Daily Papers of 2026-08-31

  1. LoopArena: Benchmarking Models as Runtime Controllers for Loop Engineering 103 upvotes, #1 of 2026-08-31
  2. Beyond Data Scaling: Representation-Centric Continued Pre-training for Vision-Language-Action Models 93 upvotes, #2 of 2026-08-31
  3. DART-SD: Diamond-topology Aware Retrieval and Tuning for Self-Distillation of Multi-Turn Tool-Calling Agents 91 upvotes, #3 of 2026-08-31
  4. Agentic Artifact Creation: Systems, Evaluation, Principles, and Opportunities 65 upvotes, #4 of 2026-08-31
  5. Code as Worlds: Agentic Discovery of Executable World Representations for Physical Reasoning 51 upvotes, #5 of 2026-08-31
  6. J-Zero: Unified Challenger--Solver--Judge Co-Evolution from Zero Data 43 upvotes, #6 of 2026-08-31
  7. StarHarness: Evolving Harnesses with Stratified Search for Enterprise Environments 40 upvotes, #7 of 2026-08-31
  8. Puro-2B: Poor Lab's Qwen2-1.5B Trained on RTX 5090 within $5090 38 upvotes, #8 of 2026-08-31
  9. Revisiting Local Context for Long-Horizon Streaming 3D Reconstruction 33 upvotes, #9 of 2026-08-31
  10. Fast Weight Attention for Continual Learning 31 upvotes, #10 of 2026-08-31
  11. Act with Intent: Distilling Behavior Intent for Vision-Language-Action Models 30 upvotes, #11 of 2026-08-31
  12. LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation 29 upvotes, #12 of 2026-08-31
  13. ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RL 27 upvotes, #13 of 2026-08-31
  14. Locate Anything in Videos: Rethinking Efficient Generative Spatio-Temporal Video Grounding 19 upvotes, #14 of 2026-08-31
  15. Blind Men and the Elephant: Probing the Epistemic Myopia of LLMs under Long-Tail Divergent Knowledge 19 upvotes, #14 of 2026-08-31
  16. Paint What You See: Benchmarking Dexterous Visual Tool Use in Multimodal Agents 17 upvotes, #16 of 2026-08-31
  17. Ring Forcing: Towards Precise Long-Term Memory for Autoregressive Video Diffusion 17 upvotes, #16 of 2026-08-31
  18. StepGuard: Learning Step-Level Guardrails with Scalable Supervision and Safety-Utility Balancing 16 upvotes, #18 of 2026-08-31
  19. Sliding-window beats linear attention 16 upvotes, #18 of 2026-08-31
  20. Video Generative Models as Geometry Learner 15 upvotes, #20 of 2026-08-31
  21. PonderPounce: A Pretrained MLLM as an Episode Context Engine for Robot Control 13 upvotes, #21 of 2026-08-31
  22. EvoUndo: Recoverability-Constrained Self-Evolution for LLM Agent Harnesses 9 upvotes, #22 of 2026-08-31
  23. Rubric-to-Code Credit Assignment for Reinforcement Learning 7 upvotes, #23 of 2026-08-31
  24. Language Chain in Alignment: Cross-lingual Ranking Preference Optimization 6 upvotes, #24 of 2026-08-31
  25. Acquire, Repair, Preserve: A Diagnosis-Guided Post-Training Recipe for Small-Model Dialogue Game Agents 6 upvotes, #24 of 2026-08-31
  26. LMSM: LLM Security Framework Inspired by Linux Security Modules 5 upvotes, #26 of 2026-08-31
  27. Training, learning and inference: unified dynamics of neural systems 4 upvotes, #27 of 2026-08-31
  28. Ask or Answer: A Decision Framework for Multi-Turn Health Misinformation Intervention 4 upvotes, #27 of 2026-08-31
  29. GGSS: Geodesic-Gated Spherical Steering for Inference-Time Debiasing of Generative Vision-Language Models 4 upvotes, #27 of 2026-08-31
  30. Lost in Compression: A Controlled Cross-Lingual Audit of Extractive Prompt Compressors 3 upvotes, #30 of 2026-08-31
  31. Generative Semantic Scene Completion 3 upvotes, #30 of 2026-08-31

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.