Daily Papers of 2024-06-28

  1. OMG-LLaVA: Bridging Image-level, Object-level, Pixel-level Reasoning and Understanding 50 upvotes, #1 of 2024-06-28
  2. Step-DPO: Step-wise Preference Optimization for Long-chain Reasoning of LLMs 37 upvotes, #2 of 2024-06-28
  3. MUMU: Bootstrapping Multimodal Image Generation from Text-to-Image Data 32 upvotes, #3 of 2024-06-28
  4. Simulating Classroom Education with LLM-Empowered Agents 27 upvotes, #4 of 2024-06-28
  5. SeaKR: Self-aware Knowledge Retrieval for Adaptive Retrieval Augmented Generation 26 upvotes, #5 of 2024-06-28
  6. Aligning Teacher with Student Preferences for Tailored Training Data Generation 22 upvotes, #6 of 2024-06-28
  7. Can LLMs Learn by Teaching? A Preliminary Study 17 upvotes, #7 of 2024-06-28
  8. Dataset Size Recovery from LoRA Weights 17 upvotes, #7 of 2024-06-28
  9. MoA: Mixture of Sparse Attention for Automatic Large Language Model Compression 12 upvotes, #9 of 2024-06-28
  10. LiveBench: A Challenging, Contamination-Free LLM Benchmark 12 upvotes, #9 of 2024-06-28
  11. Is Programming by Example solved by LLMs? 11 upvotes, #11 of 2024-06-28
  12. AUTOHALLUSION: Automatic Generation of Hallucination Benchmarks for Vision-Language Models 11 upvotes, #11 of 2024-06-28
  13. Read Anywhere Pointed: Layout-aware GUI Screen Reading with Tree-of-Lens Grounding 9 upvotes, #13 of 2024-06-28
  14. T-FREE: Tokenizer-Free Generative LLMs via Sparse Representations for Memory-Efficient Embeddings 8 upvotes, #14 of 2024-06-28
  15. Understand What LLM Needs: Dual Preference Alignment for Retrieval-Augmented Generation 5 upvotes, #15 of 2024-06-28
  16. ArzEn-LLM: Code-Switched Egyptian Arabic-English Translation and Speech Recognition Using LLMs 4 upvotes, #16 of 2024-06-28
  17. Benchmarking Mental State Representations in Language Models 3 upvotes, #17 of 2024-06-28
  18. ResumeAtlas: Revisiting Resume Classification with Large-Scale Datasets and Large Language Models 3 upvotes, #17 of 2024-06-28

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.