@inproceedings{harish-2026-sparse,
title = "Sparse-Positive Hallucination Evaluation Entangles Detection and Prevalence",
author = "Harish, Srinivas",
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.46/",
pages = "729--733",
ISBN = "979-8-89176-486-6",
abstract = "Span-level hallucination metrics aggregate outputs whose references contain hallucinated spans with outputs whose references are empty. Under scoring rules that award a perfect match to an empty prediction against empty gold, the aggregate entangles detection performance with empty-gold prevalence. We write the exact decomposition as $s=ea+(1-e)d$, where $e$ is empty-gold prevalence, $a$ is performance on empty-gold items, and $d$ is performance on non-empty-gold items. Under the benchmark{'}s metric conventions, an empty submission identifies $e$ from leaderboard values without access to closed labels. Applied to a multilingual vision-language hallucination benchmark, this recovers rates from 0.1549 to 0.2554. Assuming perfect empty-gold performance gives a maximum contribution equal to 47.3{--}64.8{\%} of the mean team score and a conservative lower bound on $d$. Under this bound, the highest-ranked language moves from best for 17 of 18 teams to best for none and falls from rank 1 to rank 3. Released annotations show that union-over-annotators gold retains 4.09 times as many characters as majority-agreement gold, while 71.7{\%} of annotated spans were marked by one of three annotators. Cross-split comparisons therefore require reporting both empty-gold prevalence and system performance on empty-gold items alongside aggregate scores."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="harish-2026-sparse">
<titleInfo>
<title>Sparse-Positive Hallucination Evaluation Entangles Detection and Prevalence</title>
</titleInfo>
<name type="personal">
<namePart type="given">Srinivas</namePart>
<namePart type="family">Harish</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>Span-level hallucination metrics aggregate outputs whose references contain hallucinated spans with outputs whose references are empty. Under scoring rules that award a perfect match to an empty prediction against empty gold, the aggregate entangles detection performance with empty-gold prevalence. We write the exact decomposition as s=ea+(1-e)d, where e is empty-gold prevalence, a is performance on empty-gold items, and d is performance on non-empty-gold items. Under the benchmark’s metric conventions, an empty submission identifies e from leaderboard values without access to closed labels. Applied to a multilingual vision-language hallucination benchmark, this recovers rates from 0.1549 to 0.2554. Assuming perfect empty-gold performance gives a maximum contribution equal to 47.3–64.8% of the mean team score and a conservative lower bound on d. Under this bound, the highest-ranked language moves from best for 17 of 18 teams to best for none and falls from rank 1 to rank 3. Released annotations show that union-over-annotators gold retains 4.09 times as many characters as majority-agreement gold, while 71.7% of annotated spans were marked by one of three annotators. Cross-split comparisons therefore require reporting both empty-gold prevalence and system performance on empty-gold items alongside aggregate scores.</abstract>
<identifier type="citekey">harish-2026-sparse</identifier>
<location>
<url>https://aclanthology.org/2026.uncertainlp-main.46/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>729</start>
<end>733</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Sparse-Positive Hallucination Evaluation Entangles Detection and Prevalence
%A Harish, Srinivas
%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 harish-2026-sparse
%X Span-level hallucination metrics aggregate outputs whose references contain hallucinated spans with outputs whose references are empty. Under scoring rules that award a perfect match to an empty prediction against empty gold, the aggregate entangles detection performance with empty-gold prevalence. We write the exact decomposition as s=ea+(1-e)d, where e is empty-gold prevalence, a is performance on empty-gold items, and d is performance on non-empty-gold items. Under the benchmark’s metric conventions, an empty submission identifies e from leaderboard values without access to closed labels. Applied to a multilingual vision-language hallucination benchmark, this recovers rates from 0.1549 to 0.2554. Assuming perfect empty-gold performance gives a maximum contribution equal to 47.3–64.8% of the mean team score and a conservative lower bound on d. Under this bound, the highest-ranked language moves from best for 17 of 18 teams to best for none and falls from rank 1 to rank 3. Released annotations show that union-over-annotators gold retains 4.09 times as many characters as majority-agreement gold, while 71.7% of annotated spans were marked by one of three annotators. Cross-split comparisons therefore require reporting both empty-gold prevalence and system performance on empty-gold items alongside aggregate scores.
%U https://aclanthology.org/2026.uncertainlp-main.46/
%P 729-733
Markdown (Informal)
[Sparse-Positive Hallucination Evaluation Entangles Detection and Prevalence](https://aclanthology.org/2026.uncertainlp-main.46/) (Harish, UncertaiNLP 2026)
ACL