Daily Papers of 2025-04-25

  1. Paper2Code: Automating Code Generation from Scientific Papers in Machine Learning 100 upvotes, #1 of 2025-04-25
  2. Step1X-Edit: A Practical Framework for General Image Editing 85 upvotes, #2 of 2025-04-25
  3. RefVNLI: Towards Scalable Evaluation of Subject-driven Text-to-image Generation 54 upvotes, #3 of 2025-04-25
  4. Breaking the Modality Barrier: Universal Embedding Learning with Multimodal LLMs 38 upvotes, #4 of 2025-04-25
  5. Perspective-Aware Reasoning in Vision-Language Models via Mental Imagery Simulation 28 upvotes, #5 of 2025-04-25
  6. Token-Shuffle: Towards High-Resolution Image Generation with Autoregressive Models 23 upvotes, #6 of 2025-04-25
  7. QuaDMix: Quality-Diversity Balanced Data Selection for Efficient LLM Pretraining 20 upvotes, #7 of 2025-04-25
  8. DiMeR: Disentangled Mesh Reconstruction Model 20 upvotes, #7 of 2025-04-25
  9. Process Reward Models That Think 16 upvotes, #9 of 2025-04-25
  10. Boosting Generative Image Modeling via Joint Image-Feature Synthesis 14 upvotes, #10 of 2025-04-25
  11. DyMU: Dynamic Merging and Virtual Unmerging for Efficient VLMs 13 upvotes, #11 of 2025-04-25
  12. TimeChat-Online: 80% Visual Tokens are Naturally Redundant in Streaming Videos 10 upvotes, #12 of 2025-04-25
  13. 3DV-TON: Textured 3D-Guided Consistent Video Try-on via Diffusion Models 10 upvotes, #12 of 2025-04-25
  14. ViSMaP: Unsupervised Hour-long Video Summarisation by Meta-Prompting 7 upvotes, #14 of 2025-04-25
  15. IberBench: LLM Evaluation on Iberian Languages 7 upvotes, #14 of 2025-04-25
  16. Distilling semantically aware orders for autoregressive image generation 5 upvotes, #16 of 2025-04-25
  17. Dynamic Camera Poses and Where to Find Them 4 upvotes, #17 of 2025-04-25
  18. Interpretable non-linear dimensionality reduction using gaussian weighted linear transformation 2 upvotes, #18 of 2025-04-25

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.