Daily Papers of 2025-07-29

  1. Agentic Reinforced Policy Optimization 127 upvotes, #1 of 2025-07-29
  2. A Survey of Self-Evolving Agents: On Path to Artificial Super Intelligence 72 upvotes, #2 of 2025-07-29
  3. ARC-Hunyuan-Video-7B: Structured Video Comprehension of Real-World Shorts 55 upvotes, #3 of 2025-07-29
  4. SmallThinker: A Family of Efficient Large Language Models Natively Trained for Local Deployment 51 upvotes, #4 of 2025-07-29
  5. Rep-MTL: Unleashing the Power of Representation-level Task Saliency for Multi-Task Learning 38 upvotes, #5 of 2025-07-29
  6. Reconstructing 4D Spatial Intelligence: A Survey 33 upvotes, #6 of 2025-07-29
  7. Geometric-Mean Policy Optimization 30 upvotes, #7 of 2025-07-29
  8. Diversity-Enhanced Reasoning for Subjective Questions 22 upvotes, #8 of 2025-07-29
  9. GPT-IMAGE-EDIT-1.5M: A Million-Scale, GPT-Generated Image Dataset 20 upvotes, #9 of 2025-07-29
  10. Region-based Cluster Discrimination for Visual Representation Learning 17 upvotes, #10 of 2025-07-29
  11. UloRL:An Ultra-Long Output Reinforcement Learning Approach for Advancing Large Language Models' Reasoning Abilities 13 upvotes, #11 of 2025-07-29
  12. Met^2Net: A Decoupled Two-Stage Spatio-Temporal Forecasting Model for Complex Meteorological Systems 12 upvotes, #12 of 2025-07-29
  13. ScenePainter: Semantically Consistent Perpetual 3D Scene Generation with Concept Relation Alignment 12 upvotes, #12 of 2025-07-29
  14. ForCenNet: Foreground-Centric Network for Document Image Rectification 11 upvotes, #14 of 2025-07-29
  15. JAM: A Tiny Flow-based Song Generator with Fine-grained Controllability and Aesthetic Alignment 8 upvotes, #15 of 2025-07-29
  16. Music Arena: Live Evaluation for Text-to-Music 8 upvotes, #15 of 2025-07-29
  17. Beyond Binary Rewards: Training LMs to Reason About Their Uncertainty 6 upvotes, #17 of 2025-07-29
  18. EDGE-GRPO: Entropy-Driven GRPO with Guided Error Correction for Advantage Diversity 6 upvotes, #17 of 2025-07-29
  19. Goal Alignment in LLM-Based User Simulators for Conversational AI 4 upvotes, #19 of 2025-07-29
  20. SAND-Math: Using LLMs to Generate Novel, Difficult and Useful Mathematics Questions and Answers 4 upvotes, #19 of 2025-07-29
  21. Running in CIRCLE? A Simple Benchmark for LLM Code Interpreter Security 1 upvotes, #21 of 2025-07-29
  22. GenoMAS: A Multi-Agent Framework for Scientific Discovery via Code-Driven Gene Expression Analysis 1 upvotes, #21 of 2025-07-29

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.