Daily Papers of 2025-11-10

1. Too Good to be Bad: On the Failure of LLMs to Role-Play Villains 50 upvotes, #1 of 2025-11-10
2. Visual Spatial Tuning 46 upvotes, #2 of 2025-11-10
3. DeepEyesV2: Toward Agentic Multimodal Model 38 upvotes, #3 of 2025-11-10
4. VeriCoT: Neuro-symbolic Chain-of-Thought Validation via Logical Consistency Checks 34 upvotes, #4 of 2025-11-10
5. Real-Time Reasoning Agents in Evolving Environments 11 upvotes, #5 of 2025-11-10
6. Dense Motion Captioning 9 upvotes, #6 of 2025-11-10
7. Towards Mitigating Hallucinations in Large Vision-Language Models by Refining Textual Embeddings 7 upvotes, #7 of 2025-11-10
8. CritiCal: Can Critique Help LLM Uncertainty or Confidence Calibration? 3 upvotes, #8 of 2025-11-10
9. HAFixAgent: History-Aware Automated Program Repair Agent 3 upvotes, #8 of 2025-11-10
10. Jailbreaking in the Haystack 3 upvotes, #8 of 2025-11-10

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.