Chain-of-Retrieval Augmented Generation

Liang Wang, Haonan Chen, Nan Yang +3

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. Conventional R…