@inproceedings{sudhi-etal-2026-attentioncd,
title = "{A}ttention{CD}: Reducing Factuality Hallucinations in Large Language Models using Task-specific Head-based Contrastive Decoding",
author = "Sudhi, Viju and
Stodden, Regina and
Cimiano, Philipp",
editor = {Aziz, Wilker and
Berant, Jonathan and
Eikema, Bryan and
Ilia, Evgenia and
de Marneffe, Marie-Catherine and
Plank, Barbara and
Shelmanov, Artem and
Swayamdipta, Swabha and
Tiedemann, J{\"o}rg and
Vazhentsev, Artem and
V{\'a}zquez, Ra{\'u}l and
Zerva, Chrysoula},
booktitle = "The Proceedings of the Third Workshop on Uncertainty-Aware {NLP}",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.uncertainlp-main.3/",
pages = "30--47",
ISBN = "979-8-89176-486-6",
abstract = "Despite the tremendous volume of data they are trained on, Large Language Models (LLMs) can generate plausibly sounding but non-factual responses. These factually hallucinated responses reduce users' trust in such models and hinder their adoption in safety-critical applications. Among the many training-free hallucination mitigation strategies proposed, contrastive decoding strategies have received wide attention. Most of the work in this direction explores how contrasting probability distributions from different decoder layers can help in reducing factuality hallucinations. This work, in contrast, investigates whether masking of attention heads can provide reliable contrastive signals. In particular, we propose AttentionCD, a decoding strategy that masks task-specific attention heads to yield better contrastive signals to reduce factuality hallucinations in LLMs. Our approach shows competitive performance with the existing layer-based contrastive decoding strategies and offers an average performance improvement of over +11.1{\%} over greedy decoding across different tasks and models."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sudhi-etal-2026-attentioncd">
<titleInfo>
<title>AttentionCD: Reducing Factuality Hallucinations in Large Language Models using Task-specific Head-based Contrastive Decoding</title>
</titleInfo>
<name type="personal">
<namePart type="given">Viju</namePart>
<namePart type="family">Sudhi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Regina</namePart>
<namePart type="family">Stodden</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Philipp</namePart>
<namePart type="family">Cimiano</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>The Proceedings of the Third Workshop on Uncertainty-Aware NLP</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wilker</namePart>
<namePart type="family">Aziz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jonathan</namePart>
<namePart type="family">Berant</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bryan</namePart>
<namePart type="family">Eikema</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Evgenia</namePart>
<namePart type="family">Ilia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marie-Catherine</namePart>
<namePart type="family">de Marneffe</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Barbara</namePart>
<namePart type="family">Plank</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Artem</namePart>
<namePart type="family">Shelmanov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Swabha</namePart>
<namePart type="family">Swayamdipta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jörg</namePart>
<namePart type="family">Tiedemann</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Artem</namePart>
<namePart type="family">Vazhentsev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Raúl</namePart>
<namePart type="family">Vázquez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chrysoula</namePart>
<namePart type="family">Zerva</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">979-8-89176-486-6</identifier>
</relatedItem>
<abstract>Despite the tremendous volume of data they are trained on, Large Language Models (LLMs) can generate plausibly sounding but non-factual responses. These factually hallucinated responses reduce users’ trust in such models and hinder their adoption in safety-critical applications. Among the many training-free hallucination mitigation strategies proposed, contrastive decoding strategies have received wide attention. Most of the work in this direction explores how contrasting probability distributions from different decoder layers can help in reducing factuality hallucinations. This work, in contrast, investigates whether masking of attention heads can provide reliable contrastive signals. In particular, we propose AttentionCD, a decoding strategy that masks task-specific attention heads to yield better contrastive signals to reduce factuality hallucinations in LLMs. Our approach shows competitive performance with the existing layer-based contrastive decoding strategies and offers an average performance improvement of over +11.1% over greedy decoding across different tasks and models.</abstract>
<identifier type="citekey">sudhi-etal-2026-attentioncd</identifier>
<location>
<url>https://aclanthology.org/2026.uncertainlp-main.3/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>30</start>
<end>47</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T AttentionCD: Reducing Factuality Hallucinations in Large Language Models using Task-specific Head-based Contrastive Decoding
%A Sudhi, Viju
%A Stodden, Regina
%A Cimiano, Philipp
%Y Aziz, Wilker
%Y Berant, Jonathan
%Y Eikema, Bryan
%Y Ilia, Evgenia
%Y de Marneffe, Marie-Catherine
%Y Plank, Barbara
%Y Shelmanov, Artem
%Y Swayamdipta, Swabha
%Y Tiedemann, Jörg
%Y Vazhentsev, Artem
%Y Vázquez, Raúl
%Y Zerva, Chrysoula
%S The Proceedings of the Third Workshop on Uncertainty-Aware NLP
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-486-6
%F sudhi-etal-2026-attentioncd
%X Despite the tremendous volume of data they are trained on, Large Language Models (LLMs) can generate plausibly sounding but non-factual responses. These factually hallucinated responses reduce users’ trust in such models and hinder their adoption in safety-critical applications. Among the many training-free hallucination mitigation strategies proposed, contrastive decoding strategies have received wide attention. Most of the work in this direction explores how contrasting probability distributions from different decoder layers can help in reducing factuality hallucinations. This work, in contrast, investigates whether masking of attention heads can provide reliable contrastive signals. In particular, we propose AttentionCD, a decoding strategy that masks task-specific attention heads to yield better contrastive signals to reduce factuality hallucinations in LLMs. Our approach shows competitive performance with the existing layer-based contrastive decoding strategies and offers an average performance improvement of over +11.1% over greedy decoding across different tasks and models.
%U https://aclanthology.org/2026.uncertainlp-main.3/
%P 30-47
Markdown (Informal)
[AttentionCD: Reducing Factuality Hallucinations in Large Language Models using Task-specific Head-based Contrastive Decoding](https://aclanthology.org/2026.uncertainlp-main.3/) (Sudhi et al., UncertaiNLP 2026)
ACL