Understand What LLM Needs: Dual Preference Alignment for Retrieval-Augmented Generation

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Dong, Guanting, Zhu, Yutao, Zhang, Chenghao, Wang, Zechen, Dou, Zhicheng, Wen, Ji-Rong
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866929425734434816
author Dong, Guanting
Zhu, Yutao
Zhang, Chenghao
Wang, Zechen
Dou, Zhicheng
Wen, Ji-Rong
author_facet Dong, Guanting
Zhu, Yutao
Zhang, Chenghao
Wang, Zechen
Dou, Zhicheng
Wen, Ji-Rong
contents Retrieval-augmented generation (RAG) has demonstrated effectiveness in mitigating the hallucination problem of large language models (LLMs). However, the difficulty of aligning the retriever with the diverse LLMs' knowledge preferences inevitably poses an inevitable challenge in developing a reliable RAG system. To address this issue, we propose DPA-RAG, a universal framework designed to align diverse knowledge preferences within RAG systems. Specifically, we initially introduce a preference knowledge construction pipline and incorporate five novel query augmentation strategies to alleviate preference data scarcity. Based on preference data, DPA-RAG accomplishes both external and internal preference alignment: 1) It jointly integrate pair-wise, point-wise, and contrastive preference alignment abilities into the reranker, achieving external preference alignment among RAG components. 2) It further introduces a pre-aligned stage before vanilla Supervised Fine-tuning (SFT), enabling LLMs to implicitly capture knowledge aligned with their reasoning preferences, achieving LLMs' internal alignment. Experimental results across four knowledge-intensive QA datasets demonstrate that DPA-RAG outperforms all baselines and seamlessly integrates both black-box and open-sourced LLM readers. Further qualitative analysis and discussions also provide empirical guidance for achieving reliable RAG systems. Our code is publicly available at https://github.com/dongguanting/DPA-RAG.
format Preprint
id arxiv_https___arxiv_org_abs_2406_18676
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Understand What LLM Needs: Dual Preference Alignment for Retrieval-Augmented Generation
Dong, Guanting
Zhu, Yutao
Zhang, Chenghao
Wang, Zechen
Dou, Zhicheng
Wen, Ji-Rong
Computation and Language
Artificial Intelligence
Machine Learning
Retrieval-augmented generation (RAG) has demonstrated effectiveness in mitigating the hallucination problem of large language models (LLMs). However, the difficulty of aligning the retriever with the diverse LLMs' knowledge preferences inevitably poses an inevitable challenge in developing a reliable RAG system. To address this issue, we propose DPA-RAG, a universal framework designed to align diverse knowledge preferences within RAG systems. Specifically, we initially introduce a preference knowledge construction pipline and incorporate five novel query augmentation strategies to alleviate preference data scarcity. Based on preference data, DPA-RAG accomplishes both external and internal preference alignment: 1) It jointly integrate pair-wise, point-wise, and contrastive preference alignment abilities into the reranker, achieving external preference alignment among RAG components. 2) It further introduces a pre-aligned stage before vanilla Supervised Fine-tuning (SFT), enabling LLMs to implicitly capture knowledge aligned with their reasoning preferences, achieving LLMs' internal alignment. Experimental results across four knowledge-intensive QA datasets demonstrate that DPA-RAG outperforms all baselines and seamlessly integrates both black-box and open-sourced LLM readers. Further qualitative analysis and discussions also provide empirical guidance for achieving reliable RAG systems. Our code is publicly available at https://github.com/dongguanting/DPA-RAG.
title Understand What LLM Needs: Dual Preference Alignment for Retrieval-Augmented Generation
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2406.18676