@inproceedings{pu-etal-2026-decoding,
title = "Decoding-Unlearning: Fact Forgetting via Entropy-Guided Inference",
author = "Pu, Jingwen and
Shi, Mingjun and
Ren, Xinrui and
Wang, Yizhe and
Zhang, Xinyu and
Wang, Zhaokun and
She, Kun",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-long.1850/",
pages = "39834--39860",
ISBN = "979-8-89176-390-6",
abstract = "Large Language Models (LLMs) exhibit powerful capabilities but inevitably memorize sensitive information, raising privacy, copyright, and safety concerns. Existing LLM unlearning methods typically rely on updating model parameters. While effective, they are often limited in real-world scenarios: fine-tuning large-scale models is costly, may introduce potential irreversible risks, and depends on both forget and retain datasets, which are often difficult to obtain in full. To address these challenges, an ideal solution is to achieve unlearning at inference time. To this end, we propose SEGUE, a training-free, plug-and-play inference-time unlearning strategy. SEGUE employs a probe to detect queries involving forgettable concepts and applies entropy-guided decoding to suppress target knowledge, enabling controllable non-factual generation while preserving overall model capabilities. Experiments on the MUSE, RWKU, and WMDP datasets, covering copyright, entity, and potential-risk knowledge, show that SEGUE effectively balances sensitive knowledge suppression and generation quality, outperforming existing most inference-time unlearning methods."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="pu-etal-2026-decoding">
<titleInfo>
<title>Decoding-Unlearning: Fact Forgetting via Entropy-Guided Inference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jingwen</namePart>
<namePart type="family">Pu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mingjun</namePart>
<namePart type="family">Shi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinrui</namePart>
<namePart type="family">Ren</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yizhe</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinyu</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhaokun</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kun</namePart>
<namePart type="family">She</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>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</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">979-8-89176-390-6</identifier>
</relatedItem>
<abstract>Large Language Models (LLMs) exhibit powerful capabilities but inevitably memorize sensitive information, raising privacy, copyright, and safety concerns. Existing LLM unlearning methods typically rely on updating model parameters. While effective, they are often limited in real-world scenarios: fine-tuning large-scale models is costly, may introduce potential irreversible risks, and depends on both forget and retain datasets, which are often difficult to obtain in full. To address these challenges, an ideal solution is to achieve unlearning at inference time. To this end, we propose SEGUE, a training-free, plug-and-play inference-time unlearning strategy. SEGUE employs a probe to detect queries involving forgettable concepts and applies entropy-guided decoding to suppress target knowledge, enabling controllable non-factual generation while preserving overall model capabilities. Experiments on the MUSE, RWKU, and WMDP datasets, covering copyright, entity, and potential-risk knowledge, show that SEGUE effectively balances sensitive knowledge suppression and generation quality, outperforming existing most inference-time unlearning methods.</abstract>
<identifier type="citekey">pu-etal-2026-decoding</identifier>
<location>
<url>https://aclanthology.org/2026.acl-long.1850/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>39834</start>
<end>39860</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Decoding-Unlearning: Fact Forgetting via Entropy-Guided Inference
%A Pu, Jingwen
%A Shi, Mingjun
%A Ren, Xinrui
%A Wang, Yizhe
%A Zhang, Xinyu
%A Wang, Zhaokun
%A She, Kun
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-390-6
%F pu-etal-2026-decoding
%X Large Language Models (LLMs) exhibit powerful capabilities but inevitably memorize sensitive information, raising privacy, copyright, and safety concerns. Existing LLM unlearning methods typically rely on updating model parameters. While effective, they are often limited in real-world scenarios: fine-tuning large-scale models is costly, may introduce potential irreversible risks, and depends on both forget and retain datasets, which are often difficult to obtain in full. To address these challenges, an ideal solution is to achieve unlearning at inference time. To this end, we propose SEGUE, a training-free, plug-and-play inference-time unlearning strategy. SEGUE employs a probe to detect queries involving forgettable concepts and applies entropy-guided decoding to suppress target knowledge, enabling controllable non-factual generation while preserving overall model capabilities. Experiments on the MUSE, RWKU, and WMDP datasets, covering copyright, entity, and potential-risk knowledge, show that SEGUE effectively balances sensitive knowledge suppression and generation quality, outperforming existing most inference-time unlearning methods.
%U https://aclanthology.org/2026.acl-long.1850/
%P 39834-39860
Markdown (Informal)
[Decoding-Unlearning: Fact Forgetting via Entropy-Guided Inference](https://aclanthology.org/2026.acl-long.1850/) (Pu et al., ACL 2026)
ACL
- Jingwen Pu, Mingjun Shi, Xinrui Ren, Yizhe Wang, Xinyu Zhang, Zhaokun Wang, and Kun She. 2026. Decoding-Unlearning: Fact Forgetting via Entropy-Guided Inference. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 39834–39860, San Diego, California, United States. Association for Computational Linguistics.