Daily Papers of 2024-10-09

  1. Only-IF:Revealing the Decisive Effect of Instruction Diversity on Generalization 17 upvotes, #1 of 2024-10-09
  2. LongGenBench: Long-context Generation Benchmark 16 upvotes, #2 of 2024-10-09
  3. A Spark of Vision-Language Intelligence: 2-Dimensional Autoregressive Transformer for Efficient Finegrained Image Generation 13 upvotes, #3 of 2024-10-09
  4. RevisEval: Improving LLM-as-a-Judge via Response-Adapted References 11 upvotes, #4 of 2024-10-09
  5. DOTS: Learning to Reason Dynamically in LLMs via Optimal Reasoning Trajectories Search 9 upvotes, #5 of 2024-10-09
  6. Inference Scaling for Long-Context Retrieval Augmented Generation 8 upvotes, #6 of 2024-10-09
  7. ControlAR: Controllable Image Generation with Autoregressive Models 7 upvotes, #7 of 2024-10-09
  8. Hyper-multi-step: The Truth Behind Difficult Long-context Tasks 7 upvotes, #7 of 2024-10-09
  9. ε-VAE: Denoising as Visual Decoding 6 upvotes, #9 of 2024-10-09
  10. TidalDecode: Fast and Accurate LLM Decoding with Position Persistent Sparse Attention 6 upvotes, #9 of 2024-10-09
  11. MA-RLHF: Reinforcement Learning from Human Feedback with Macro Actions 5 upvotes, #11 of 2024-10-09
  12. Grounded-VideoLLM: Sharpening Fine-grained Temporal Grounding in Video Large Language Models 5 upvotes, #11 of 2024-10-09
  13. EBES: Easy Benchmarking for Event Sequences 4 upvotes, #13 of 2024-10-09

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.