@inproceedings{fahmid-etal-2026-cultact,
title = "{C}ult{A}ct-{ML}: Measuring Cultural Value Defaults in Multilingual {LLM} Agents Under Action",
author = "Fahmid, Ahmad and
Mim, Jannatul Ferdous and
Arefeen, Md Adnan",
editor = "Ranaldi, Leonardo and
Pucci, Giulia and
Ranaldi, Federico and
Valentino, Marco and
Birch, Alexandra and
Zanzotto, Fabio Massimo",
booktitle = "The Proceedings of the Workshop on Open Reasoning Across Cultures Languages ({ORACLE})",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.oracle-main.22/",
pages = "329--338",
ISBN = "979-8-89176-483-5",
abstract = "Measuring which cultural values a languagemodel agent falls back on when it must act is a fundamental problem in multilingual and culturally-grounded evaluation. Current agent benchmarks are designed primarily to measure task success in high-resource languages such as English, which creates major limitations when the question is whose culture an agent{'}s committed action encodes. The key challenges are the absence of an acting-based value-conflict benchmark and the difficulty of preserving culturallyloaded meaning when a scenario is carried into a low-resource language such as Bangla. In this paper, to study these cultural defaults under action, we present CultAct-ML, a benchmark of 39 value-conflict dilemmas recast as secondperson agent actions and evaluated with a twophase plan{--}act agent over English, Indonesian, and Bangla. We further introduce three metrics that measure the Western-cluster default, its reliability across languages, and the drift of the agent{'}s plan, and every Bangla item is verified by a native speaker. Across three models and three repeats, a total of 1,053 plan{--}act decisions are produced, all valid. We find that a Western default holds above chance for Qwen2.5-72B and Kimi K2 in every language but fades to chance for Qwen3-235B outside English. After stratifying the plan drift by the language of the plan, we further find that Qwen3-235B plans in English rather than Bangla on 70.1{\%} of Bangla scenarios, highlighting a cross-language effect that a pooled analysis hides entirely. The scenarios, code, and all metrics are publicly released.1"
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="fahmid-etal-2026-cultact">
<titleInfo>
<title>CultAct-ML: Measuring Cultural Value Defaults in Multilingual LLM Agents Under Action</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ahmad</namePart>
<namePart type="family">Fahmid</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jannatul</namePart>
<namePart type="given">Ferdous</namePart>
<namePart type="family">Mim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Md</namePart>
<namePart type="given">Adnan</namePart>
<namePart type="family">Arefeen</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>The Proceedings of the Workshop on Open Reasoning Across Cultures Languages (ORACLE)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Leonardo</namePart>
<namePart type="family">Ranaldi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Giulia</namePart>
<namePart type="family">Pucci</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Federico</namePart>
<namePart type="family">Ranaldi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marco</namePart>
<namePart type="family">Valentino</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alexandra</namePart>
<namePart type="family">Birch</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fabio</namePart>
<namePart type="given">Massimo</namePart>
<namePart type="family">Zanzotto</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">979-8-89176-483-5</identifier>
</relatedItem>
<abstract>Measuring which cultural values a languagemodel agent falls back on when it must act is a fundamental problem in multilingual and culturally-grounded evaluation. Current agent benchmarks are designed primarily to measure task success in high-resource languages such as English, which creates major limitations when the question is whose culture an agent’s committed action encodes. The key challenges are the absence of an acting-based value-conflict benchmark and the difficulty of preserving culturallyloaded meaning when a scenario is carried into a low-resource language such as Bangla. In this paper, to study these cultural defaults under action, we present CultAct-ML, a benchmark of 39 value-conflict dilemmas recast as secondperson agent actions and evaluated with a twophase plan–act agent over English, Indonesian, and Bangla. We further introduce three metrics that measure the Western-cluster default, its reliability across languages, and the drift of the agent’s plan, and every Bangla item is verified by a native speaker. Across three models and three repeats, a total of 1,053 plan–act decisions are produced, all valid. We find that a Western default holds above chance for Qwen2.5-72B and Kimi K2 in every language but fades to chance for Qwen3-235B outside English. After stratifying the plan drift by the language of the plan, we further find that Qwen3-235B plans in English rather than Bangla on 70.1% of Bangla scenarios, highlighting a cross-language effect that a pooled analysis hides entirely. The scenarios, code, and all metrics are publicly released.1</abstract>
<identifier type="citekey">fahmid-etal-2026-cultact</identifier>
<location>
<url>https://aclanthology.org/2026.oracle-main.22/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>329</start>
<end>338</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CultAct-ML: Measuring Cultural Value Defaults in Multilingual LLM Agents Under Action
%A Fahmid, Ahmad
%A Mim, Jannatul Ferdous
%A Arefeen, Md Adnan
%Y Ranaldi, Leonardo
%Y Pucci, Giulia
%Y Ranaldi, Federico
%Y Valentino, Marco
%Y Birch, Alexandra
%Y Zanzotto, Fabio Massimo
%S The Proceedings of the Workshop on Open Reasoning Across Cultures Languages (ORACLE)
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-483-5
%F fahmid-etal-2026-cultact
%X Measuring which cultural values a languagemodel agent falls back on when it must act is a fundamental problem in multilingual and culturally-grounded evaluation. Current agent benchmarks are designed primarily to measure task success in high-resource languages such as English, which creates major limitations when the question is whose culture an agent’s committed action encodes. The key challenges are the absence of an acting-based value-conflict benchmark and the difficulty of preserving culturallyloaded meaning when a scenario is carried into a low-resource language such as Bangla. In this paper, to study these cultural defaults under action, we present CultAct-ML, a benchmark of 39 value-conflict dilemmas recast as secondperson agent actions and evaluated with a twophase plan–act agent over English, Indonesian, and Bangla. We further introduce three metrics that measure the Western-cluster default, its reliability across languages, and the drift of the agent’s plan, and every Bangla item is verified by a native speaker. Across three models and three repeats, a total of 1,053 plan–act decisions are produced, all valid. We find that a Western default holds above chance for Qwen2.5-72B and Kimi K2 in every language but fades to chance for Qwen3-235B outside English. After stratifying the plan drift by the language of the plan, we further find that Qwen3-235B plans in English rather than Bangla on 70.1% of Bangla scenarios, highlighting a cross-language effect that a pooled analysis hides entirely. The scenarios, code, and all metrics are publicly released.1
%U https://aclanthology.org/2026.oracle-main.22/
%P 329-338
Markdown (Informal)
[CultAct-ML: Measuring Cultural Value Defaults in Multilingual LLM Agents Under Action](https://aclanthology.org/2026.oracle-main.22/) (Fahmid et al., ORACLE 2026)
ACL