Daily Papers of 2025-07-28

1. Deep Researcher with Test-Time Diffusion 53 upvotes, #1 of 2025-07-28
2. The Geometry of LLM Quantization: GPTQ as Babai's Nearest Plane Algorithm 37 upvotes, #2 of 2025-07-28
3. MMBench-GUI: Hierarchical Multi-Platform Evaluation Framework for GUI Agents 28 upvotes, #3 of 2025-07-28
4. When Tokens Talk Too Much: A Survey of Multimodal Long-Context Token Compression across Images, Videos, and Audios 24 upvotes, #4 of 2025-07-28
5. GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning 20 upvotes, #5 of 2025-07-28
6. CLEAR: Error Analysis via LLM-as-a-Judge Made Easy 17 upvotes, #6 of 2025-07-28
7. PRIX: Learning to Plan from Raw Pixels for End-to-End Autonomous Driving 5 upvotes, #7 of 2025-07-28
8. Specification Self-Correction: Mitigating In-Context Reward Hacking Through Test-Time Refinement 5 upvotes, #7 of 2025-07-28
9. Chat with AI: The Surprising Turn of Real-time Video Communication from Human to AI 4 upvotes, #9 of 2025-07-28
10. Frontier AI Risk Management Framework in Practice: A Risk Analysis Technical Report 4 upvotes, #9 of 2025-07-28
11. AFRDA: Attentive Feature Refinement for Domain Adaptive Semantic Segmentation 1 upvotes, #11 of 2025-07-28

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.