@inproceedings{tian-etal-2026-mirage,
title = "{MIRAGE}: Reliability Assessment via Grounded Evidence for Table Reasoning",
author = "Tian, Jinlong and
Jin, Ruochun and
Liu, Shixuan and
Zhou, Yanfang and
Li, Mengmeng and
Wang, Haotian and
Xu, Xinhai and
Tang, Yuhua",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.1427/",
pages = "32610--32623",
ISBN = "979-8-89176-459-0",
abstract = "Uncertainty estimation for language models often treats reliability as a property of the answer distribution: high confidence, stable samples, or low semantic entropy are taken as signs that an answer can be trusted. Table reasoning exposes a fundamentally different failure mode. A model may consistently return the same answer while linking the wrong entity, selecting the wrong column, applying the wrong operation, or relying on incomplete or insufficient evidence. We introduce MIRAGE{---}Model Inference Reliability Assessment via Grounded Evidence, a structure-grounded reliability framework for selective table reasoning. MIRAGE estimates residual structural uncertainty from test-time traces, combining outcome consistency with grounding and computation signals, and produces calibrated answer-or-abstain decisions with lightweight reliability certificates. On a controlled four-family diagnostic benchmark, structural reliability signals substantially reduce selective risk compared with generic confidence, sampling, semantic-entropy, and hallucination-detection baselines."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="tian-etal-2026-mirage">
<titleInfo>
<title>MIRAGE: Reliability Assessment via Grounded Evidence for Table Reasoning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jinlong</namePart>
<namePart type="family">Tian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruochun</namePart>
<namePart type="family">Jin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shixuan</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yanfang</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mengmeng</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haotian</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinhai</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuhua</namePart>
<namePart type="family">Tang</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>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Uncertainty estimation for language models often treats reliability as a property of the answer distribution: high confidence, stable samples, or low semantic entropy are taken as signs that an answer can be trusted. Table reasoning exposes a fundamentally different failure mode. A model may consistently return the same answer while linking the wrong entity, selecting the wrong column, applying the wrong operation, or relying on incomplete or insufficient evidence. We introduce MIRAGE—Model Inference Reliability Assessment via Grounded Evidence, a structure-grounded reliability framework for selective table reasoning. MIRAGE estimates residual structural uncertainty from test-time traces, combining outcome consistency with grounding and computation signals, and produces calibrated answer-or-abstain decisions with lightweight reliability certificates. On a controlled four-family diagnostic benchmark, structural reliability signals substantially reduce selective risk compared with generic confidence, sampling, semantic-entropy, and hallucination-detection baselines.</abstract>
<identifier type="citekey">tian-etal-2026-mirage</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1427/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>32610</start>
<end>32623</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T MIRAGE: Reliability Assessment via Grounded Evidence for Table Reasoning
%A Tian, Jinlong
%A Jin, Ruochun
%A Liu, Shixuan
%A Zhou, Yanfang
%A Li, Mengmeng
%A Wang, Haotian
%A Xu, Xinhai
%A Tang, Yuhua
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F tian-etal-2026-mirage
%X Uncertainty estimation for language models often treats reliability as a property of the answer distribution: high confidence, stable samples, or low semantic entropy are taken as signs that an answer can be trusted. Table reasoning exposes a fundamentally different failure mode. A model may consistently return the same answer while linking the wrong entity, selecting the wrong column, applying the wrong operation, or relying on incomplete or insufficient evidence. We introduce MIRAGE—Model Inference Reliability Assessment via Grounded Evidence, a structure-grounded reliability framework for selective table reasoning. MIRAGE estimates residual structural uncertainty from test-time traces, combining outcome consistency with grounding and computation signals, and produces calibrated answer-or-abstain decisions with lightweight reliability certificates. On a controlled four-family diagnostic benchmark, structural reliability signals substantially reduce selective risk compared with generic confidence, sampling, semantic-entropy, and hallucination-detection baselines.
%U https://aclanthology.org/2026.emnlp-main.1427/
%P 32610-32623
Markdown (Informal)
[MIRAGE: Reliability Assessment via Grounded Evidence for Table Reasoning](https://aclanthology.org/2026.emnlp-main.1427/) (Tian et al., EMNLP 2026)
ACL
- Jinlong Tian, Ruochun Jin, Shixuan Liu, Yanfang Zhou, Mengmeng Li, Haotian Wang, Xinhai Xu, and Yuhua Tang. 2026. MIRAGE: Reliability Assessment via Grounded Evidence for Table Reasoning. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 32610–32623, Budapest, Hungary. Association for Computational Linguistics.