Daily Papers of 2024-10-31

  1. CORAL: Benchmarking Multi-turn Conversational Retrieval-Augmentation Generation 52 upvotes, #1 of 2024-10-31
  2. A Large Recurrent Action Model: xLSTM enables Fast Inference for Robotics Tasks 21 upvotes, #2 of 2024-10-31
  3. TokenFormer: Rethinking Transformer Scaling with Tokenized Model Parameters 17 upvotes, #3 of 2024-10-31
  4. ReferEverything: Towards Segmenting Everything We Can Speak of in Videos 17 upvotes, #3 of 2024-10-31
  5. Decoding Reading Goals from Eye Movements 15 upvotes, #5 of 2024-10-31
  6. On Memorization of Large Language Models in Logical Reasoning 14 upvotes, #6 of 2024-10-31
  7. Stealing User Prompts from Mixture of Experts 13 upvotes, #7 of 2024-10-31
  8. Toxicity of the Commons: Curating Open-Source Pre-Training Data 8 upvotes, #8 of 2024-10-31
  9. AutoMIR: Effective Zero-Shot Medical Information Retrieval without Relevance Labels 7 upvotes, #9 of 2024-10-31
  10. SlowFast-VGen: Slow-Fast Learning for Action-Driven Long Video Generation 7 upvotes, #9 of 2024-10-31
  11. Can Models Help Us Create Better Models? Evaluating LLMs as Data Scientists 6 upvotes, #11 of 2024-10-31

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.