Daily Papers of 2025-07-22

  1. GUI-G^2: Gaussian Reward Modeling for GUI Grounding 122 upvotes, #1 of 2025-07-22
  2. MiroMind-M1: An Open-Source Advancement in Mathematical Reasoning via Context-Aware Multi-Stage Policy Optimization 116 upvotes, #2 of 2025-07-22
  3. The Invisible Leash: Why RLVR May Not Escape Its Origin 81 upvotes, #3 of 2025-07-22
  4. NoHumansRequired: Autonomous High-Quality Image Editing Triplet Mining 53 upvotes, #4 of 2025-07-22
  5. WebShaper: Agentically Data Synthesizing via Information-Seeking Formalization 44 upvotes, #5 of 2025-07-22
  6. GR-3 Technical Report 44 upvotes, #5 of 2025-07-22
  7. Robust 3D-Masked Part-level Editing in 3D Gaussian Splatting with Regularized Score Distillation Sampling 37 upvotes, #7 of 2025-07-22
  8. SeC: Advancing Complex Video Object Segmentation via Progressive Concept Construction 37 upvotes, #7 of 2025-07-22
  9. Being-H0: Vision-Language-Action Pretraining from Large-Scale Human Videos 33 upvotes, #9 of 2025-07-22
  10. Inverse Scaling in Test-Time Compute 25 upvotes, #10 of 2025-07-22
  11. STITCH: Simultaneous Thinking and Talking with Chunked Reasoning for Spoken Language Models 25 upvotes, #10 of 2025-07-22
  12. Gaussian Splatting with Discretized SDF for Relightable Assets 22 upvotes, #12 of 2025-07-22
  13. Towards Video Thinking Test: A Holistic Benchmark for Advanced Video Reasoning and Understanding 20 upvotes, #13 of 2025-07-22
  14. Stabilizing Knowledge, Promoting Reasoning: Dual-Token Constraints for RLVR 19 upvotes, #14 of 2025-07-22
  15. MCPEval: Automatic MCP-based Deep Evaluation for AI Agent Models 16 upvotes, #15 of 2025-07-22
  16. Streaming 4D Visual Geometry Transformer 14 upvotes, #16 of 2025-07-22
  17. "PhyWorldBench": A Comprehensive Evaluation of Physical Realism in Text-to-Video Models 14 upvotes, #16 of 2025-07-22
  18. A Simple "Try Again" Can Elicit Multi-Turn LLM Reasoning 13 upvotes, #18 of 2025-07-22
  19. Latent Denoising Makes Good Visual Tokenizers 9 upvotes, #19 of 2025-07-22
  20. The Serial Scaling Hypothesis 8 upvotes, #20 of 2025-07-22
  21. TokensGen: Harnessing Condensed Tokens for Long Video Generation 6 upvotes, #21 of 2025-07-22
  22. LLM Economist: Large Population Models and Mechanism Design in Multi-Agent Generative Simulacra 6 upvotes, #21 of 2025-07-22
  23. PhysGym: Benchmarking LLMs in Interactive Physics Discovery with Controlled Priors 4 upvotes, #23 of 2025-07-22
  24. Data Mixing Agent: Learning to Re-weight Domains for Continual Pre-training 3 upvotes, #24 of 2025-07-22
  25. GeoDistill: Geometry-Guided Self-Distillation for Weakly Supervised Cross-View Localization 1 upvotes, #25 of 2025-07-22
  26. ParaStudent: Generating and Evaluating Realistic Student Code by Teaching LLMs to Struggle 1 upvotes, #26 of 2025-07-22
  27. UGPL: Uncertainty-Guided Progressive Learning for Evidence-Based Classification in Computed Tomography 1 upvotes, #26 of 2025-07-22

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.