Kuicai Dong

Benchmarking Retrieval-Augmented Multimomal Generation for Document Question Answering

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