@inproceedings{liu-etal-2025-abstain,
title = "Do not Abstain! Identify and Solve the Uncertainty",
author = "Liu, Jingyu and
JingquanPeng, JingquanPeng and
Wu, Xiaopeng and
Li, Xubin and
Ge, Tiezheng and
Zheng, Bo and
Liu, Yong",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.acl-long.840/",
doi = "10.18653/v1/2025.acl-long.840",
pages = "17177--17197",
ISBN = "979-8-89176-251-0",
abstract = "Despite the widespread application of Large Language Models (LLMs) across various domains, they frequently exhibit overconfidence when encountering uncertain scenarios, yet existing solutions primarily rely on evasive responses (e.g., ``I don{'}t know'') overlooks the opportunity of identifying and addressing the uncertainty to generate more satisfactory responses. To systematically investigate and improve LLMs' ability of recognizing and addressing the source of uncertainty, we introduce ConfuseBench, a benchmark mainly focus on three types of uncertainty: document scarcity, limited capability, and query ambiguity. Experiments with ConfuseBench reveal that current LLMs struggle to accurately identify the root cause of uncertainty and solve it. They prefer to attribute uncertainty to query ambiguity while overlooking capability limitations, especially for those weaker models. To tackle this challenge, we first generate context-aware inquiries that highlight the confusing aspect of the original query. Then we judge the source of uncertainty based on the uniqueness of the inquiry{'}s answer. Further we use an on-policy training method, InteractDPO to generate better inquiries. Experimental results demonstrate the efficacy of our approach."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="liu-etal-2025-abstain">
<titleInfo>
<title>Do not Abstain! Identify and Solve the Uncertainty</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jingyu</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">JingquanPeng</namePart>
<namePart type="family">JingquanPeng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaopeng</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xubin</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tiezheng</namePart>
<namePart type="family">Ge</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bo</namePart>
<namePart type="family">Zheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yong</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-251-0</identifier>
</relatedItem>
<abstract>Despite the widespread application of Large Language Models (LLMs) across various domains, they frequently exhibit overconfidence when encountering uncertain scenarios, yet existing solutions primarily rely on evasive responses (e.g., “I don’t know”) overlooks the opportunity of identifying and addressing the uncertainty to generate more satisfactory responses. To systematically investigate and improve LLMs’ ability of recognizing and addressing the source of uncertainty, we introduce ConfuseBench, a benchmark mainly focus on three types of uncertainty: document scarcity, limited capability, and query ambiguity. Experiments with ConfuseBench reveal that current LLMs struggle to accurately identify the root cause of uncertainty and solve it. They prefer to attribute uncertainty to query ambiguity while overlooking capability limitations, especially for those weaker models. To tackle this challenge, we first generate context-aware inquiries that highlight the confusing aspect of the original query. Then we judge the source of uncertainty based on the uniqueness of the inquiry’s answer. Further we use an on-policy training method, InteractDPO to generate better inquiries. Experimental results demonstrate the efficacy of our approach.</abstract>
<identifier type="citekey">liu-etal-2025-abstain</identifier>
<identifier type="doi">10.18653/v1/2025.acl-long.840</identifier>
<location>
<url>https://aclanthology.org/2025.acl-long.840/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>17177</start>
<end>17197</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Do not Abstain! Identify and Solve the Uncertainty
%A Liu, Jingyu
%A JingquanPeng, JingquanPeng
%A Wu, Xiaopeng
%A Li, Xubin
%A Ge, Tiezheng
%A Zheng, Bo
%A Liu, Yong
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-251-0
%F liu-etal-2025-abstain
%X Despite the widespread application of Large Language Models (LLMs) across various domains, they frequently exhibit overconfidence when encountering uncertain scenarios, yet existing solutions primarily rely on evasive responses (e.g., “I don’t know”) overlooks the opportunity of identifying and addressing the uncertainty to generate more satisfactory responses. To systematically investigate and improve LLMs’ ability of recognizing and addressing the source of uncertainty, we introduce ConfuseBench, a benchmark mainly focus on three types of uncertainty: document scarcity, limited capability, and query ambiguity. Experiments with ConfuseBench reveal that current LLMs struggle to accurately identify the root cause of uncertainty and solve it. They prefer to attribute uncertainty to query ambiguity while overlooking capability limitations, especially for those weaker models. To tackle this challenge, we first generate context-aware inquiries that highlight the confusing aspect of the original query. Then we judge the source of uncertainty based on the uniqueness of the inquiry’s answer. Further we use an on-policy training method, InteractDPO to generate better inquiries. Experimental results demonstrate the efficacy of our approach.
%R 10.18653/v1/2025.acl-long.840
%U https://aclanthology.org/2025.acl-long.840/
%U https://doi.org/10.18653/v1/2025.acl-long.840
%P 17177-17197
Markdown (Informal)
[Do not Abstain! Identify and Solve the Uncertainty](https://aclanthology.org/2025.acl-long.840/) (Liu et al., ACL 2025)
ACL
- Jingyu Liu, JingquanPeng JingquanPeng, Xiaopeng Wu, Xubin Li, Tiezheng Ge, Bo Zheng, and Yong Liu. 2025. Do not Abstain! Identify and Solve the Uncertainty. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 17177–17197, Vienna, Austria. Association for Computational Linguistics.