Mirroring Cultural Dominance

Disclosing Large Language Models Social Values, Attitudes and Stereotypes

Bibliographic Data

ID13024474
AuthorsKristian Đokić (0000-0003-4358-9065, University of Osijek, corresponding author), Barbara Pisker (0000-0001-9434-5541, University of Osijek), Bojan Radišić (0000-0001-7870-193X, University of Osijek)
Year2025
Volume15
Issue5
Pages142-142
Publication date2025-05-21
Open AccessYes
TypeARTICLE
VenueSocieties (JOURNAL)
Journal identifiersISSN: 2075-4698 • E-ISSN: 2075-4698
PublisherMDPI AG (PUBLISHER • IT)
DOI10.3390/soc15050142
OpenAlexW4410560601
LanguageEN
Citations received1
References cited30

The paper aims to address large language models’ (LLMs) cultural bias using the World Value Survey Wave 7 (WVS) questionnaire on social values, attitudes, and stereotypes. Comparative analysis and LLMs interview methods measure the Euclidean distance of response vectors of four culturally diverse LLMs (USA, China, Russia, UAE) in a multidimensional vector space to contrast originated WVS research countries and population positions. The results confirmed the initial hypotheses reflecting culturally and linguistically biased LLM answers, considering specific socio-cultural contexts and English language and Latin script digital dominance in available training materials. USA-constructed LLMs showed the most liberal attitudes, followed by China, Russia, and the UAE. LLM interview results also show WVS results closest to the United States population, positioning the similarity of the responses in first place for China and Russia followed by the USA and the UAE. Mitigating initiatives in LLMs’ cultural and linguistic debiasing is required to preserve cultural and linguistic diversity in the digital space

Dominance (genetics · Mirroring · Sociology · Multilingual Education and Policy · Psychology · Social Media and Politics · Social Psychology

Tools

Open DOIOpen Access