@inproceedings{waseem-fawaz-2026-false,
title = "A False Positive in Cross-Lingual Unlearning Evaluation",
author = "Waseem, Mohamed and
Fawaz, Mohammed",
editor = "Adelani, David Ifeoluwa and
Alabi, Jesujoba and
Arnett, Catherine and
Chang, Tyler and
Downey, C.M. and
Goldman, Omer and
Li, Senyu and
Miranda, Lester James Validad and
Qi, Jirui and
Stap, David and
Winata, Genta Indra and
Zhang, Xinyu Crystina",
booktitle = "Proceedings of the 6th Workshop on Multilingual Representation Learning ({MRL} 2026)",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.mrl-1.43/",
pages = "533--541",
ISBN = "979-8-89176-469-9",
abstract = "Cross-lingual audits of LLM unlearning ask whether a fact forgotten in one language remains recoverable in another, and read a small loss increase in the second language as residual recall. We show that this inference is unsafe without a control that is routinely omitted. Auditing GradDiff unlearning of TOFU fictional-persona facts on Qwen2.5-0.5B across English, Modern Standard Arabic, Egyptian Arabic and Hindi, we find the expected signature of leakage: the unlearned model{'}s forget-set loss rises by +2.922 nats in English but only +0.774 in MSA, 26{\%} as much. Measuring the base model as well shows that signature to be an artefact. English-only fine-tuning left no measurable trace of the facts in the other three varieties {---} it raised forget-set loss there, and raised retain-set loss by an indistinguishable amount {---} so there is no evidence of cross-lingual memory for unlearning to have left behind. Once the never-learned and the uniform-degradation components are removed, the MSA effect is +0.054 nats, 2.5{\%} of English rather than 26{\%}. We argue that a base-model transfer control and a retain-set control are jointly necessary for any cross-lingual unlearning claim, and report the audit design that yields them at no additional training cost."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="waseem-fawaz-2026-false">
<titleInfo>
<title>A False Positive in Cross-Lingual Unlearning Evaluation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mohamed</namePart>
<namePart type="family">Waseem</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammed</namePart>
<namePart type="family">Fawaz</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 6th Workshop on Multilingual Representation Learning (MRL 2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="given">Ifeoluwa</namePart>
<namePart type="family">Adelani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jesujoba</namePart>
<namePart type="family">Alabi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Catherine</namePart>
<namePart type="family">Arnett</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tyler</namePart>
<namePart type="family">Chang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">C</namePart>
<namePart type="given">M</namePart>
<namePart type="family">Downey</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Omer</namePart>
<namePart type="family">Goldman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Senyu</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lester</namePart>
<namePart type="given">James</namePart>
<namePart type="given">Validad</namePart>
<namePart type="family">Miranda</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jirui</namePart>
<namePart type="family">Qi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Stap</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Genta</namePart>
<namePart type="given">Indra</namePart>
<namePart type="family">Winata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinyu</namePart>
<namePart type="given">Crystina</namePart>
<namePart type="family">Zhang</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">979-8-89176-469-9</identifier>
</relatedItem>
<abstract>Cross-lingual audits of LLM unlearning ask whether a fact forgotten in one language remains recoverable in another, and read a small loss increase in the second language as residual recall. We show that this inference is unsafe without a control that is routinely omitted. Auditing GradDiff unlearning of TOFU fictional-persona facts on Qwen2.5-0.5B across English, Modern Standard Arabic, Egyptian Arabic and Hindi, we find the expected signature of leakage: the unlearned model’s forget-set loss rises by +2.922 nats in English but only +0.774 in MSA, 26% as much. Measuring the base model as well shows that signature to be an artefact. English-only fine-tuning left no measurable trace of the facts in the other three varieties — it raised forget-set loss there, and raised retain-set loss by an indistinguishable amount — so there is no evidence of cross-lingual memory for unlearning to have left behind. Once the never-learned and the uniform-degradation components are removed, the MSA effect is +0.054 nats, 2.5% of English rather than 26%. We argue that a base-model transfer control and a retain-set control are jointly necessary for any cross-lingual unlearning claim, and report the audit design that yields them at no additional training cost.</abstract>
<identifier type="citekey">waseem-fawaz-2026-false</identifier>
<location>
<url>https://aclanthology.org/2026.mrl-1.43/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>533</start>
<end>541</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T A False Positive in Cross-Lingual Unlearning Evaluation
%A Waseem, Mohamed
%A Fawaz, Mohammed
%Y Adelani, David Ifeoluwa
%Y Alabi, Jesujoba
%Y Arnett, Catherine
%Y Chang, Tyler
%Y Downey, C. M.
%Y Goldman, Omer
%Y Li, Senyu
%Y Miranda, Lester James Validad
%Y Qi, Jirui
%Y Stap, David
%Y Winata, Genta Indra
%Y Zhang, Xinyu Crystina
%S Proceedings of the 6th Workshop on Multilingual Representation Learning (MRL 2026)
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-469-9
%F waseem-fawaz-2026-false
%X Cross-lingual audits of LLM unlearning ask whether a fact forgotten in one language remains recoverable in another, and read a small loss increase in the second language as residual recall. We show that this inference is unsafe without a control that is routinely omitted. Auditing GradDiff unlearning of TOFU fictional-persona facts on Qwen2.5-0.5B across English, Modern Standard Arabic, Egyptian Arabic and Hindi, we find the expected signature of leakage: the unlearned model’s forget-set loss rises by +2.922 nats in English but only +0.774 in MSA, 26% as much. Measuring the base model as well shows that signature to be an artefact. English-only fine-tuning left no measurable trace of the facts in the other three varieties — it raised forget-set loss there, and raised retain-set loss by an indistinguishable amount — so there is no evidence of cross-lingual memory for unlearning to have left behind. Once the never-learned and the uniform-degradation components are removed, the MSA effect is +0.054 nats, 2.5% of English rather than 26%. We argue that a base-model transfer control and a retain-set control are jointly necessary for any cross-lingual unlearning claim, and report the audit design that yields them at no additional training cost.
%U https://aclanthology.org/2026.mrl-1.43/
%P 533-541
Markdown (Informal)
[A False Positive in Cross-Lingual Unlearning Evaluation](https://aclanthology.org/2026.mrl-1.43/) (Waseem & Fawaz, MRL 2026)
ACL