Daily Papers of 2024-10-28

  1. ROCKET-1: Master Open-World Interaction with Visual-Temporal Context Prompting 48 upvotes, #1 of 2024-10-28
  2. Continuous Speech Synthesis using per-token Latent Diffusion 28 upvotes, #2 of 2024-10-28
  3. Teach Multimodal LLMs to Comprehend Electrocardiographic Images 22 upvotes, #3 of 2024-10-28
  4. FasterCache: Training-Free Video Diffusion Model Acceleration with High Quality 20 upvotes, #4 of 2024-10-28
  5. MMAU: A Massive Multi-Task Audio Understanding and Reasoning Benchmark 19 upvotes, #5 of 2024-10-28
  6. Infinity-MM: Scaling Multimodal Performance with Large-Scale and High-Quality Instruction Data 17 upvotes, #6 of 2024-10-28
  7. Are LLMs Better than Reported? Detecting Label Errors and Mitigating Their Effect on Model Performance 15 upvotes, #7 of 2024-10-28
  8. Read-ME: Refactorizing LLMs as Router-Decoupled Mixture of Experts with System Co-Design 15 upvotes, #7 of 2024-10-28
  9. Hybrid Preferences: Learning to Route Instances for Human vs. AI Feedback 10 upvotes, #9 of 2024-10-28
  10. Fictitious Synthetic Data Can Improve LLM Factuality via Prerequisite Learning 10 upvotes, #9 of 2024-10-28
  11. Counting Ability of Large Language Models and Impact of Tokenization 10 upvotes, #9 of 2024-10-28
  12. Analysing the Residual Stream of Language Models Under Knowledge Conflicts 6 upvotes, #12 of 2024-10-28
  13. Dynamic 3D Gaussian Tracking for Graph-Based Neural Dynamics Modeling 6 upvotes, #12 of 2024-10-28
  14. Reflection-Bench: probing AI intelligence with reflection 5 upvotes, #14 of 2024-10-28
  15. Mapping the Media Landscape: Predicting Factual Reporting and Political Bias Through Web Interactions 5 upvotes, #14 of 2024-10-28
  16. Leveraging Skills from Unlabeled Prior Data for Efficient Online Exploration 4 upvotes, #16 of 2024-10-28

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.