Daily Papers of 2026-08-26

  1. Annotations as Rollouts: Efficient and Scalable Reinforcement Learning for Video MLLMs 110 upvotes, #1 of 2026-08-26
  2. GigaBrain-0.7: Scaling Embodied Foundation Models to Emergent Capabilities with a Three-System Architecture 103 upvotes, #2 of 2026-08-26
  3. WeMM-Embedding: WeChat Multi-Modal Embedding Technical Report 68 upvotes, #3 of 2026-08-26
  4. On-Policy Self-Distillation in Diffusion Models 66 upvotes, #4 of 2026-08-26
  5. AutoSaddler: Automatic Harness Optimization with Durable Updates from Agent Execution Traces 64 upvotes, #5 of 2026-08-26
  6. SecOPD: Mitigating Adaptive Prompt Injections by On-Policy Distillation 41 upvotes, #6 of 2026-08-26
  7. CyberFactory: Scaling Cyber Security Capabilities with Instances from the Wild 33 upvotes, #7 of 2026-08-26
  8. Recursive Experiential-Working Memory Evolution for Long-Horizon Agent Harnesses 27 upvotes, #8 of 2026-08-26
  9. Best Practice Critic Optimization 17 upvotes, #9 of 2026-08-26
  10. On-policy Distillation with Verifiable Reward 17 upvotes, #9 of 2026-08-26
  11. Meta^n: Recursive Self-Improvement through Emergent Depth 15 upvotes, #11 of 2026-08-26
  12. LAION-BVD: A 10-Million-Hour Open Video Dataset for Multimodal Pre-training 15 upvotes, #11 of 2026-08-26
  13. Game2World Engine: Unlocking In-the-Wild Gameplay Videos for World Model Training 11 upvotes, #13 of 2026-08-26
  14. From Seeing to Acting: Smart Glasses as First-Person Intelligence Platforms 10 upvotes, #14 of 2026-08-26
  15. MARS: Multi-Specialist LLM Relay System for Competitive Programming 9 upvotes, #15 of 2026-08-26
  16. AgentRoom: Concurrent Multi-Agent Coding in a CRDT-Backed Shared Workspace 7 upvotes, #16 of 2026-08-26
  17. When "Must" Becomes "Maybe": Constraint Weakening in LLM Agent Workflows 6 upvotes, #17 of 2026-08-26
  18. CAFE: Self-Improving Search Agents Need Co-Evolving Feedback 6 upvotes, #17 of 2026-08-26
  19. DREAM Technical Report 5 upvotes, #19 of 2026-08-26
  20. Automata from Agent Traces: Failure and Next-Step Prediction 5 upvotes, #19 of 2026-08-26
  21. Length-Adaptive Decoding for Masked Diffusion Machine Translation 4 upvotes, #21 of 2026-08-26
  22. TorchMorph: CUDA-accelerated Morphological Transforms 4 upvotes, #21 of 2026-08-26
  23. Autonomous Mathematical Discovery in an Open-World Multi-Agent Environment 3 upvotes, #23 of 2026-08-26
  24. MoTE: Mixture of Task Experts for Multi-Task Video Understanding 2 upvotes, #24 of 2026-08-26
  25. MemUse: Moving Memory Evaluation from Direct QA to Natural Integration in Long-Term Human-AI Conversation 1 upvotes, #25 of 2026-08-26
  26. Latent Action as Intention Enables Efficient Future Imagination for World Action Models 1 upvotes, #25 of 2026-08-26

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.