RbFT: Robust Fine-tuning for Retrieval-Augmented Generation against Retrieval Defects

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tu, Yiteng, Su, Weihang, Zhou, Yujia, Liu, Yiqun, Ai, Qingyao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909470162944000
author Tu, Yiteng
Su, Weihang
Zhou, Yujia
Liu, Yiqun
Ai, Qingyao
author_facet Tu, Yiteng
Su, Weihang
Zhou, Yujia
Liu, Yiqun
Ai, Qingyao
contents Retrieval-augmented generation (RAG) enhances large language models (LLMs) by integrating external knowledge retrieved from a knowledge base. However, its effectiveness is fundamentally constrained by the reliability of both the retriever and the knowledge base. In real-world scenarios, imperfections in these components often lead to the retrieval of noisy, irrelevant, or misleading counterfactual information, ultimately undermining the trustworthiness of RAG systems. To address this challenge, we propose Robust Fine-Tuning (RbFT), a method designed to enhance the resilience of LLMs against retrieval defects through two targeted fine-tuning tasks. Experimental results demonstrate that RbFT significantly improves the robustness of RAG systems across diverse retrieval conditions, surpassing existing methods while maintaining high inference efficiency and compatibility with other robustness techniques.
format Preprint
id arxiv_https___arxiv_org_abs_2501_18365
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle RbFT: Robust Fine-tuning for Retrieval-Augmented Generation against Retrieval Defects
Tu, Yiteng
Su, Weihang
Zhou, Yujia
Liu, Yiqun
Ai, Qingyao
Computation and Language
Information Retrieval
Retrieval-augmented generation (RAG) enhances large language models (LLMs) by integrating external knowledge retrieved from a knowledge base. However, its effectiveness is fundamentally constrained by the reliability of both the retriever and the knowledge base. In real-world scenarios, imperfections in these components often lead to the retrieval of noisy, irrelevant, or misleading counterfactual information, ultimately undermining the trustworthiness of RAG systems. To address this challenge, we propose Robust Fine-Tuning (RbFT), a method designed to enhance the resilience of LLMs against retrieval defects through two targeted fine-tuning tasks. Experimental results demonstrate that RbFT significantly improves the robustness of RAG systems across diverse retrieval conditions, surpassing existing methods while maintaining high inference efficiency and compatibility with other robustness techniques.
title RbFT: Robust Fine-tuning for Retrieval-Augmented Generation against Retrieval Defects
topic Computation and Language
Information Retrieval
url https://arxiv.org/abs/2501.18365