Daily Papers of 2024-12-19

  1. Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference 105 upvotes, #1 of 2024-12-19
  2. AniDoc: Animation Creation Made Easier 48 upvotes, #2 of 2024-12-19
  3. TheAgentCompany: Benchmarking LLM Agents on Consequential Real World Tasks 43 upvotes, #3 of 2024-12-19
  4. No More Adam: Learning Rate Scaling at Initialization is All You Need 41 upvotes, #4 of 2024-12-19
  5. GUI Agents: A Survey 22 upvotes, #5 of 2024-12-19
  6. Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces 22 upvotes, #5 of 2024-12-19
  7. Mix-LN: Unleashing the Power of Deeper Layers by Combining Pre-LN and Post-LN 18 upvotes, #7 of 2024-12-19
  8. LLaVA-UHD v2: an MLLM Integrating High-Resolution Feature Pyramid via Hierarchical Window Transformer 17 upvotes, #8 of 2024-12-19
  9. FashionComposer: Compositional Fashion Image Generation 16 upvotes, #9 of 2024-12-19
  10. FastVLM: Efficient Vision Encoding for Vision Language Models 13 upvotes, #10 of 2024-12-19
  11. Autoregressive Video Generation without Vector Quantization 13 upvotes, #10 of 2024-12-19
  12. Prompting Depth Anything for 4K Resolution Accurate Metric Depth Estimation 12 upvotes, #12 of 2024-12-19
  13. Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for Multitask Learning 11 upvotes, #13 of 2024-12-19
  14. AnySat: An Earth Observation Model for Any Resolutions, Scales, and Modalities 11 upvotes, #13 of 2024-12-19
  15. Learning from Massive Human Videos for Universal Humanoid Pose Control 10 upvotes, #15 of 2024-12-19
  16. RAG-RewardBench: Benchmarking Reward Models in Retrieval Augmented Generation for Preference Alignment 9 upvotes, #16 of 2024-12-19
  17. ChatDiT: A Training-Free Baseline for Task-Agnostic Free-Form Chatting with Diffusion Transformers 8 upvotes, #17 of 2024-12-19
  18. VidTok: A Versatile and Open-Source Video Tokenizer 7 upvotes, #18 of 2024-12-19
  19. Alignment faking in large language models 7 upvotes, #18 of 2024-12-19
  20. CAD-Recode: Reverse Engineering CAD Code from Point Clouds 5 upvotes, #20 of 2024-12-19
  21. Predicting the Original Appearance of Damaged Historical Documents 4 upvotes, #21 of 2024-12-19
  22. AntiLeak-Bench: Preventing Data Contamination by Automatically Constructing Benchmarks with Updated Real-World Knowledge 4 upvotes, #21 of 2024-12-19

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.