@inproceedings{sohail-pasha-qureshi-etal-2026-halluvision,
title = "{H}allu{V}ision at {SHROOM}-Vision: Hallucination Span Detection with a Classifier over a Vision-Language Backbone",
author = "Sohail Pasha Qureshi, Mohammed and
Charan, Surya and
Kuila, Alapan",
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.64/",
pages = "869--876",
ISBN = "979-8-89176-486-6",
abstract = "We present our system for the SHROOM-Visions 2026 shared task on fine-grained hallucination detection in vision-language model (VLM) outputs. Our approach uses three token-level internal signals from Qwen2.5-VL-3B-Instruct: hidden representations, the change in token confidence when the image is removed, and predictive uncertainty. A lightweight classifier combines these signals to predict hallucinated spans and their confidence, while LoRA enables parameter-efficient adaptation of the VLM backbone. Our submission achieves IoU scores of 0.36{--}0.42 and Spearman correlations of 0.41{--}0.46 on English, French, and Italian, ranking 10th in both French and Italian. We additionally analyze continued-training experiments that improved our local validation scores but consistently degraded official test performance. Our analysis attributes this discrepancy primarily to threshold tuning on a small validation split, which was only partially mitigated by subsequent adjustments. Increasing the backbone size from 3B to 7B parameters did not resolve the issue. These findings highlight the sensitivity of hallucination detection systems to validation size and decision-threshold selection."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sohail-pasha-qureshi-etal-2026-halluvision">
<titleInfo>
<title>HalluVision at SHROOM-Vision: Hallucination Span Detection with a Classifier over a Vision-Language Backbone</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mohammed</namePart>
<namePart type="family">Sohail Pasha Qureshi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Surya</namePart>
<namePart type="family">Charan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alapan</namePart>
<namePart type="family">Kuila</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>We present our system for the SHROOM-Visions 2026 shared task on fine-grained hallucination detection in vision-language model (VLM) outputs. Our approach uses three token-level internal signals from Qwen2.5-VL-3B-Instruct: hidden representations, the change in token confidence when the image is removed, and predictive uncertainty. A lightweight classifier combines these signals to predict hallucinated spans and their confidence, while LoRA enables parameter-efficient adaptation of the VLM backbone. Our submission achieves IoU scores of 0.36–0.42 and Spearman correlations of 0.41–0.46 on English, French, and Italian, ranking 10th in both French and Italian. We additionally analyze continued-training experiments that improved our local validation scores but consistently degraded official test performance. Our analysis attributes this discrepancy primarily to threshold tuning on a small validation split, which was only partially mitigated by subsequent adjustments. Increasing the backbone size from 3B to 7B parameters did not resolve the issue. These findings highlight the sensitivity of hallucination detection systems to validation size and decision-threshold selection.</abstract>
<identifier type="citekey">sohail-pasha-qureshi-etal-2026-halluvision</identifier>
<location>
<url>https://aclanthology.org/2026.uncertainlp-main.64/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>869</start>
<end>876</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T HalluVision at SHROOM-Vision: Hallucination Span Detection with a Classifier over a Vision-Language Backbone
%A Sohail Pasha Qureshi, Mohammed
%A Charan, Surya
%A Kuila, Alapan
%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 sohail-pasha-qureshi-etal-2026-halluvision
%X We present our system for the SHROOM-Visions 2026 shared task on fine-grained hallucination detection in vision-language model (VLM) outputs. Our approach uses three token-level internal signals from Qwen2.5-VL-3B-Instruct: hidden representations, the change in token confidence when the image is removed, and predictive uncertainty. A lightweight classifier combines these signals to predict hallucinated spans and their confidence, while LoRA enables parameter-efficient adaptation of the VLM backbone. Our submission achieves IoU scores of 0.36–0.42 and Spearman correlations of 0.41–0.46 on English, French, and Italian, ranking 10th in both French and Italian. We additionally analyze continued-training experiments that improved our local validation scores but consistently degraded official test performance. Our analysis attributes this discrepancy primarily to threshold tuning on a small validation split, which was only partially mitigated by subsequent adjustments. Increasing the backbone size from 3B to 7B parameters did not resolve the issue. These findings highlight the sensitivity of hallucination detection systems to validation size and decision-threshold selection.
%U https://aclanthology.org/2026.uncertainlp-main.64/
%P 869-876
Markdown (Informal)
[HalluVision at SHROOM-Vision: Hallucination Span Detection with a Classifier over a Vision-Language Backbone](https://aclanthology.org/2026.uncertainlp-main.64/) (Sohail Pasha Qureshi et al., UncertaiNLP 2026)
ACL