Daily Papers of 2025-07-17

1. Towards Agentic RAG with Deep Reasoning: A Survey of RAG-Reasoning Systems in LLMs 73 upvotes, #1 of 2025-07-17
2. SWE-Perf: Can Language Models Optimize Code Performance on Real-World Repositories? 37 upvotes, #2 of 2025-07-17
3. PhysX: Physical-Grounded 3D Asset Generation 37 upvotes, #2 of 2025-07-17
4. DrafterBench: Benchmarking Large Language Models for Tasks Automation in Civil Engineering 30 upvotes, #4 of 2025-07-17
5. Seq vs Seq: An Open Suite of Paired Encoders and Decoders 23 upvotes, #5 of 2025-07-17
6. MMHU: A Massive-Scale Multimodal Benchmark for Human Behavior Understanding 23 upvotes, #5 of 2025-07-17
7. MOSPA: Human Motion Generation Driven by Spatial Audio 20 upvotes, #7 of 2025-07-17
8. Lizard: An Efficient Linearization Framework for Large Language Models 16 upvotes, #8 of 2025-07-17
9. Replacing thinking with tool usage enables reasoning in small language models 14 upvotes, #9 of 2025-07-17
10. SpatialTrackerV2: 3D Point Tracking Made Easy 13 upvotes, #10 of 2025-07-17
11. AnyI2V: Animating Any Conditional Image with Motion Control 11 upvotes, #11 of 2025-07-17
12. GitChameleon: Evaluating AI Code Generation Against Python Library Version Incompatibilities 6 upvotes, #12 of 2025-07-17
13. RLEP: Reinforcement Learning with Experience Replay for LLM Reasoning 3 upvotes, #13 of 2025-07-17
14. AI Wizards at CheckThat! 2025: Enhancing Transformer-Based Embeddings with Sentiment for Subjectivity Detection in News Articles 2 upvotes, #14 of 2025-07-17
15. MST-Distill: Mixture of Specialized Teachers for Cross-Modal Knowledge Distillation 1 upvotes, #15 of 2025-07-17
16. (Almost) Free Modality Stitching of Foundation Models 2 upvotes, #15 of 2025-07-17

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.