Attribute First, then Generate: Locally-attributable Grounded Text Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Slobodkin, Aviv, Hirsch, Eran, Cattan, Arie, Schuster, Tal, Dagan, Ido
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914857710780416
author Slobodkin, Aviv
Hirsch, Eran
Cattan, Arie
Schuster, Tal
Dagan, Ido
author_facet Slobodkin, Aviv
Hirsch, Eran
Cattan, Arie
Schuster, Tal
Dagan, Ido
contents Recent efforts to address hallucinations in Large Language Models (LLMs) have focused on attributed text generation, which supplements generated texts with citations of supporting sources for post-generation fact-checking and corrections. Yet, these citations often point to entire documents or paragraphs, burdening users with extensive verification work. In this paper, we introduce a locally-attributable text generation approach, prioritizing concise attributions. Our method, named "Attribute First, then Generate", breaks down the conventional end-to-end generation process into three intuitive steps: content selection, sentence planning, and sequential sentence generation. By initially identifying relevant source segments ("select first") and then conditioning the generation process on them ("then generate"), we ensure these segments also act as the output's fine-grained attributions ("select" becomes "attribute"). Tested on Multi-document Summarization and Long-form Question-answering, our method not only yields more concise citations than the baselines but also maintains - and in some cases enhances - both generation quality and attribution accuracy. Furthermore, it significantly reduces the time required for fact verification by human assessors.
format Preprint
id arxiv_https___arxiv_org_abs_2403_17104
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Attribute First, then Generate: Locally-attributable Grounded Text Generation
Slobodkin, Aviv
Hirsch, Eran
Cattan, Arie
Schuster, Tal
Dagan, Ido
Computation and Language
Recent efforts to address hallucinations in Large Language Models (LLMs) have focused on attributed text generation, which supplements generated texts with citations of supporting sources for post-generation fact-checking and corrections. Yet, these citations often point to entire documents or paragraphs, burdening users with extensive verification work. In this paper, we introduce a locally-attributable text generation approach, prioritizing concise attributions. Our method, named "Attribute First, then Generate", breaks down the conventional end-to-end generation process into three intuitive steps: content selection, sentence planning, and sequential sentence generation. By initially identifying relevant source segments ("select first") and then conditioning the generation process on them ("then generate"), we ensure these segments also act as the output's fine-grained attributions ("select" becomes "attribute"). Tested on Multi-document Summarization and Long-form Question-answering, our method not only yields more concise citations than the baselines but also maintains - and in some cases enhances - both generation quality and attribution accuracy. Furthermore, it significantly reduces the time required for fact verification by human assessors.
title Attribute First, then Generate: Locally-attributable Grounded Text Generation
topic Computation and Language
url https://arxiv.org/abs/2403.17104