HER: Human-like Reasoning and Reinforcement Learning for LLM Role-playing Paper • 2601.21459 • Published 19 days ago • 9