Daily Papers of 2025-03-25

  1. I Have Covered All the Bases Here: Interpreting Reasoning Features in Large Language Models via Sparse Autoencoders 110 upvotes, #1 of 2025-03-25
  2. Video-T1: Test-Time Scaling for Video Generation 84 upvotes, #2 of 2025-03-25
  3. Position: Interactive Generative Video as Next-Generation Game Engine 59 upvotes, #3 of 2025-03-25
  4. SimpleRL-Zoo: Investigating and Taming Zero Reinforcement Learning for Open Base Models in the Wild 27 upvotes, #4 of 2025-03-25
  5. Aether: Geometric-Aware Unified World Modeling 26 upvotes, #5 of 2025-03-25
  6. OmnimatteZero: Training-free Real-time Omnimatte with Pre-trained Video Diffusion Models 23 upvotes, #6 of 2025-03-25
  7. AgentRxiv: Towards Collaborative Autonomous Research 21 upvotes, #7 of 2025-03-25
  8. Judge Anything: MLLM as a Judge Across Any Modality 19 upvotes, #8 of 2025-03-25
  9. Vision-R1: Evolving Human-Free Alignment in Large Vision-Language Models via Vision-Guided Reinforcement Learning 18 upvotes, #9 of 2025-03-25
  10. Defeating Prompt Injections by Design 18 upvotes, #9 of 2025-03-25
  11. CFG-Zero*: Improved Classifier-Free Guidance for Flow Matching Models 18 upvotes, #9 of 2025-03-25
  12. FFN Fusion: Rethinking Sequential Computation in Large Language Models 16 upvotes, #12 of 2025-03-25
  13. Equivariant Image Modeling 14 upvotes, #13 of 2025-03-25
  14. Optimized Minimal 3D Gaussian Splatting 13 upvotes, #14 of 2025-03-25
  15. LEMMA: Learning from Errors for MatheMatical Advancement in LLMs 13 upvotes, #14 of 2025-03-25
  16. Feather-SQL: A Lightweight NL2SQL Framework with Dual-Model Collaboration Paradigm for Small Language Models 13 upvotes, #14 of 2025-03-25
  17. Reasoning to Learn from Latent Thoughts 13 upvotes, #14 of 2025-03-25
  18. Training-free Diffusion Acceleration with Bottleneck Sampling 12 upvotes, #18 of 2025-03-25
  19. Video SimpleQA: Towards Factuality Evaluation in Large Video Language Models 11 upvotes, #19 of 2025-03-25
  20. AlphaSpace: Enabling Robotic Actions through Semantic Tokenization and Symbolic Reasoning 9 upvotes, #20 of 2025-03-25
  21. MagicComp: Training-free Dual-Phase Refinement for Compositional Video Generation 8 upvotes, #21 of 2025-03-25
  22. Typed-RAG: Type-aware Multi-Aspect Decomposition for Non-Factoid Question Answering 6 upvotes, #22 of 2025-03-25
  23. Lost in Cultural Translation: Do LLMs Struggle with Math Across Cultural Contexts? 6 upvotes, #22 of 2025-03-25
  24. Diffusion-4K: Ultra-High-Resolution Image Synthesis with Latent Diffusion Models 6 upvotes, #22 of 2025-03-25
  25. V-Seek: Accelerating LLM Reasoning on Open-hardware Server-class RISC-V Platforms 5 upvotes, #25 of 2025-03-25
  26. AMD-Hummingbird: Towards an Efficient Text-to-Video Model 5 upvotes, #25 of 2025-03-25
  27. Variance Control via Weight Rescaling in LLM Pre-training 4 upvotes, #27 of 2025-03-25
  28. RDTF: Resource-efficient Dual-mask Training Framework for Multi-frame Animated Sticker Generation 3 upvotes, #28 of 2025-03-25
  29. CODA: Repurposing Continuous VAEs for Discrete Tokenization 3 upvotes, #28 of 2025-03-25
  30. Mind with Eyes: from Language Reasoning to Multimodal Reasoning 3 upvotes, #28 of 2025-03-25
  31. Instruct-CLIP: Improving Instruction-Guided Image Editing with Automated Data Refinement Using Contrastive Learning 3 upvotes, #28 of 2025-03-25
  32. MetaSpatial: Reinforcing 3D Spatial Reasoning in VLMs for the Metaverse 3 upvotes, #28 of 2025-03-25
  33. Verbal Process Supervision Elicits Better Coding Agents 2 upvotes, #33 of 2025-03-25
  34. Rethinking Image Evaluation in Super-Resolution 1 upvotes, #34 of 2025-03-25
  35. Revisiting Image Fusion for Multi-Illuminant White-Balance Correction 1 upvotes, #34 of 2025-03-25
  36. Human Motion Unlearning 1 upvotes, #34 of 2025-03-25
  37. DynamicVis: An Efficient and General Visual Foundation Model for Remote Sensing Image Understanding 0 upvotes, #37 of 2025-03-25
  38. QuartDepth: Post-Training Quantization for Real-Time Depth Estimation on the Edge 0 upvotes, #37 of 2025-03-25
  39. Global-Local Tree Search for Language Guided 3D Scene Generation 0 upvotes, #37 of 2025-03-25

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.