Daily Papers of 2025-07-21

  1. The Devil behind the mask: An emergent safety vulnerability of Diffusion LLMs 56 upvotes, #1 of 2025-07-21
  2. A Data-Centric Framework for Addressing Phonetic and Prosodic Challenges in Russian Speech Generative Models 48 upvotes, #2 of 2025-07-21
  3. Franca: Nested Matryoshka Clustering for Scalable Visual Representation Learning 27 upvotes, #3 of 2025-07-21
  4. Inverse Reinforcement Learning Meets Large Language Model Post-Training: Basics, Advances, and Opportunities 24 upvotes, #4 of 2025-07-21
  5. CSD-VAR: Content-Style Decomposition in Visual Autoregressive Models 21 upvotes, #5 of 2025-07-21
  6. Mono-InternVL-1.5: Towards Cheaper and Faster Monolithic Multimodal Large Language Models 14 upvotes, #6 of 2025-07-21
  7. OpenBEATs: A Fully Open-Source General-Purpose Audio Encoder 8 upvotes, #7 of 2025-07-21
  8. RedOne: Revealing Domain-specific LLM Post-Training in Social Networking Services 7 upvotes, #8 of 2025-07-21
  9. Mitigating Object Hallucinations via Sentence-Level Early Intervention 6 upvotes, #9 of 2025-07-21
  10. The Generative Energy Arena (GEA): Incorporating Energy Awareness in Large Language Model (LLM) Human Evaluations 4 upvotes, #10 of 2025-07-21
  11. Quantitative Risk Management in Volatile Markets with an Expectile-Based Framework for the FTSE Index 4 upvotes, #10 of 2025-07-21

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.