Saved in:
Bibliographic Details
Main Authors: Zhang, Siyuan, Zhang, Yichi, Dong, Yinpeng, Su, Hang
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.19127
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909836027887616
author Zhang, Siyuan
Zhang, Yichi
Dong, Yinpeng
Su, Hang
author_facet Zhang, Siyuan
Zhang, Yichi
Dong, Yinpeng
Su, Hang
contents Large Language Models (LLMs) often struggle to align their responses with objective facts, resulting in the issue of factual hallucinations, which can be difficult to detect and mislead users without relevant knowledge. Although post-training techniques have been employed to mitigate the issue, existing methods usually suffer from poor generalization and trade-offs in other different capabilities. In this paper, we propose to address these by directly augmenting LLM's fundamental ability to precisely leverage its knowledge and introduce PKUE (Precise Knowledge Utilization Enhancement), which fine-tunes the model on self-generated responses to precise and simple factual questions through preference optimization. Furthermore, we construct FactualBench, a comprehensive and precise factual QA dataset containing 181k Chinese data spanning 21 domains, to facilitate both evaluation and training. Extensive experiments demonstrate that PKUE significantly improves LLM overall performance, with consistent enhancement across factual tasks of various forms, general tasks beyond factuality, and tasks in different language.
format Preprint
id arxiv_https___arxiv_org_abs_2502_19127
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Exploring the Generalizability of Factual Hallucination Mitigation via Enhancing Precise Knowledge Utilization
Zhang, Siyuan
Zhang, Yichi
Dong, Yinpeng
Su, Hang
Computation and Language
Large Language Models (LLMs) often struggle to align their responses with objective facts, resulting in the issue of factual hallucinations, which can be difficult to detect and mislead users without relevant knowledge. Although post-training techniques have been employed to mitigate the issue, existing methods usually suffer from poor generalization and trade-offs in other different capabilities. In this paper, we propose to address these by directly augmenting LLM's fundamental ability to precisely leverage its knowledge and introduce PKUE (Precise Knowledge Utilization Enhancement), which fine-tunes the model on self-generated responses to precise and simple factual questions through preference optimization. Furthermore, we construct FactualBench, a comprehensive and precise factual QA dataset containing 181k Chinese data spanning 21 domains, to facilitate both evaluation and training. Extensive experiments demonstrate that PKUE significantly improves LLM overall performance, with consistent enhancement across factual tasks of various forms, general tasks beyond factuality, and tasks in different language.
title Exploring the Generalizability of Factual Hallucination Mitigation via Enhancing Precise Knowledge Utilization
topic Computation and Language
url https://arxiv.org/abs/2502.19127