Derong Xu

Process vs. Outcome Reward: Which is Better for Agentic RAG Reinforcement Learning