Daily Papers of 2025-05-20

  1. Chain-of-Model Learning for Language Model 107 upvotes, #1 of 2025-05-20
  2. AdaptThink: Reasoning Models Can Learn When to Think 72 upvotes, #2 of 2025-05-20
  3. AdaCoT: Pareto-Optimal Adaptive Chain-of-Thought Triggering via Reinforcement Learning 54 upvotes, #3 of 2025-05-20
  4. Delta Attention: Fast and Accurate Sparse Attention Inference by Delta Correction 47 upvotes, #4 of 2025-05-20
  5. Thinkless: LLM Learns When to Think 46 upvotes, #5 of 2025-05-20
  6. Scaling Computer-Use Grounding via User Interface Decomposition and Synthesis 44 upvotes, #6 of 2025-05-20
  7. Model Merging in Pre-training of Large Language Models 35 upvotes, #7 of 2025-05-20
  8. Faster Video Diffusion with Trainable Sparse Attention 34 upvotes, #8 of 2025-05-20
  9. Through the Looking Glass: Common Sense Consistency Evaluation of Weird Images 29 upvotes, #9 of 2025-05-20
  10. Hybrid 3D-4D Gaussian Splatting for Fast Dynamic Scene Representation 27 upvotes, #10 of 2025-05-20
  11. Seek in the Dark: Reasoning via Test-Time Instance-Level Policy Gradient in Latent Space 26 upvotes, #11 of 2025-05-20
  12. MM-PRM: Enhancing Multimodal Mathematical Reasoning with Scalable Step-Level Supervision 25 upvotes, #12 of 2025-05-20
  13. CPGD: Toward Stable Rule-based Reinforcement Learning for Language Models 23 upvotes, #13 of 2025-05-20
  14. FedSVD: Adaptive Orthogonalization for Private Federated Learning with LoRA 21 upvotes, #14 of 2025-05-20
  15. Fractured Chain-of-Thought Reasoning 21 upvotes, #14 of 2025-05-20
  16. EfficientLLM: Efficiency in Large Language Models 21 upvotes, #14 of 2025-05-20
  17. SEED-GRPO: Semantic Entropy Enhanced GRPO for Uncertainty-Aware Policy Optimization 19 upvotes, #17 of 2025-05-20
  18. Neuro-Symbolic Query Compiler 16 upvotes, #18 of 2025-05-20
  19. ChartMuseum: Testing Visual Reasoning Capabilities of Large Vision-Language Models 16 upvotes, #18 of 2025-05-20
  20. VisionReasoner: Unified Visual Perception and Reasoning via Reinforcement Learning 15 upvotes, #20 of 2025-05-20
  21. ViPlan: A Benchmark for Visual Planning with Symbolic Predicates and Vision-Language Models 13 upvotes, #21 of 2025-05-20
  22. R3: Robust Rubric-Agnostic Reward Models 11 upvotes, #22 of 2025-05-20
  23. When AI Co-Scientists Fail: SPOT-a Benchmark for Automated Verification of Scientific Research 9 upvotes, #23 of 2025-05-20
  24. MTVCrafter: 4D Motion Tokenization for Open-World Human Image Animation 8 upvotes, #24 of 2025-05-20
  25. Efficient Speech Language Modeling via Energy Distance in Continuous Latent Space 8 upvotes, #24 of 2025-05-20
  26. Accelerate TarFlow Sampling with GS-Jacobi Iteration 7 upvotes, #26 of 2025-05-20
  27. MedCaseReasoning: Evaluating and learning diagnostic reasoning from clinical case reports 6 upvotes, #27 of 2025-05-20
  28. Tiny QA Benchmark++: Ultra-Lightweight, Synthetic Multilingual Dataset Generation & Smoke-Tests for Continuous LLM Evaluation 6 upvotes, #27 of 2025-05-20
  29. SoftCoT++: Test-Time Scaling with Soft Chain-of-Thought Reasoning 5 upvotes, #29 of 2025-05-20
  30. FinePhys: Fine-grained Human Action Generation by Explicitly Incorporating Physical Laws for Effective Skeletal Guidance 5 upvotes, #29 of 2025-05-20
  31. QVGen: Pushing the Limit of Quantized Video Generative Models 4 upvotes, #31 of 2025-05-20
  32. Creating General User Models from Computer Use 3 upvotes, #32 of 2025-05-20
  33. HISTAI: An Open-Source, Large-Scale Whole Slide Image Dataset for Computational Pathology 3 upvotes, #32 of 2025-05-20
  34. ExTrans: Multilingual Deep Reasoning Translation via Exemplar-Enhanced Reinforcement Learning 3 upvotes, #32 of 2025-05-20
  35. Learned Lightweight Smartphone ISP with Unpaired Data 2 upvotes, #35 of 2025-05-20
  36. HelpSteer3-Preference: Open Human-Annotated Preference Data across Diverse Tasks and Languages 2 upvotes, #35 of 2025-05-20
  37. TechniqueRAG: Retrieval Augmented Generation for Adversarial Technique Annotation in Cyber Threat Intelligence Text 2 upvotes, #35 of 2025-05-20
  38. A Token is Worth over 1,000 Tokens: Efficient Knowledge Distillation through Low-Rank Clone 2 upvotes, #35 of 2025-05-20
  39. From Grunts to Grammar: Emergent Language from Cooperative Foraging 2 upvotes, #35 of 2025-05-20
  40. AI-Driven Scholarly Peer Review via Persistent Workflow Prompting, Meta-Prompting, and Meta-Reasoning 1 upvotes, #40 of 2025-05-20
  41. LLM Context Conditioning and PWP Prompting for Multimodal Validation of Chemical Formulas 1 upvotes, #40 of 2025-05-20
  42. Can AI Freelancers Compete? Benchmarking Earnings, Reliability, and Task Success at Scale 1 upvotes, #40 of 2025-05-20
  43. Fast, Not Fancy: Rethinking G2P with Rich Data and Rule-Based Models 3 upvotes, #43 of 2025-05-20

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.