@inproceedings{huq-etal-2026-modeling,
title = "Modeling Distinct Human Interaction in Web Agents",
author = "Huq, Faria and
Wang, Zora Zhiruo and
Guo, Zhanqiu and
Arangarajan, Venu Arvind and
Ou, Tianyue and
Xu, Frank F. and
Zhou, Shuyan and
Neubig, Graham and
Bigham, Jeffrey P.",
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.2029/",
doi = "10.18653/2026.findings-emnlp.2029",
pages = "43195--43212",
ISBN = "979-8-89176-462-0",
abstract = "Despite rapid progress in autonomous web agents, human involvement remains essential for shaping preferences and correcting agent behavior as tasks unfold. However, current agentic systems lack a principled understanding of when and why humans intervene, often proceeding autonomously past critical decision points or requesting unnecessary confirmation. In this work, we introduce the task of modeling human intervention to support collaborative web task execution. We collect CowCorpus, a dataset of 400 real-user web navigation trajectories containing over 4200 interleaved human and agent actions. We identify four distinct patterns of user interaction with agents {--} hands-off supervision, hands-on oversight, collaborative task-solving, and full user takeover. Leveraging these insights, we train language models (LMs) to anticipate when users are likely to intervene based on their interaction styles, yielding 11.6-18.4{\%} improvement over base LMs. Finally, we deploy these intervention-aware models in live web navigation agents and evaluate them in a user study, finding a 36.8{\%} increase in user-rated agent usefulness. Together, our results show structured modeling of human intervention leads to more adaptive, collaborative agents."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="huq-etal-2026-modeling">
<titleInfo>
<title>Modeling Distinct Human Interaction in Web Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Faria</namePart>
<namePart type="family">Huq</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zora</namePart>
<namePart type="given">Zhiruo</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhanqiu</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Venu</namePart>
<namePart type="given">Arvind</namePart>
<namePart type="family">Arangarajan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianyue</namePart>
<namePart type="family">Ou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Frank</namePart>
<namePart type="given">F</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shuyan</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Graham</namePart>
<namePart type="family">Neubig</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jeffrey</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Bigham</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>Despite rapid progress in autonomous web agents, human involvement remains essential for shaping preferences and correcting agent behavior as tasks unfold. However, current agentic systems lack a principled understanding of when and why humans intervene, often proceeding autonomously past critical decision points or requesting unnecessary confirmation. In this work, we introduce the task of modeling human intervention to support collaborative web task execution. We collect CowCorpus, a dataset of 400 real-user web navigation trajectories containing over 4200 interleaved human and agent actions. We identify four distinct patterns of user interaction with agents – hands-off supervision, hands-on oversight, collaborative task-solving, and full user takeover. Leveraging these insights, we train language models (LMs) to anticipate when users are likely to intervene based on their interaction styles, yielding 11.6-18.4% improvement over base LMs. Finally, we deploy these intervention-aware models in live web navigation agents and evaluate them in a user study, finding a 36.8% increase in user-rated agent usefulness. Together, our results show structured modeling of human intervention leads to more adaptive, collaborative agents.</abstract>
<identifier type="citekey">huq-etal-2026-modeling</identifier>
<identifier type="doi">10.18653/2026.findings-emnlp.2029</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.2029/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>43195</start>
<end>43212</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Modeling Distinct Human Interaction in Web Agents
%A Huq, Faria
%A Wang, Zora Zhiruo
%A Guo, Zhanqiu
%A Arangarajan, Venu Arvind
%A Ou, Tianyue
%A Xu, Frank F.
%A Zhou, Shuyan
%A Neubig, Graham
%A Bigham, Jeffrey P.
%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 huq-etal-2026-modeling
%X Despite rapid progress in autonomous web agents, human involvement remains essential for shaping preferences and correcting agent behavior as tasks unfold. However, current agentic systems lack a principled understanding of when and why humans intervene, often proceeding autonomously past critical decision points or requesting unnecessary confirmation. In this work, we introduce the task of modeling human intervention to support collaborative web task execution. We collect CowCorpus, a dataset of 400 real-user web navigation trajectories containing over 4200 interleaved human and agent actions. We identify four distinct patterns of user interaction with agents – hands-off supervision, hands-on oversight, collaborative task-solving, and full user takeover. Leveraging these insights, we train language models (LMs) to anticipate when users are likely to intervene based on their interaction styles, yielding 11.6-18.4% improvement over base LMs. Finally, we deploy these intervention-aware models in live web navigation agents and evaluate them in a user study, finding a 36.8% increase in user-rated agent usefulness. Together, our results show structured modeling of human intervention leads to more adaptive, collaborative agents.
%R 10.18653/2026.findings-emnlp.2029
%U https://aclanthology.org/2026.findings-emnlp.2029/
%U https://doi.org/10.18653/2026.findings-emnlp.2029
%P 43195-43212
Markdown (Informal)
[Modeling Distinct Human Interaction in Web Agents](https://aclanthology.org/2026.findings-emnlp.2029/) (Huq et al., Findings 2026)
ACL
- Faria Huq, Zora Zhiruo Wang, Zhanqiu Guo, Venu Arvind Arangarajan, Tianyue Ou, Frank F. Xu, Shuyan Zhou, Graham Neubig, and Jeffrey P. Bigham. 2026. Modeling Distinct Human Interaction in Web Agents. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 43195–43212, Budapest, Hungary. Association for Computational Linguistics.