@inproceedings{qian-etal-2026-userbench,
title = "{U}ser{B}ench: An Interactive Gym Environment for User-Centric Agents",
author = "Qian, Cheng and
Liu, Zuxin and
Prabhakar, Akshara and
Liu, Zhiwei and
Zhang, Jianguo and
Chen, Haolin and
Ji, Heng and
Yao, Weiran and
Heinecke, Shelby and
Savarese, Silvio and
Xiong, Caiming and
Wang, Huan",
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.772/",
pages = "17936--17966",
ISBN = "979-8-89176-459-0",
abstract = "Large Language Model agents increasingly solve tool-use tasks, yet often fail to determine what users actually want when goals are underspecified, incremental, or indirectly expressed. We introduce **UserBench**, a controlled Gymnasium environment for evaluating proactive preference elicitation in multi-turn, tool-grounded interactions. Instantiated in travel planning, UserBench hides user preferences behind implicit natural-language cues and requires agents to ask targeted clarification questions, search structured option databases, and select user-aligned options. Evaluating leading open- and closed-source models reveals a large gap between tool execution and user understanding: models achieve high valid-search rates but elicit only a minority of hidden preferences, and even stronger models remain far below a human oracle. Further analysis shows that many additional interaction turns are broad or non-informative rather than preference-relevant. UserBench provides a reproducible testbed for studying and training user-centric agent interaction while intentionally abstracting away full travel negotiation and infeasible-constraint recovery."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="qian-etal-2026-userbench">
<titleInfo>
<title>UserBench: An Interactive Gym Environment for User-Centric Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Cheng</namePart>
<namePart type="family">Qian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zuxin</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Akshara</namePart>
<namePart type="family">Prabhakar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhiwei</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jianguo</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haolin</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Heng</namePart>
<namePart type="family">Ji</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Weiran</namePart>
<namePart type="family">Yao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shelby</namePart>
<namePart type="family">Heinecke</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Silvio</namePart>
<namePart type="family">Savarese</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Caiming</namePart>
<namePart type="family">Xiong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Huan</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>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>Large Language Model agents increasingly solve tool-use tasks, yet often fail to determine what users actually want when goals are underspecified, incremental, or indirectly expressed. We introduce **UserBench**, a controlled Gymnasium environment for evaluating proactive preference elicitation in multi-turn, tool-grounded interactions. Instantiated in travel planning, UserBench hides user preferences behind implicit natural-language cues and requires agents to ask targeted clarification questions, search structured option databases, and select user-aligned options. Evaluating leading open- and closed-source models reveals a large gap between tool execution and user understanding: models achieve high valid-search rates but elicit only a minority of hidden preferences, and even stronger models remain far below a human oracle. Further analysis shows that many additional interaction turns are broad or non-informative rather than preference-relevant. UserBench provides a reproducible testbed for studying and training user-centric agent interaction while intentionally abstracting away full travel negotiation and infeasible-constraint recovery.</abstract>
<identifier type="citekey">qian-etal-2026-userbench</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.772/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>17936</start>
<end>17966</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T UserBench: An Interactive Gym Environment for User-Centric Agents
%A Qian, Cheng
%A Liu, Zuxin
%A Prabhakar, Akshara
%A Liu, Zhiwei
%A Zhang, Jianguo
%A Chen, Haolin
%A Ji, Heng
%A Yao, Weiran
%A Heinecke, Shelby
%A Savarese, Silvio
%A Xiong, Caiming
%A Wang, Huan
%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 qian-etal-2026-userbench
%X Large Language Model agents increasingly solve tool-use tasks, yet often fail to determine what users actually want when goals are underspecified, incremental, or indirectly expressed. We introduce **UserBench**, a controlled Gymnasium environment for evaluating proactive preference elicitation in multi-turn, tool-grounded interactions. Instantiated in travel planning, UserBench hides user preferences behind implicit natural-language cues and requires agents to ask targeted clarification questions, search structured option databases, and select user-aligned options. Evaluating leading open- and closed-source models reveals a large gap between tool execution and user understanding: models achieve high valid-search rates but elicit only a minority of hidden preferences, and even stronger models remain far below a human oracle. Further analysis shows that many additional interaction turns are broad or non-informative rather than preference-relevant. UserBench provides a reproducible testbed for studying and training user-centric agent interaction while intentionally abstracting away full travel negotiation and infeasible-constraint recovery.
%U https://aclanthology.org/2026.emnlp-main.772/
%P 17936-17966
Markdown (Informal)
[UserBench: An Interactive Gym Environment for User-Centric Agents](https://aclanthology.org/2026.emnlp-main.772/) (Qian et al., EMNLP 2026)
ACL
- Cheng Qian, Zuxin Liu, Akshara Prabhakar, Zhiwei Liu, Jianguo Zhang, Haolin Chen, Heng Ji, Weiran Yao, Shelby Heinecke, Silvio Savarese, Caiming Xiong, and Huan Wang. 2026. UserBench: An Interactive Gym Environment for User-Centric Agents. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 17936–17966, Budapest, Hungary. Association for Computational Linguistics.