<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="to-le-2025-harry">
    <titleInfo>
        <title>Harry Potter is Still Here! Probing Knowledge Leakage in Targeted Unlearned Large Language Models</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Bang</namePart>
        <namePart type="given">Trinh</namePart>
        <namePart type="given">Tran</namePart>
        <namePart type="family">To</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Thai</namePart>
        <namePart type="family">Le</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: EMNLP 2025</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Christos</namePart>
            <namePart type="family">Christodoulopoulos</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Tanmoy</namePart>
            <namePart type="family">Chakraborty</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Carolyn</namePart>
            <namePart type="family">Rose</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Violet</namePart>
            <namePart type="family">Peng</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Suzhou, China</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-335-7</identifier>
    </relatedItem>
    <abstract>This work presents LURK (Latent Unlearned Knowledge), a novel framework that probes for undesired knowledge retention in unlearned LLMs through adversarial suffix prompting. LURK automatically generates adversarial prompt suffixes designed to elicit residual knowledge about the Harry Potter domain, a commonly used benchmark for unlearning. Our experiments reveal that even models deemed successfully unlearned can leak idiosyncratic information under targeted adversarial conditions, highlighting critical limitations of current unlearning evaluation standards. By uncovering implicit knowledge through indirect probing, LURK offers a more rigorous and diagnostic tool for assessing the robustness of unlearning algorithms. Code and data will be available at https://github.com/Rachel1809/LURK.</abstract>
    <identifier type="citekey">to-le-2025-harry</identifier>
    <location>
        <url>https://aclanthology.org/2025.findings-emnlp.778/</url>
    </location>
    <part>
        <date>2025-11</date>
        <extent unit="page">
            <start>14427</start>
            <end>14439</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Harry Potter is Still Here! Probing Knowledge Leakage in Targeted Unlearned Large Language Models
%A To, Bang Trinh Tran
%A Le, Thai
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Findings of the Association for Computational Linguistics: EMNLP 2025
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-335-7
%F to-le-2025-harry
%X This work presents LURK (Latent Unlearned Knowledge), a novel framework that probes for undesired knowledge retention in unlearned LLMs through adversarial suffix prompting. LURK automatically generates adversarial prompt suffixes designed to elicit residual knowledge about the Harry Potter domain, a commonly used benchmark for unlearning. Our experiments reveal that even models deemed successfully unlearned can leak idiosyncratic information under targeted adversarial conditions, highlighting critical limitations of current unlearning evaluation standards. By uncovering implicit knowledge through indirect probing, LURK offers a more rigorous and diagnostic tool for assessing the robustness of unlearning algorithms. Code and data will be available at https://github.com/Rachel1809/LURK.
%U https://aclanthology.org/2025.findings-emnlp.778/
%P 14427-14439

Markdown (Informal)

[Harry Potter is Still Here! Probing Knowledge Leakage in Targeted Unlearned Large Language Models](https://aclanthology.org/2025.findings-emnlp.778/) (To & Le, Findings 2025)

• Harry Potter is Still Here! Probing Knowledge Leakage in Targeted Unlearned Large Language Models (To & Le, Findings 2025)

ACL

• Bang Trinh Tran To and Thai Le. 2025. Harry Potter is Still Here! Probing Knowledge Leakage in Targeted Unlearned Large Language Models. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 14427–14439, Suzhou, China. Association for Computational Linguistics.