Daily Papers of 2026-02-05

  1. ERNIE 5.0 Technical Report 249 upvotes, #1 of 2026-02-05
  2. FASA: Frequency-aware Sparse Attention 146 upvotes, #2 of 2026-02-05
  3. WideSeek-R1: Exploring Width Scaling for Broad Information Seeking via Multi-Agent Reinforcement Learning 92 upvotes, #3 of 2026-02-05
  4. Training Data Efficiency in Multimodal Process Reward Models 75 upvotes, #4 of 2026-02-05
  5. OmniSIFT: Modality-Asymmetric Token Compression for Efficient Omni-modal Large Language Models 46 upvotes, #5 of 2026-02-05
  6. HySparse: A Hybrid Sparse Attention Architecture with Oracle Token Selection and KV Cache Sharing 42 upvotes, #6 of 2026-02-05
  7. EgoActor: Grounding Task Planning into Spatial-aware Egocentric Actions for Humanoid Robots via Visual-Language Models 37 upvotes, #7 of 2026-02-05
  8. Quant VideoGen: Auto-Regressive Long Video Generation via 2-Bit KV-Cache Quantization 33 upvotes, #8 of 2026-02-05
  9. TIDE: Trajectory-based Diagnostic Evaluation of Test-Time Improvement in LLM Agents 32 upvotes, #9 of 2026-02-05
  10. Residual Context Diffusion Language Models 31 upvotes, #10 of 2026-02-05
  11. SoMA: A Real-to-Sim Neural Simulator for Robotic Soft-body Manipulation 31 upvotes, #10 of 2026-02-05
  12. Rethinking the Trust Region in LLM Reinforcement Learning 30 upvotes, #12 of 2026-02-05
  13. Self-Hinting Language Models Enhance Reinforcement Learning 27 upvotes, #13 of 2026-02-05
  14. Semantic Routing: Exploring Multi-Layer LLM Feature Weighting for Diffusion Transformers 27 upvotes, #13 of 2026-02-05
  15. Learning to Repair Lean Proofs from Compiler Feedback 26 upvotes, #15 of 2026-02-05
  16. CL-bench: A Benchmark for Context Learning 22 upvotes, #16 of 2026-02-05
  17. HY3D-Bench: Generation of 3D Assets 22 upvotes, #16 of 2026-02-05
  18. VLS: Steering Pretrained Robot Policies via Vision-Language Models 22 upvotes, #16 of 2026-02-05
  19. AutoFigure: Generating and Refining Publication-Ready Scientific Illustrations 20 upvotes, #19 of 2026-02-05
  20. PaperSearchQA: Learning to Search and Reason over Scientific Papers with RLVR 19 upvotes, #20 of 2026-02-05
  21. A-RAG: Scaling Agentic Retrieval-Augmented Generation via Hierarchical Retrieval Interfaces 19 upvotes, #20 of 2026-02-05
  22. MEnvAgent: Scalable Polyglot Environment Construction for Verifiable Software Engineering 17 upvotes, #22 of 2026-02-05
  23. Vibe AIGC: A New Paradigm for Content Generation via Agentic Orchestration 17 upvotes, #22 of 2026-02-05
  24. Horizon-LM: A RAM-Centric Architecture for LLM Training 16 upvotes, #24 of 2026-02-05
  25. From Data to Behavior: Predicting Unintended Model Behaviors Before Training 15 upvotes, #25 of 2026-02-05
  26. D-CORE: Incentivizing Task Decomposition in Large Reasoning Models for Complex Tool Use 14 upvotes, #26 of 2026-02-05
  27. Agent-Omit: Training Efficient LLM Agents for Adaptive Thought and Observation Omission via Agentic Reinforcement Learning 13 upvotes, #27 of 2026-02-05
  28. Quantifying the Gap between Understanding and Generation within Unified Multimodal Models 12 upvotes, #28 of 2026-02-05
  29. SpatiaLab: Can Vision-Language Models Perform Spatial Reasoning in the Wild? 11 upvotes, #29 of 2026-02-05
  30. MeKi: Memory-based Expert Knowledge Injection for Efficient LLM Scaling 9 upvotes, #30 of 2026-02-05
  31. Efficient Autoregressive Video Diffusion with Dummy Head 8 upvotes, #31 of 2026-02-05
  32. A2Eval: Agentic and Automated Evaluation for Embodied Brain 8 upvotes, #31 of 2026-02-05
  33. BatCoder: Self-Supervised Bidirectional Code-Documentation Learning via Back-Translation 8 upvotes, #31 of 2026-02-05
  34. AgentArk: Distilling Multi-Agent Intelligence into a Single LLM Agent 8 upvotes, #31 of 2026-02-05
  35. Likelihood-Based Reward Designs for General LLM Reasoning 8 upvotes, #31 of 2026-02-05
  36. Beyond Unimodal Shortcuts: MLLMs as Cross-Modal Reasoners for Grounded Named Entity Recognition 6 upvotes, #36 of 2026-02-05
  37. Skin Tokens: A Learned Compact Representation for Unified Autoregressive Rigging 6 upvotes, #36 of 2026-02-05
  38. Self-Rewarding Sequential Monte Carlo for Masked Diffusion Language Models 5 upvotes, #38 of 2026-02-05
  39. Context Learning for Multi-Agent Discussion 4 upvotes, #39 of 2026-02-05
  40. HalluHard: A Hard Multi-Turn Hallucination Benchmark 3 upvotes, #40 of 2026-02-05
  41. Proxy Compression for Language Modeling 3 upvotes, #40 of 2026-02-05
  42. No One-Size-Fits-All: Building Systems For Translation to Bashkir, Kazakh, Kyrgyz, Tatar and Chuvash Using Synthetic And Original Data 3 upvotes, #40 of 2026-02-05
  43. Protein Autoregressive Modeling via Multiscale Structure Generation 3 upvotes, #40 of 2026-02-05
  44. Reward-free Alignment for Conflicting Objectives 2 upvotes, #44 of 2026-02-05
  45. FOTBCD: A Large-Scale Building Change Detection Benchmark from French Orthophotos and Topographic Data 1 upvotes, #45 of 2026-02-05
  46. LongVPO: From Anchored Cues to Self-Reasoning for Long-Form Video Preference Optimization 1 upvotes, #45 of 2026-02-05
  47. "I May Not Have Articulated Myself Clearly": Diagnosing Dynamic Instability in LLM Reasoning at Inference Time 1 upvotes, #45 of 2026-02-05
  48. SkeletonGaussian: Editable 4D Generation through Gaussian Skeletonization 1 upvotes, #45 of 2026-02-05
  49. OmniRad: A Radiological Foundation Model for Multi-Task Medical Image Analysis 1 upvotes, #45 of 2026-02-05
  50. Trust The Typical 1 upvotes, #45 of 2026-02-05
  51. RexBERT: Context Specialized Bidirectional Encoders for E-commerce 1 upvotes, #45 of 2026-02-05
  52. SAFE: Stable Alignment Finetuning with Entropy-Aware Predictive Control for RLHF 1 upvotes, #45 of 2026-02-05
  53. EntRGi: Entropy Aware Reward Guidance for Diffusion Language Models 1 upvotes, #45 of 2026-02-05

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.