PhD student, Renmin University of China
1 paper at NeurIPS 2025
This paper introduces an approach for training o1-like RAG models that retrieve and reason over relevant information step by step before generating the final answer.
