@inproceedings{li-etal-2026-beyond-output,
title = "Beyond Output Confidence: Epistemic-Aware Hallucination Detection with Answer-Level Signals",
author = "Li, Jieran and
Hu, Xiuyuan and
Zhao, Yang and
Sun, Dongbiao and
Zhang, Hao",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-acl.674/",
pages = "13796--13806",
ISBN = "979-8-89176-395-1",
abstract = "Despite their strong generative capabilities, large language models frequently exhibit hallucinations, particularly due to outside-boundary confidence where incorrect assertions are produced with high statistical certainty. Existing approaches commonly use output probability as a proxy for truthfulness; however, this signal is confounded by epistemic uncertainty and cannot reliably distinguish genuine uncertainty from fabricated content. We argue that effective hallucination detection requires integrating surface-level confidence with signals that reflect the model{'}s underlying epistemic state. To this end, we propose Answer-level Intrinsic Cognition (AIC), a model-agnostic metric that captures epistemic boundary deviations by measuring answer-level stability across multiple stochastic forward passes. By coupling AIC with conventional output uncertainty, we derive a composite metric that disentangles within-boundary uncertainty from outside-boundary confidence. Across three public question-answering benchmarks and diverse model scales, the two-dimensional score consistently outperforms strong uncertainty-only baselines, with larger gains on adversarially constructed hallucination sets. The code is available at: https://github.com/HXYfighter/AIC-ACL2026."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2026-beyond-output">
<titleInfo>
<title>Beyond Output Confidence: Epistemic-Aware Hallucination Detection with Answer-Level Signals</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jieran</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiuyuan</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yang</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dongbiao</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hao</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-395-1</identifier>
</relatedItem>
<abstract>Despite their strong generative capabilities, large language models frequently exhibit hallucinations, particularly due to outside-boundary confidence where incorrect assertions are produced with high statistical certainty. Existing approaches commonly use output probability as a proxy for truthfulness; however, this signal is confounded by epistemic uncertainty and cannot reliably distinguish genuine uncertainty from fabricated content. We argue that effective hallucination detection requires integrating surface-level confidence with signals that reflect the model’s underlying epistemic state. To this end, we propose Answer-level Intrinsic Cognition (AIC), a model-agnostic metric that captures epistemic boundary deviations by measuring answer-level stability across multiple stochastic forward passes. By coupling AIC with conventional output uncertainty, we derive a composite metric that disentangles within-boundary uncertainty from outside-boundary confidence. Across three public question-answering benchmarks and diverse model scales, the two-dimensional score consistently outperforms strong uncertainty-only baselines, with larger gains on adversarially constructed hallucination sets. The code is available at: https://github.com/HXYfighter/AIC-ACL2026.</abstract>
<identifier type="citekey">li-etal-2026-beyond-output</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.674/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>13796</start>
<end>13806</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Beyond Output Confidence: Epistemic-Aware Hallucination Detection with Answer-Level Signals
%A Li, Jieran
%A Hu, Xiuyuan
%A Zhao, Yang
%A Sun, Dongbiao
%A Zhang, Hao
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F li-etal-2026-beyond-output
%X Despite their strong generative capabilities, large language models frequently exhibit hallucinations, particularly due to outside-boundary confidence where incorrect assertions are produced with high statistical certainty. Existing approaches commonly use output probability as a proxy for truthfulness; however, this signal is confounded by epistemic uncertainty and cannot reliably distinguish genuine uncertainty from fabricated content. We argue that effective hallucination detection requires integrating surface-level confidence with signals that reflect the model’s underlying epistemic state. To this end, we propose Answer-level Intrinsic Cognition (AIC), a model-agnostic metric that captures epistemic boundary deviations by measuring answer-level stability across multiple stochastic forward passes. By coupling AIC with conventional output uncertainty, we derive a composite metric that disentangles within-boundary uncertainty from outside-boundary confidence. Across three public question-answering benchmarks and diverse model scales, the two-dimensional score consistently outperforms strong uncertainty-only baselines, with larger gains on adversarially constructed hallucination sets. The code is available at: https://github.com/HXYfighter/AIC-ACL2026.
%U https://aclanthology.org/2026.findings-acl.674/
%P 13796-13806
Markdown (Informal)
[Beyond Output Confidence: Epistemic-Aware Hallucination Detection with Answer-Level Signals](https://aclanthology.org/2026.findings-acl.674/) (Li et al., Findings 2026)
ACL