Retrieving, Rethinking and Revising: The Chain-of-Verification Can Improve Retrieval Augmented Generation

TL;DR

This paper introduces CoV-RAG, a chain-of-verification method that improves retrieval accuracy and answer consistency in retrieval-augmented generation by integrating verification, rewriting, and reasoning modules, significantly outperforming existing baselines.

Contribution

The paper proposes a novel chain-of-verification framework for RAG that enhances retrieval correctness and generation consistency through integrated verification and reasoning modules.

Findings

CoV-RAG significantly outperforms state-of-the-art baselines.

The verification module improves retrieval accuracy.

Unified QA and verification with Chain-of-Thought enhances consistency.

Abstract

Recent Retrieval Augmented Generation (RAG) aims to enhance Large Language Models (LLMs) by incorporating extensive knowledge retrieved from external sources. However, such approach encounters some challenges: Firstly, the original queries may not be suitable for precise retrieval, resulting in erroneous contextual knowledge; Secondly, the language model can easily generate inconsistent answer with external references due to their knowledge boundary limitation. To address these issues, we propose the chain-of-verification (CoV-RAG) to enhance the external retrieval correctness and internal generation consistency. Specifically, we integrate the verification module into the RAG, engaging in scoring, judgment, and rewriting. To correct external retrieval errors, CoV-RAG retrieves new knowledge using a revised query. To correct internal generation errors, we unify QA and verification tasks with a Chain-of-Thought (CoT) reasoning during training. Our comprehensive experiments across various LLMs demonstrate the effectiveness and adaptability compared with other strong baselines. Especially, our CoV-RAG can significantly surpass the state-of-the-art baselines using different LLM backbones.