@inproceedings{xu-etal-2026-domains,
title = "From Domains to Instances: Dual-Granularity Data Synthesis for {LLM} Unlearning",
author = "Xu, Xiaoyu and
Du, Minxin and
LI, Zitong and
Liang, Zi and
Guo, Zhibiao and
Shiyu, Zhang and
Hu, Peizhao and
Ye, Qingqing and
Hu, Haibo",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-acl.1424/",
pages = "28540--28556",
ISBN = "979-8-89176-395-1",
abstract = "Although machine unlearning is essential for removing private, harmful, or copyrighted content from LLMs, current benchmarks often fail to faithfully represent the true ``forgetting scope'' learned by the model. We formalize two distinct unlearning granularities, domain-level and instance-level, and propose , an automated framework for synthesizing high-quality forget sets.Unlike prior work relying on \textit{external} generators, exploits the target model per se to elicit data that matches its internal knowledge distribution through seed-guided and adversarial prompting. Our experiments across diverse benchmarks show that it achieves a superior balance of relevance, diversity, and efficiency. Quantitatively, in the Harry Potter domain, it improves relevance by ${\sim}20$ and diversity by ${\sim}$0.05 while \textit{halving} the total data size compared to SOTAs. Ultimately, it facilitates more robust forgetting and better utility preservation, providing a more rigorous foundation for evaluating LLM unlearning."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="xu-etal-2026-domains">
<titleInfo>
<title>From Domains to Instances: Dual-Granularity Data Synthesis for LLM Unlearning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xiaoyu</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Minxin</namePart>
<namePart type="family">Du</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zitong</namePart>
<namePart type="family">LI</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zi</namePart>
<namePart type="family">Liang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhibiao</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhang</namePart>
<namePart type="family">Shiyu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Peizhao</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Qingqing</namePart>
<namePart type="family">Ye</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haibo</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-395-1</identifier>
</relatedItem>
<abstract>Although machine unlearning is essential for removing private, harmful, or copyrighted content from LLMs, current benchmarks often fail to faithfully represent the true “forgetting scope” learned by the model. We formalize two distinct unlearning granularities, domain-level and instance-level, and propose , an automated framework for synthesizing high-quality forget sets.Unlike prior work relying on external generators, exploits the target model per se to elicit data that matches its internal knowledge distribution through seed-guided and adversarial prompting. Our experiments across diverse benchmarks show that it achieves a superior balance of relevance, diversity, and efficiency. Quantitatively, in the Harry Potter domain, it improves relevance by \sim20 and diversity by \sim0.05 while halving the total data size compared to SOTAs. Ultimately, it facilitates more robust forgetting and better utility preservation, providing a more rigorous foundation for evaluating LLM unlearning.</abstract>
<identifier type="citekey">xu-etal-2026-domains</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.1424/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>28540</start>
<end>28556</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T From Domains to Instances: Dual-Granularity Data Synthesis for LLM Unlearning
%A Xu, Xiaoyu
%A Du, Minxin
%A LI, Zitong
%A Liang, Zi
%A Guo, Zhibiao
%A Shiyu, Zhang
%A Hu, Peizhao
%A Ye, Qingqing
%A Hu, Haibo
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F xu-etal-2026-domains
%X Although machine unlearning is essential for removing private, harmful, or copyrighted content from LLMs, current benchmarks often fail to faithfully represent the true “forgetting scope” learned by the model. We formalize two distinct unlearning granularities, domain-level and instance-level, and propose , an automated framework for synthesizing high-quality forget sets.Unlike prior work relying on external generators, exploits the target model per se to elicit data that matches its internal knowledge distribution through seed-guided and adversarial prompting. Our experiments across diverse benchmarks show that it achieves a superior balance of relevance, diversity, and efficiency. Quantitatively, in the Harry Potter domain, it improves relevance by \sim20 and diversity by \sim0.05 while halving the total data size compared to SOTAs. Ultimately, it facilitates more robust forgetting and better utility preservation, providing a more rigorous foundation for evaluating LLM unlearning.
%U https://aclanthology.org/2026.findings-acl.1424/
%P 28540-28556
Markdown (Informal)
[From Domains to Instances: Dual-Granularity Data Synthesis for LLM Unlearning](https://aclanthology.org/2026.findings-acl.1424/) (Xu et al., Findings 2026)
ACL
- Xiaoyu Xu, Minxin Du, Zitong LI, Zi Liang, Zhibiao Guo, Zhang Shiyu, Peizhao Hu, Qingqing Ye, and Haibo Hu. 2026. From Domains to Instances: Dual-Granularity Data Synthesis for LLM Unlearning. In Findings of the Association for Computational Linguistics: ACL 2026, pages 28540–28556, San Diego, California, United States. Association for Computational Linguistics.