@inproceedings{wang-etal-2024-sss,
title = "{SSS}: Editing Factual Knowledge in Language Models towards Semantic Sparse Space",
author = "Wang, Huazheng and
Sun, Haifeng and
Wang, Jingyu and
Qi, Qi and
Xia, Zixuan and
Zhang, Menghao and
Liao, Jianxin",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2024",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-acl.331/",
doi = "10.18653/v1/2024.findings-acl.331",
pages = "5559--5570",
abstract = "Language Models (LMs) acquire factual knowledge during pre-training and store it in the parameters, which can be valuable for downstream tasks. As world evolves, some facts may be incorrectly induced or become obsolete over time. Various model editing methods have been proposed to modify specific examples in LMs. However, existing training-based methods still suffer from sub-optimal locality, where irrelevant neighborhood examples can be adversely influenced. Model{'}s gradients are still struggling to identify the appropriate direction when updating the parameters. To address this issue, we find that directing the hidden state of the edit example towards spaces where semantics are sparse tends to help preserve the semantics of irrelevant neighborhood examples. Based on this hypothesis, we propose a novel metric, named SSS, to evaluate the degree of sparsity around a sentence embedding in the semantic space without any human or machine annotation. Subsequently, we incorporate SSS into the original loss function of the existing training-based methods to enhance locality. Experiments conducted on two datasets across various models demonstrate that SSS is effective in improving both locality and reasoning capability."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2024-sss">
<titleInfo>
<title>SSS: Editing Factual Knowledge in Language Models towards Semantic Sparse Space</title>
</titleInfo>
<name type="personal">
<namePart type="given">Huazheng</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haifeng</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jingyu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qi</namePart>
<namePart type="family">Qi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zixuan</namePart>
<namePart type="family">Xia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Menghao</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jianxin</namePart>
<namePart type="family">Liao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Lun-Wei</namePart>
<namePart type="family">Ku</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andre</namePart>
<namePart type="family">Martins</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivek</namePart>
<namePart type="family">Srikumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Bangkok, Thailand</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Language Models (LMs) acquire factual knowledge during pre-training and store it in the parameters, which can be valuable for downstream tasks. As world evolves, some facts may be incorrectly induced or become obsolete over time. Various model editing methods have been proposed to modify specific examples in LMs. However, existing training-based methods still suffer from sub-optimal locality, where irrelevant neighborhood examples can be adversely influenced. Model’s gradients are still struggling to identify the appropriate direction when updating the parameters. To address this issue, we find that directing the hidden state of the edit example towards spaces where semantics are sparse tends to help preserve the semantics of irrelevant neighborhood examples. Based on this hypothesis, we propose a novel metric, named SSS, to evaluate the degree of sparsity around a sentence embedding in the semantic space without any human or machine annotation. Subsequently, we incorporate SSS into the original loss function of the existing training-based methods to enhance locality. Experiments conducted on two datasets across various models demonstrate that SSS is effective in improving both locality and reasoning capability.</abstract>
<identifier type="citekey">wang-etal-2024-sss</identifier>
<identifier type="doi">10.18653/v1/2024.findings-acl.331</identifier>
<location>
<url>https://aclanthology.org/2024.findings-acl.331/</url>
</location>
<part>
<date>2024-08</date>
<extent unit="page">
<start>5559</start>
<end>5570</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T SSS: Editing Factual Knowledge in Language Models towards Semantic Sparse Space
%A Wang, Huazheng
%A Sun, Haifeng
%A Wang, Jingyu
%A Qi, Qi
%A Xia, Zixuan
%A Zhang, Menghao
%A Liao, Jianxin
%Y Ku, Lun-Wei
%Y Martins, Andre
%Y Srikumar, Vivek
%S Findings of the Association for Computational Linguistics: ACL 2024
%D 2024
%8 August
%I Association for Computational Linguistics
%C Bangkok, Thailand
%F wang-etal-2024-sss
%X Language Models (LMs) acquire factual knowledge during pre-training and store it in the parameters, which can be valuable for downstream tasks. As world evolves, some facts may be incorrectly induced or become obsolete over time. Various model editing methods have been proposed to modify specific examples in LMs. However, existing training-based methods still suffer from sub-optimal locality, where irrelevant neighborhood examples can be adversely influenced. Model’s gradients are still struggling to identify the appropriate direction when updating the parameters. To address this issue, we find that directing the hidden state of the edit example towards spaces where semantics are sparse tends to help preserve the semantics of irrelevant neighborhood examples. Based on this hypothesis, we propose a novel metric, named SSS, to evaluate the degree of sparsity around a sentence embedding in the semantic space without any human or machine annotation. Subsequently, we incorporate SSS into the original loss function of the existing training-based methods to enhance locality. Experiments conducted on two datasets across various models demonstrate that SSS is effective in improving both locality and reasoning capability.
%R 10.18653/v1/2024.findings-acl.331
%U https://aclanthology.org/2024.findings-acl.331/
%U https://doi.org/10.18653/v1/2024.findings-acl.331
%P 5559-5570
Markdown (Informal)
[SSS: Editing Factual Knowledge in Language Models towards Semantic Sparse Space](https://aclanthology.org/2024.findings-acl.331/) (Wang et al., Findings 2024)
ACL