@inproceedings{niu-etal-2025-reledit,
title = "{R}el{E}dit: Evaluating Conceptual Knowledge Editing in Language Models via Relational Reasoning",
author = "Niu, Yifan and
Peng, Miao and
Chen, Nuo and
Bian, Yatao and
Xu, Tingyang and
Li, Jia",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-acl.533/",
doi = "10.18653/v1/2025.findings-acl.533",
pages = "10220--10238",
ISBN = "979-8-89176-256-5",
abstract = "The conceptual knowledge in Large Language Models (LLMs) can become outdated over time, and concept editing is often an option. Current evaluations on conceptual knowledge editing primarily focus on whether the definitions of concepts are successfully edited, neglecting the impact on the model{'}s related beliefs. To address this gap, we introduce a benchmark called RelEdit, which includes criteria and questions to assess both concept-level and instance-level relational reasoning abilities of edited models. Our findings reveal that existing knowledge editing methods struggle to reason about related conceptual knowledge effectively. Additionally, we introduce a simple memory-based in-context editing baseline, MICE, which prompts the language model to generate answers that align with the stored edited concepts in external memory. In addition, we find that MICE obtains the best scores on our benchmark, suggesting a promising research direction for model editing."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="niu-etal-2025-reledit">
<titleInfo>
<title>RelEdit: Evaluating Conceptual Knowledge Editing in Language Models via Relational Reasoning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yifan</namePart>
<namePart type="family">Niu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Miao</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nuo</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yatao</namePart>
<namePart type="family">Bian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tingyang</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jia</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-256-5</identifier>
</relatedItem>
<abstract>The conceptual knowledge in Large Language Models (LLMs) can become outdated over time, and concept editing is often an option. Current evaluations on conceptual knowledge editing primarily focus on whether the definitions of concepts are successfully edited, neglecting the impact on the model’s related beliefs. To address this gap, we introduce a benchmark called RelEdit, which includes criteria and questions to assess both concept-level and instance-level relational reasoning abilities of edited models. Our findings reveal that existing knowledge editing methods struggle to reason about related conceptual knowledge effectively. Additionally, we introduce a simple memory-based in-context editing baseline, MICE, which prompts the language model to generate answers that align with the stored edited concepts in external memory. In addition, we find that MICE obtains the best scores on our benchmark, suggesting a promising research direction for model editing.</abstract>
<identifier type="citekey">niu-etal-2025-reledit</identifier>
<identifier type="doi">10.18653/v1/2025.findings-acl.533</identifier>
<location>
<url>https://aclanthology.org/2025.findings-acl.533/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>10220</start>
<end>10238</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T RelEdit: Evaluating Conceptual Knowledge Editing in Language Models via Relational Reasoning
%A Niu, Yifan
%A Peng, Miao
%A Chen, Nuo
%A Bian, Yatao
%A Xu, Tingyang
%A Li, Jia
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Findings of the Association for Computational Linguistics: ACL 2025
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-256-5
%F niu-etal-2025-reledit
%X The conceptual knowledge in Large Language Models (LLMs) can become outdated over time, and concept editing is often an option. Current evaluations on conceptual knowledge editing primarily focus on whether the definitions of concepts are successfully edited, neglecting the impact on the model’s related beliefs. To address this gap, we introduce a benchmark called RelEdit, which includes criteria and questions to assess both concept-level and instance-level relational reasoning abilities of edited models. Our findings reveal that existing knowledge editing methods struggle to reason about related conceptual knowledge effectively. Additionally, we introduce a simple memory-based in-context editing baseline, MICE, which prompts the language model to generate answers that align with the stored edited concepts in external memory. In addition, we find that MICE obtains the best scores on our benchmark, suggesting a promising research direction for model editing.
%R 10.18653/v1/2025.findings-acl.533
%U https://aclanthology.org/2025.findings-acl.533/
%U https://doi.org/10.18653/v1/2025.findings-acl.533
%P 10220-10238
Markdown (Informal)
[RelEdit: Evaluating Conceptual Knowledge Editing in Language Models via Relational Reasoning](https://aclanthology.org/2025.findings-acl.533/) (Niu et al., Findings 2025)
ACL