@inproceedings{spliethover-etal-2026-cultural,
title = "Cultural (In)Competence: Investigating Culture-Specific Stereotype Detection Capabilities of {LLM}s",
author = {Splieth{\"o}ver, Maximilian and
L{\"o}tfering, Jannes and
Bachmann, Dominik and
Espig, Alexander and
Zuidema, Willem and
Schulz, Katrin and
Wachsmuth, Henning},
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.426/",
pages = "8964--8983",
ISBN = "979-8-89176-462-0",
abstract = "Stereotypes are subjective constructs that are mediated by the cultural context in which they are viewed. Successfully modeling them thus also requires cultural competence. When using large language models (LLMs) for stereotype-related tasks, such as their detection, it is therefore crucial that models can access relevant cultural knowledge and adjust responses accordingly. Since LLMs have shown disparities in what cultural knowledge is available to them, their capacity regarding stereotypes may also be disparate across languages and cultural contexts (e.g., for non-English languages or for non-US contexts). In this paper, we systematically investigate the effectiveness of LLMs for culture-specific stereotype detection across three factors: (1){~}the prompting language, (2){~}the instruction tuning language, and (3){~}the fine-tuning dataset size. We conduct experiments for US, Dutch, and Chinese cultural contexts. Our results suggest that LLMs do not reliably account for cultural contexts and overpredict the presence of stereotypes. Non-English prompts seem to improve consideration of cultural context and thus lead to more effective predictions. Moreover, we find that additional instruction tuning reduces false-positive rates and that training data size moderates detection effectiveness non-monotonically."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="spliethover-etal-2026-cultural">
<titleInfo>
<title>Cultural (In)Competence: Investigating Culture-Specific Stereotype Detection Capabilities of LLMs</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maximilian</namePart>
<namePart type="family">Spliethöver</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jannes</namePart>
<namePart type="family">Lötfering</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dominik</namePart>
<namePart type="family">Bachmann</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alexander</namePart>
<namePart type="family">Espig</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Willem</namePart>
<namePart type="family">Zuidema</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Katrin</namePart>
<namePart type="family">Schulz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Henning</namePart>
<namePart type="family">Wachsmuth</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>Stereotypes are subjective constructs that are mediated by the cultural context in which they are viewed. Successfully modeling them thus also requires cultural competence. When using large language models (LLMs) for stereotype-related tasks, such as their detection, it is therefore crucial that models can access relevant cultural knowledge and adjust responses accordingly. Since LLMs have shown disparities in what cultural knowledge is available to them, their capacity regarding stereotypes may also be disparate across languages and cultural contexts (e.g., for non-English languages or for non-US contexts). In this paper, we systematically investigate the effectiveness of LLMs for culture-specific stereotype detection across three factors: (1) the prompting language, (2) the instruction tuning language, and (3) the fine-tuning dataset size. We conduct experiments for US, Dutch, and Chinese cultural contexts. Our results suggest that LLMs do not reliably account for cultural contexts and overpredict the presence of stereotypes. Non-English prompts seem to improve consideration of cultural context and thus lead to more effective predictions. Moreover, we find that additional instruction tuning reduces false-positive rates and that training data size moderates detection effectiveness non-monotonically.</abstract>
<identifier type="citekey">spliethover-etal-2026-cultural</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.426/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>8964</start>
<end>8983</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Cultural (In)Competence: Investigating Culture-Specific Stereotype Detection Capabilities of LLMs
%A Spliethöver, Maximilian
%A Lötfering, Jannes
%A Bachmann, Dominik
%A Espig, Alexander
%A Zuidema, Willem
%A Schulz, Katrin
%A Wachsmuth, Henning
%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 spliethover-etal-2026-cultural
%X Stereotypes are subjective constructs that are mediated by the cultural context in which they are viewed. Successfully modeling them thus also requires cultural competence. When using large language models (LLMs) for stereotype-related tasks, such as their detection, it is therefore crucial that models can access relevant cultural knowledge and adjust responses accordingly. Since LLMs have shown disparities in what cultural knowledge is available to them, their capacity regarding stereotypes may also be disparate across languages and cultural contexts (e.g., for non-English languages or for non-US contexts). In this paper, we systematically investigate the effectiveness of LLMs for culture-specific stereotype detection across three factors: (1) the prompting language, (2) the instruction tuning language, and (3) the fine-tuning dataset size. We conduct experiments for US, Dutch, and Chinese cultural contexts. Our results suggest that LLMs do not reliably account for cultural contexts and overpredict the presence of stereotypes. Non-English prompts seem to improve consideration of cultural context and thus lead to more effective predictions. Moreover, we find that additional instruction tuning reduces false-positive rates and that training data size moderates detection effectiveness non-monotonically.
%U https://aclanthology.org/2026.findings-emnlp.426/
%P 8964-8983
Markdown (Informal)
[Cultural (In)Competence: Investigating Culture-Specific Stereotype Detection Capabilities of LLMs](https://aclanthology.org/2026.findings-emnlp.426/) (Spliethöver et al., Findings 2026)
ACL