@inproceedings{gromada-etal-2026-simulation,
title = "From Simulation to Reality: Validating Automated Evaluation of Conversational Agents",
author = "Gromada, Justyna and
Isakova, Ekaterina and
Brudnik, Joanna and
Kasicka, Alicja and
Krawczyk, Natalia and
Przyby{\l}, Bartosz and
Krajewski, Lukasz and
Szczerbak, Micha{\l} K.",
editor = "Atanasova, Pepa and
Haffari, Gholamreza and
Wang, Minghan",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: Industry Track",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-industry.46/",
pages = "763--784",
ISBN = "979-8-89176-460-6",
abstract = "Evaluating conversational agents typically requires human participants, a process that is costly and difficult to scale. LLM-based user simulators and LLM-as-a-judge evaluators offer a promising alternative, but how close they are to replacing real users remains unclear. We study this through a human user study (52 participants) and a simulated user study (four LLM-based simulators) applied to an experimental conversational sales agent, comparing LLM-as-a-judge assessments against human self-reported ratings. The results show that none of our simulators fully replicates human interaction patterns, and none of the LLM-based evaluators shows strong correlation with human ratings on real dialogues. We provide practical guidelines for combining automated and human evaluation, identifying key areas for improvement and highlighting when real-user studies remain essential."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="gromada-etal-2026-simulation">
<titleInfo>
<title>From Simulation to Reality: Validating Automated Evaluation of Conversational Agents</title>
</titleInfo>
<name type="personal">
<namePart type="given">Justyna</namePart>
<namePart type="family">Gromada</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Isakova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joanna</namePart>
<namePart type="family">Brudnik</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alicja</namePart>
<namePart type="family">Kasicka</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Natalia</namePart>
<namePart type="family">Krawczyk</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bartosz</namePart>
<namePart type="family">Przybył</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lukasz</namePart>
<namePart type="family">Krajewski</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Michał</namePart>
<namePart type="given">K</namePart>
<namePart type="family">Szczerbak</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: Industry Track</title>
</titleInfo>
<name type="personal">
<namePart type="given">Pepa</namePart>
<namePart type="family">Atanasova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gholamreza</namePart>
<namePart type="family">Haffari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Minghan</namePart>
<namePart type="family">Wang</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">979-8-89176-460-6</identifier>
</relatedItem>
<abstract>Evaluating conversational agents typically requires human participants, a process that is costly and difficult to scale. LLM-based user simulators and LLM-as-a-judge evaluators offer a promising alternative, but how close they are to replacing real users remains unclear. We study this through a human user study (52 participants) and a simulated user study (four LLM-based simulators) applied to an experimental conversational sales agent, comparing LLM-as-a-judge assessments against human self-reported ratings. The results show that none of our simulators fully replicates human interaction patterns, and none of the LLM-based evaluators shows strong correlation with human ratings on real dialogues. We provide practical guidelines for combining automated and human evaluation, identifying key areas for improvement and highlighting when real-user studies remain essential.</abstract>
<identifier type="citekey">gromada-etal-2026-simulation</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-industry.46/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>763</start>
<end>784</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T From Simulation to Reality: Validating Automated Evaluation of Conversational Agents
%A Gromada, Justyna
%A Isakova, Ekaterina
%A Brudnik, Joanna
%A Kasicka, Alicja
%A Krawczyk, Natalia
%A Przybył, Bartosz
%A Krajewski, Lukasz
%A Szczerbak, Michał K.
%Y Atanasova, Pepa
%Y Haffari, Gholamreza
%Y Wang, Minghan
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: Industry Track
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-460-6
%F gromada-etal-2026-simulation
%X Evaluating conversational agents typically requires human participants, a process that is costly and difficult to scale. LLM-based user simulators and LLM-as-a-judge evaluators offer a promising alternative, but how close they are to replacing real users remains unclear. We study this through a human user study (52 participants) and a simulated user study (four LLM-based simulators) applied to an experimental conversational sales agent, comparing LLM-as-a-judge assessments against human self-reported ratings. The results show that none of our simulators fully replicates human interaction patterns, and none of the LLM-based evaluators shows strong correlation with human ratings on real dialogues. We provide practical guidelines for combining automated and human evaluation, identifying key areas for improvement and highlighting when real-user studies remain essential.
%U https://aclanthology.org/2026.emnlp-industry.46/
%P 763-784
Markdown (Informal)
[From Simulation to Reality: Validating Automated Evaluation of Conversational Agents](https://aclanthology.org/2026.emnlp-industry.46/) (Gromada et al., EMNLP 2026)
ACL
- Justyna Gromada, Ekaterina Isakova, Joanna Brudnik, Alicja Kasicka, Natalia Krawczyk, Bartosz Przybył, Lukasz Krajewski, and Michał K. Szczerbak. 2026. From Simulation to Reality: Validating Automated Evaluation of Conversational Agents. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: Industry Track, pages 763–784, Budapest, Hungary. Association for Computational Linguistics.