mingcai yuan

Evidence-R1: Fine-Grained Evidence-Driven Explicit Reasoning and Implicit Reflection for Enhancing RAG Explainability via Reinforcement Learning