FaithfulRAG: Fact-Level Conflict Modeling for Context-Faithful Retrieval-Augmented Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Qinggang, Xiang, Zhishang, Xiao, Yilin, Wang, Le, Li, Junhui, Wang, Xinrun, Su, Jinsong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908439646568448
author Zhang, Qinggang
Xiang, Zhishang
Xiao, Yilin
Wang, Le
Li, Junhui
Wang, Xinrun
Su, Jinsong
author_facet Zhang, Qinggang
Xiang, Zhishang
Xiao, Yilin
Wang, Le
Li, Junhui
Wang, Xinrun
Su, Jinsong
contents Large language models (LLMs) augmented with retrieval systems have demonstrated significant potential in handling knowledge-intensive tasks. However, these models often struggle with unfaithfulness issues, generating outputs that either ignore the retrieved context or inconsistently blend it with the LLM`s parametric knowledge. This issue is particularly severe in cases of knowledge conflict, where the retrieved context conflicts with the model`s parametric knowledge. While existing faithful RAG approaches enforce strict context adherence through well-designed prompts or modified decoding strategies, our analysis reveals a critical limitation: they achieve faithfulness by forcibly suppressing the model`s parametric knowledge, which undermines the model`s internal knowledge structure and increases the risk of misinterpreting the context. To this end, this paper proposes FaithfulRAG, a novel framework that resolves knowledge conflicts by explicitly modeling discrepancies between the model`s parametric knowledge and retrieved context. Specifically, FaithfulRAG identifies conflicting knowledge at the fact level and designs a self-thinking process, allowing LLMs to reason about and integrate conflicting facts before generating responses. Extensive experiments demonstrate that our method outperforms state-of-the-art methods. The code is available at https://github.com/DeepLearnXMU/Faithful-RAG
format Preprint
id arxiv_https___arxiv_org_abs_2506_08938
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FaithfulRAG: Fact-Level Conflict Modeling for Context-Faithful Retrieval-Augmented Generation
Zhang, Qinggang
Xiang, Zhishang
Xiao, Yilin
Wang, Le
Li, Junhui
Wang, Xinrun
Su, Jinsong
Computation and Language
Large language models (LLMs) augmented with retrieval systems have demonstrated significant potential in handling knowledge-intensive tasks. However, these models often struggle with unfaithfulness issues, generating outputs that either ignore the retrieved context or inconsistently blend it with the LLM`s parametric knowledge. This issue is particularly severe in cases of knowledge conflict, where the retrieved context conflicts with the model`s parametric knowledge. While existing faithful RAG approaches enforce strict context adherence through well-designed prompts or modified decoding strategies, our analysis reveals a critical limitation: they achieve faithfulness by forcibly suppressing the model`s parametric knowledge, which undermines the model`s internal knowledge structure and increases the risk of misinterpreting the context. To this end, this paper proposes FaithfulRAG, a novel framework that resolves knowledge conflicts by explicitly modeling discrepancies between the model`s parametric knowledge and retrieved context. Specifically, FaithfulRAG identifies conflicting knowledge at the fact level and designs a self-thinking process, allowing LLMs to reason about and integrate conflicting facts before generating responses. Extensive experiments demonstrate that our method outperforms state-of-the-art methods. The code is available at https://github.com/DeepLearnXMU/Faithful-RAG
title FaithfulRAG: Fact-Level Conflict Modeling for Context-Faithful Retrieval-Augmented Generation
topic Computation and Language
url https://arxiv.org/abs/2506.08938