Daily Papers of 2026-01-09

  1. GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL Optimization 191 upvotes, #1 of 2026-01-09
  2. RL-AWB: Deep Reinforcement Learning for Auto White Balance Correction in Low-Light Night-time Scenes 44 upvotes, #2 of 2026-01-09
  3. Learnable Multipliers: Freeing the Scale of Language Model Matrix Layers 40 upvotes, #3 of 2026-01-09
  4. Token-Level LLM Collaboration via FusionRoute 39 upvotes, #4 of 2026-01-09
  5. VideoAuto-R1: Video Auto Reasoning via Thinking Once, Answering Twice 32 upvotes, #5 of 2026-01-09
  6. RelayLLM: Efficient Reasoning via Collaborative Decoding 27 upvotes, #6 of 2026-01-09
  7. AT^2PO: Agentic Turn-based Policy Optimization via Tree Search 26 upvotes, #7 of 2026-01-09
  8. RoboVIP: Multi-View Video Generation with Visual Identity Prompting Augments Robot Manipulation 23 upvotes, #8 of 2026-01-09
  9. Few Tokens Matter: Entropy Guided Attacks on Vision-Language Models 20 upvotes, #9 of 2026-01-09
  10. Agent-as-a-Judge 16 upvotes, #10 of 2026-01-09
  11. VerseCrafter: Dynamic Realistic Video World Model with 4D Geometric Control 16 upvotes, #10 of 2026-01-09
  12. The Illusion of Specialization: Unveiling the Domain-Invariant "Standing Committee" in Mixture-of-Experts Models 15 upvotes, #12 of 2026-01-09
  13. DiffCoT: Diffusion-styled Chain-of-Thought Reasoning in LLMs 12 upvotes, #13 of 2026-01-09
  14. Plenoptic Video Generation 11 upvotes, #14 of 2026-01-09
  15. One Sample to Rule Them All: Extreme Data Efficiency in RL Scaling 8 upvotes, #15 of 2026-01-09
  16. CoV: Chain-of-View Prompting for Spatial Reasoning 8 upvotes, #15 of 2026-01-09
  17. Scaling Behavior Cloning Improves Causal Reasoning: An Open Model for Real-Time Video Game Playing 6 upvotes, #17 of 2026-01-09
  18. Re-Align: Structured Reasoning-guided Alignment for In-Context Image Generation and Editing 5 upvotes, #18 of 2026-01-09
  19. DocDancer: Towards Agentic Document-Grounded Information Seeking 4 upvotes, #19 of 2026-01-09
  20. ReHyAt: Recurrent Hybrid Attention for Video Diffusion Transformers 3 upvotes, #20 of 2026-01-09
  21. ProFuse: Efficient Cross-View Context Fusion for Open-Vocabulary 3D Gaussian Splatting 3 upvotes, #20 of 2026-01-09
  22. Memorization in 3D Shape Generation: An Empirical Study 2 upvotes, #22 of 2026-01-09
  23. Towards Open-Vocabulary Industrial Defect Understanding with a Large-Scale Multimodal Dataset 2 upvotes, #22 of 2026-01-09
  24. Guardians of the Hair: Rescuing Soft Boundaries in Depth, Stereo, and Novel Views 2 upvotes, #22 of 2026-01-09
  25. Beyond Binary Preference: Aligning Diffusion Models to Fine-grained Criteria by Decoupling Attributes 2 upvotes, #22 of 2026-01-09
  26. PyramidalWan: On Making Pretrained Video Model Pyramidal for Efficient Inference 2 upvotes, #22 of 2026-01-09
  27. Multi-Scale Local Speculative Decoding for Image Generation 2 upvotes, #22 of 2026-01-09
  28. Enhancing Object Detection with Privileged Information: A Model-Agnostic Teacher-Student Approach 1 upvotes, #28 of 2026-01-09
  29. Learning User Preferences Through Interaction for Long-Term Collaboration 1 upvotes, #28 of 2026-01-09
  30. LEMAS: Large A 150K-Hour Large-scale Extensible Multilingual Audio Suite with Generative Speech Models 1 upvotes, #28 of 2026-01-09
  31. AgentDevel: Reframing Self-Evolving LLM Agents as Release Engineering 1 upvotes, #28 of 2026-01-09
  32. Safety at One Shot: Patching Fine-Tuned LLMs with A Single Instance 1 upvotes, #32 of 2026-01-09
  33. VERSE: Visual Embedding Reduction and Space Exploration. Clustering-Guided Insights for Training Data Enhancement in Visually-Rich Document Understanding 2 upvotes, #32 of 2026-01-09

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.