Ground Every Sentence: Improving Retrieval-Augmented LLMs with Interleaved Reference-Claim Generation
This work addresses the need for improved credibility and verifiability in RAG systems for knowledge-intensive applications, representing an incremental advancement over prior coarse-grained attribution methods.
The paper tackles the problem of coarse-grained attributions in Retrieval-Augmented Generation (RAG) systems by proposing ReClaim, a fine-grained method that provides sentence-level citations, achieving a citation accuracy rate of 90% in long-form question-answering tasks.
Retrieval-Augmented Generation (RAG) has been widely adopted to enhance Large Language Models (LLMs) in knowledge-intensive tasks. To enhance credibility and verifiability in RAG systems, Attributed Text Generation (ATG) is proposed, which provides citations to retrieval knowledge in LLM-generated responses. Prior methods mainly adopt coarse-grained attributions, with passage-level or paragraph-level references or citations, which fall short in verifiability. This paper proposes ReClaim (Refer & Claim), a fine-grained ATG method that alternates the generation of references and answers step by step. Different from previous coarse-grained attribution, ReClaim provides sentence-level citations in long-form question-answering tasks. With extensive experiments, we verify the effectiveness of ReClaim in extensive settings, achieving a citation accuracy rate of 90%.