@inproceedings{shiba-etal-2026-leap,
title = "{LEAP}: Learning Edit-Adaptable Parameters for Multi-Hop Knowledge Propagation",
author = {Shiba, Taichi and
M{\"u}ller-Eberstein, Max and
Isonuma, Masaru and
Oseki, Yohei and
Miyao, Yusuke},
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2150/",
pages = "48872--48887",
ISBN = "979-8-89176-459-0",
abstract = "Knowledge editing aims to update factual knowledge in language models without full retraining. However, edits should also propagate to downstream queries that require combining the edited fact with other knowledge, such as multi-hop queries. Existing approaches address this propagation problem with more elaborate edit-time procedures, through augmented reasoning examples or external editing modules. In this work, we take a different approach: we meta-learn model parameters from which simple edits are more likely to propagate. We propose LEAP (Learning Edit-Adaptable Parameters), a MAML-based framework that learns edit-adaptable parameters such that simple gradient-based fine-tuning on new single-hop facts improves answers to their multi-hop consequences. Experiments on MQuAKE show that LEAP consistently improves over standard LoRA editing and achieves a favorable accuracy-efficiency trade-off compared with representative knowledge-editing baselines. Further analysis suggests that LEAP makes bridge entities more accessible and strengthens the bridge-to-answer transition needed for multi-hop reasoning. These results suggest a new direction for improving edit propagation: preparing the model itself to be more amenable to propagation-friendly updates, rather than relying only on increasingly complex edit-time procedures."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="shiba-etal-2026-leap">
<titleInfo>
<title>LEAP: Learning Edit-Adaptable Parameters for Multi-Hop Knowledge Propagation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Taichi</namePart>
<namePart type="family">Shiba</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Max</namePart>
<namePart type="family">Müller-Eberstein</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Masaru</namePart>
<namePart type="family">Isonuma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yohei</namePart>
<namePart type="family">Oseki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yusuke</namePart>
<namePart type="family">Miyao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Knowledge editing aims to update factual knowledge in language models without full retraining. However, edits should also propagate to downstream queries that require combining the edited fact with other knowledge, such as multi-hop queries. Existing approaches address this propagation problem with more elaborate edit-time procedures, through augmented reasoning examples or external editing modules. In this work, we take a different approach: we meta-learn model parameters from which simple edits are more likely to propagate. We propose LEAP (Learning Edit-Adaptable Parameters), a MAML-based framework that learns edit-adaptable parameters such that simple gradient-based fine-tuning on new single-hop facts improves answers to their multi-hop consequences. Experiments on MQuAKE show that LEAP consistently improves over standard LoRA editing and achieves a favorable accuracy-efficiency trade-off compared with representative knowledge-editing baselines. Further analysis suggests that LEAP makes bridge entities more accessible and strengthens the bridge-to-answer transition needed for multi-hop reasoning. These results suggest a new direction for improving edit propagation: preparing the model itself to be more amenable to propagation-friendly updates, rather than relying only on increasingly complex edit-time procedures.</abstract>
<identifier type="citekey">shiba-etal-2026-leap</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2150/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>48872</start>
<end>48887</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T LEAP: Learning Edit-Adaptable Parameters for Multi-Hop Knowledge Propagation
%A Shiba, Taichi
%A Müller-Eberstein, Max
%A Isonuma, Masaru
%A Oseki, Yohei
%A Miyao, Yusuke
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F shiba-etal-2026-leap
%X Knowledge editing aims to update factual knowledge in language models without full retraining. However, edits should also propagate to downstream queries that require combining the edited fact with other knowledge, such as multi-hop queries. Existing approaches address this propagation problem with more elaborate edit-time procedures, through augmented reasoning examples or external editing modules. In this work, we take a different approach: we meta-learn model parameters from which simple edits are more likely to propagate. We propose LEAP (Learning Edit-Adaptable Parameters), a MAML-based framework that learns edit-adaptable parameters such that simple gradient-based fine-tuning on new single-hop facts improves answers to their multi-hop consequences. Experiments on MQuAKE show that LEAP consistently improves over standard LoRA editing and achieves a favorable accuracy-efficiency trade-off compared with representative knowledge-editing baselines. Further analysis suggests that LEAP makes bridge entities more accessible and strengthens the bridge-to-answer transition needed for multi-hop reasoning. These results suggest a new direction for improving edit propagation: preparing the model itself to be more amenable to propagation-friendly updates, rather than relying only on increasingly complex edit-time procedures.
%U https://aclanthology.org/2026.emnlp-main.2150/
%P 48872-48887
Markdown (Informal)
[LEAP: Learning Edit-Adaptable Parameters for Multi-Hop Knowledge Propagation](https://aclanthology.org/2026.emnlp-main.2150/) (Shiba et al., EMNLP 2026)
ACL