: Note-Frame Dual Encoding for Piano Performance-to-Score Transcription 3 upvotes, #60 of 2026-10-19
  62. AI-Driven Disaster Response and Displacement Monitoring 3 upvotes, #60 of 2026-10-19
  63. ARC-SORT: Adaptive, Robust, Compensated Multi-Object Tracking with Interacting Multiple Models 3 upvotes, #60 of 2026-10-19
  64. GeoState: Learning View-Decoupled Geometry States for 3D World Modeling 3 upvotes, #60 of 2026-10-19
  65. Relational Learning as World Modeling 2 upvotes, #65 of 2026-10-19

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.