@inproceedings{duan-etal-2026-counterfactual,
title = "When Counterfactual Debiasing Backfires: Identity Cues as Task-Dependent Evidence in Harmful Language Understanding",
author = "Duan, Jia and
Lu, Xiaoyu and
Hu, Wenjing and
Ding, Zihao",
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.1062/",
pages = "24392--24410",
ISBN = "979-8-89176-459-0",
abstract = "Counterfactual debiasing is often used to reduce reliance on identity-linked shortcuts in harmful language understanding. We show that the same intervention can be harmful when identity mentions are not shortcuts but label-defining evidence. In multi-task cyberbullying modeling, identity cues may be spurious for cyberbullying detection (CD) and sentiment analysis (SA), yet necessary for fine-grained target identification (TI). We propose task-aware routing, which sends CD and SA through a debiased representation while preserving the original representation for TI. We evaluate this approach on the English-translated BullyExplain benchmark, HateXplain, and an external Measuring Hate Speech diagnostic. Experiments show that task-aware routing improves TI, while auxiliary detection and sentiment performance remain largely unchanged where applicable. Ablation studies attribute this improvement to representation routing that preserves identity-bearing evidence for TI rather than to class-weighting choices, with reduced-architecture diagnostics suggesting that the effect is not specific to the graph-enhanced model. These results support a task-conditional view of counterfactual debiasing: the same identity cue can warrant suppression in one task and preservation in another."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="duan-etal-2026-counterfactual">
<titleInfo>
<title>When Counterfactual Debiasing Backfires: Identity Cues as Task-Dependent Evidence in Harmful Language Understanding</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jia</namePart>
<namePart type="family">Duan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaoyu</namePart>
<namePart type="family">Lu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wenjing</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zihao</namePart>
<namePart type="family">Ding</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>Counterfactual debiasing is often used to reduce reliance on identity-linked shortcuts in harmful language understanding. We show that the same intervention can be harmful when identity mentions are not shortcuts but label-defining evidence. In multi-task cyberbullying modeling, identity cues may be spurious for cyberbullying detection (CD) and sentiment analysis (SA), yet necessary for fine-grained target identification (TI). We propose task-aware routing, which sends CD and SA through a debiased representation while preserving the original representation for TI. We evaluate this approach on the English-translated BullyExplain benchmark, HateXplain, and an external Measuring Hate Speech diagnostic. Experiments show that task-aware routing improves TI, while auxiliary detection and sentiment performance remain largely unchanged where applicable. Ablation studies attribute this improvement to representation routing that preserves identity-bearing evidence for TI rather than to class-weighting choices, with reduced-architecture diagnostics suggesting that the effect is not specific to the graph-enhanced model. These results support a task-conditional view of counterfactual debiasing: the same identity cue can warrant suppression in one task and preservation in another.</abstract>
<identifier type="citekey">duan-etal-2026-counterfactual</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1062/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>24392</start>
<end>24410</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T When Counterfactual Debiasing Backfires: Identity Cues as Task-Dependent Evidence in Harmful Language Understanding
%A Duan, Jia
%A Lu, Xiaoyu
%A Hu, Wenjing
%A Ding, Zihao
%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 duan-etal-2026-counterfactual
%X Counterfactual debiasing is often used to reduce reliance on identity-linked shortcuts in harmful language understanding. We show that the same intervention can be harmful when identity mentions are not shortcuts but label-defining evidence. In multi-task cyberbullying modeling, identity cues may be spurious for cyberbullying detection (CD) and sentiment analysis (SA), yet necessary for fine-grained target identification (TI). We propose task-aware routing, which sends CD and SA through a debiased representation while preserving the original representation for TI. We evaluate this approach on the English-translated BullyExplain benchmark, HateXplain, and an external Measuring Hate Speech diagnostic. Experiments show that task-aware routing improves TI, while auxiliary detection and sentiment performance remain largely unchanged where applicable. Ablation studies attribute this improvement to representation routing that preserves identity-bearing evidence for TI rather than to class-weighting choices, with reduced-architecture diagnostics suggesting that the effect is not specific to the graph-enhanced model. These results support a task-conditional view of counterfactual debiasing: the same identity cue can warrant suppression in one task and preservation in another.
%U https://aclanthology.org/2026.emnlp-main.1062/
%P 24392-24410
Markdown (Informal)
[When Counterfactual Debiasing Backfires: Identity Cues as Task-Dependent Evidence in Harmful Language Understanding](https://aclanthology.org/2026.emnlp-main.1062/) (Duan et al., EMNLP 2026)
ACL