Salvato in:
Dettagli Bibliografici
Autori principali: Xie, Yiqing, Zhou, Wenxuan, Prakash, Pradyot, Jin, Di, Mao, Yuning, Fettes, Quintin, Talebzadeh, Arya, Wang, Sinong, Fang, Han, Rose, Carolyn, Fried, Daniel, Zhang, Hejia
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2410.18359
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908388530585600
author Xie, Yiqing
Zhou, Wenxuan
Prakash, Pradyot
Jin, Di
Mao, Yuning
Fettes, Quintin
Talebzadeh, Arya
Wang, Sinong
Fang, Han
Rose, Carolyn
Fried, Daniel
Zhang, Hejia
author_facet Xie, Yiqing
Zhou, Wenxuan
Prakash, Pradyot
Jin, Di
Mao, Yuning
Fettes, Quintin
Talebzadeh, Arya
Wang, Sinong
Fang, Han
Rose, Carolyn
Fried, Daniel
Zhang, Hejia
contents Factuality evaluation aims to detect factual errors produced by language models (LMs) and hence guide the development of more factual models. Towards this goal, we train a factuality evaluator, FenCE, that provides LM generators with claim-level factuality feedback. We conduct data augmentation on a combination of public judgment datasets to train FenCE to (1) generate textual critiques along with scores and (2) make claim-level judgment based on diverse source documents obtained by various tools. We then present a framework that leverages FenCE to improve the factuality of LM generators by constructing training data. Specifically, we generate a set of candidate responses, leverage FenCE to revise and score each response without introducing lesser-known facts, and train the generator by preferring highly scored revised responses. Experiments show that our data augmentation methods improve the evaluator's accuracy by 2.9% on LLM-AggreFact. With FenCE, we improve Llama2-7B-chat and Llama3-8B-chat's factuality rate by 16.86% and 14.45% on FActScore, outperforming state-of-the-art factuality finetuning methods by 8.83% and 6.96%.
format Preprint
id arxiv_https___arxiv_org_abs_2410_18359
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Improving Model Factuality with Fine-grained Critique-based Evaluator
Xie, Yiqing
Zhou, Wenxuan
Prakash, Pradyot
Jin, Di
Mao, Yuning
Fettes, Quintin
Talebzadeh, Arya
Wang, Sinong
Fang, Han
Rose, Carolyn
Fried, Daniel
Zhang, Hejia
Computation and Language
Factuality evaluation aims to detect factual errors produced by language models (LMs) and hence guide the development of more factual models. Towards this goal, we train a factuality evaluator, FenCE, that provides LM generators with claim-level factuality feedback. We conduct data augmentation on a combination of public judgment datasets to train FenCE to (1) generate textual critiques along with scores and (2) make claim-level judgment based on diverse source documents obtained by various tools. We then present a framework that leverages FenCE to improve the factuality of LM generators by constructing training data. Specifically, we generate a set of candidate responses, leverage FenCE to revise and score each response without introducing lesser-known facts, and train the generator by preferring highly scored revised responses. Experiments show that our data augmentation methods improve the evaluator's accuracy by 2.9% on LLM-AggreFact. With FenCE, we improve Llama2-7B-chat and Llama3-8B-chat's factuality rate by 16.86% and 14.45% on FActScore, outperforming state-of-the-art factuality finetuning methods by 8.83% and 6.96%.
title Improving Model Factuality with Fine-grained Critique-based Evaluator
topic Computation and Language
url https://arxiv.org/abs/2410.18359