go back ..
Retrieval Augmented Generation
Retrieval Augmented Generation
RAG
Retrieval-Augmented Generation (RAG) is a promising approach that leverages the strengths of information retrieval systems and reduces the hallucinations of modern large language models (LLMs) through the incorporation of external knowledge in order to provide more accurate and relevant responses. However, the influence of RAG on LLMs has not been rigorously evaluated in the literature, which makes it difficult to pinpoint any potential bottlenecks in RAG's capabilities for state-of-the-art LLMs . Given the growing research interest in RAG, a dedicated study of this new field applied to LLMs is needed.