Model Internals-based Answer Attribution for Trustworthy Retrieval-Augmented Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Qi, Jirui, Sarti, Gabriele, Fernández, Raquel, Bisazza, Arianna
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917850670694400
author Qi, Jirui
Sarti, Gabriele
Fernández, Raquel
Bisazza, Arianna
author_facet Qi, Jirui
Sarti, Gabriele
Fernández, Raquel
Bisazza, Arianna
contents Ensuring the verifiability of model answers is a fundamental challenge for retrieval-augmented generation (RAG) in the question answering (QA) domain. Recently, self-citation prompting was proposed to make large language models (LLMs) generate citations to supporting documents along with their answers. However, self-citing LLMs often struggle to match the required format, refer to non-existent sources, and fail to faithfully reflect LLMs' context usage throughout the generation. In this work, we present MIRAGE --Model Internals-based RAG Explanations -- a plug-and-play approach using model internals for faithful answer attribution in RAG applications. MIRAGE detects context-sensitive answer tokens and pairs them with retrieved documents contributing to their prediction via saliency methods. We evaluate our proposed approach on a multilingual extractive QA dataset, finding high agreement with human answer attribution. On open-ended QA, MIRAGE achieves citation quality and efficiency comparable to self-citation while also allowing for a finer-grained control of attribution parameters. Our qualitative evaluation highlights the faithfulness of MIRAGE's attributions and underscores the promising application of model internals for RAG answer attribution.
format Preprint
id arxiv_https___arxiv_org_abs_2406_13663
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Model Internals-based Answer Attribution for Trustworthy Retrieval-Augmented Generation
Qi, Jirui
Sarti, Gabriele
Fernández, Raquel
Bisazza, Arianna
Computation and Language
Artificial Intelligence
Machine Learning
Ensuring the verifiability of model answers is a fundamental challenge for retrieval-augmented generation (RAG) in the question answering (QA) domain. Recently, self-citation prompting was proposed to make large language models (LLMs) generate citations to supporting documents along with their answers. However, self-citing LLMs often struggle to match the required format, refer to non-existent sources, and fail to faithfully reflect LLMs' context usage throughout the generation. In this work, we present MIRAGE --Model Internals-based RAG Explanations -- a plug-and-play approach using model internals for faithful answer attribution in RAG applications. MIRAGE detects context-sensitive answer tokens and pairs them with retrieved documents contributing to their prediction via saliency methods. We evaluate our proposed approach on a multilingual extractive QA dataset, finding high agreement with human answer attribution. On open-ended QA, MIRAGE achieves citation quality and efficiency comparable to self-citation while also allowing for a finer-grained control of attribution parameters. Our qualitative evaluation highlights the faithfulness of MIRAGE's attributions and underscores the promising application of model internals for RAG answer attribution.
title Model Internals-based Answer Attribution for Trustworthy Retrieval-Augmented Generation
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2406.13663