Daily Papers of 2025-05-06

  1. Grokking in the Wild: Data Augmentation for Real-World Multi-Hop Reasoning with Transformers 86 upvotes, #1 of 2025-05-06
  2. Voila: Voice-Language Foundation Models for Real-Time Autonomous Interaction and Voice Role-Play 79 upvotes, #2 of 2025-05-06
  3. RM-R1: Reward Modeling as Reasoning 66 upvotes, #3 of 2025-05-06
  4. Practical Efficiency of Muon for Pretraining 36 upvotes, #4 of 2025-05-06
  5. Agentic Reasoning and Tool Integration for LLMs via Reinforcement Learning 35 upvotes, #5 of 2025-05-06
  6. A Survey on Inference Engines for Large Language Models: Perspectives on Optimization and Efficiency 31 upvotes, #6 of 2025-05-06
  7. FormalMATH: Benchmarking Formal Mathematical Reasoning of Large Language Models 27 upvotes, #7 of 2025-05-06
  8. ReplaceMe: Network Simplification via Layer Pruning and Linear Transformations 24 upvotes, #8 of 2025-05-06
  9. Optimizing Chain-of-Thought Reasoners via Gradient Variance Minimization in Rejection Sampling and RL 22 upvotes, #9 of 2025-05-06
  10. R1-Reward: Training Multimodal Reward Model Through Stable Reinforcement Learning 22 upvotes, #9 of 2025-05-06
  11. LLaMA-Omni2: LLM-based Real-time Spoken Chatbot with Autoregressive Streaming Speech Synthesis 20 upvotes, #11 of 2025-05-06
  12. SkillMimic-V2: Learning Robust and Generalizable Interaction Skills from Sparse and Noisy Demonstrations 17 upvotes, #12 of 2025-05-06
  13. Think on your Feet: Adaptive Thinking via Reinforcement Learning for Social Agents 17 upvotes, #12 of 2025-05-06
  14. SuperEdit: Rectifying and Facilitating Supervision for Instruction-Based Image Editing 12 upvotes, #14 of 2025-05-06
  15. Ming-Lite-Uni: Advancements in Unified Architecture for Natural Multimodal Interaction 12 upvotes, #14 of 2025-05-06
  16. Low-Precision Training of Large Language Models: Methods, Challenges, and Opportunities 9 upvotes, #16 of 2025-05-06
  17. TEMPURA: Temporal Event Masked Prediction and Understanding for Reasoning in Action 9 upvotes, #16 of 2025-05-06
  18. MUSAR: Exploring Multi-Subject Customization from Single-Subject Dataset via Attention Routing 5 upvotes, #18 of 2025-05-06
  19. Learning Heterogeneous Mixture of Scene Experts for Large-scale Neural Radiance Fields 3 upvotes, #19 of 2025-05-06
  20. Attention Mechanisms Perspective: Exploring LLM Processing of Graph-Structured Data 3 upvotes, #19 of 2025-05-06
  21. Unlearning Sensitive Information in Multimodal LLMs: Benchmark and Attack-Defense Evaluation 2 upvotes, #21 of 2025-05-06
  22. Rethinking RGB-Event Semantic Segmentation with a Novel Bidirectional Motion-enhanced Event Representation 1 upvotes, #22 of 2025-05-06

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.