@inproceedings{xiao-etal-2026-reverse,
title = "Reverse Probing: Supervised Token-level Uncertainty Quantification for Large Language Models in Clinical Text",
author = "Xiao, Bushi and
Soni, Sarvesh and
Wang, Daisy Zhe",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.407/",
pages = "8599--8616",
ISBN = "979-8-89176-462-0",
abstract = "As large language models are increasingly deployed for clinical text, ensuring they can reliably signal their own uncertainty becomes critical. Most existing uncertainty quantification (UQ) methods are designed for open-domain generation and cannot localize uncertainty at the token or span level in long clinical text. We propose Reverse Probing, the first uncertainty quantification framework specialized for clinical summarization, which estimates token-level uncertainty directly from pre-existing clinical summaries. Rather than sampling new outputs, Reverse Probing treats the text as a probe into the model{'}s internal state, extracting uncertainty features from four categories of signals. We evaluate Reverse Probing on two expert-annotated clinical datasets and outperform eight adapted baselines, achieving up to $4\times$ higher AUPRC while running over $15\times$ faster than sampling-based methods. Feature analysis reveals that the log-logit difference induced by the source record and the neighborhood context are the most consistent predictors across all models. This study offers interpretable insights into how models internally respond to unsupported clinical content."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="xiao-etal-2026-reverse">
<titleInfo>
<title>Reverse Probing: Supervised Token-level Uncertainty Quantification for Large Language Models in Clinical Text</title>
</titleInfo>
<name type="personal">
<namePart type="given">Bushi</namePart>
<namePart type="family">Xiao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sarvesh</namePart>
<namePart type="family">Soni</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Daisy</namePart>
<namePart type="given">Zhe</namePart>
<namePart type="family">Wang</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>As large language models are increasingly deployed for clinical text, ensuring they can reliably signal their own uncertainty becomes critical. Most existing uncertainty quantification (UQ) methods are designed for open-domain generation and cannot localize uncertainty at the token or span level in long clinical text. We propose Reverse Probing, the first uncertainty quantification framework specialized for clinical summarization, which estimates token-level uncertainty directly from pre-existing clinical summaries. Rather than sampling new outputs, Reverse Probing treats the text as a probe into the model’s internal state, extracting uncertainty features from four categories of signals. We evaluate Reverse Probing on two expert-annotated clinical datasets and outperform eight adapted baselines, achieving up to 4\times higher AUPRC while running over 15\times faster than sampling-based methods. Feature analysis reveals that the log-logit difference induced by the source record and the neighborhood context are the most consistent predictors across all models. This study offers interpretable insights into how models internally respond to unsupported clinical content.</abstract>
<identifier type="citekey">xiao-etal-2026-reverse</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.407/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>8599</start>
<end>8616</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Reverse Probing: Supervised Token-level Uncertainty Quantification for Large Language Models in Clinical Text
%A Xiao, Bushi
%A Soni, Sarvesh
%A Wang, Daisy Zhe
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F xiao-etal-2026-reverse
%X As large language models are increasingly deployed for clinical text, ensuring they can reliably signal their own uncertainty becomes critical. Most existing uncertainty quantification (UQ) methods are designed for open-domain generation and cannot localize uncertainty at the token or span level in long clinical text. We propose Reverse Probing, the first uncertainty quantification framework specialized for clinical summarization, which estimates token-level uncertainty directly from pre-existing clinical summaries. Rather than sampling new outputs, Reverse Probing treats the text as a probe into the model’s internal state, extracting uncertainty features from four categories of signals. We evaluate Reverse Probing on two expert-annotated clinical datasets and outperform eight adapted baselines, achieving up to 4\times higher AUPRC while running over 15\times faster than sampling-based methods. Feature analysis reveals that the log-logit difference induced by the source record and the neighborhood context are the most consistent predictors across all models. This study offers interpretable insights into how models internally respond to unsupported clinical content.
%U https://aclanthology.org/2026.findings-emnlp.407/
%P 8599-8616
Markdown (Informal)
[Reverse Probing: Supervised Token-level Uncertainty Quantification for Large Language Models in Clinical Text](https://aclanthology.org/2026.findings-emnlp.407/) (Xiao et al., Findings 2026)
ACL