Cross-Cultural Transfer of Commonsense Reasoning in LLMs: Evidence from the Arab World

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Almheiri, Saeed, Hossam, Rania, Attia, Mena, Wang, Chenxi, Nakov, Preslav, Baldwin, Timothy, Koto, Fajri
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916964942741504
author Almheiri, Saeed
Hossam, Rania
Attia, Mena
Wang, Chenxi
Nakov, Preslav
Baldwin, Timothy
Koto, Fajri
author_facet Almheiri, Saeed
Hossam, Rania
Attia, Mena
Wang, Chenxi
Nakov, Preslav
Baldwin, Timothy
Koto, Fajri
contents Large language models (LLMs) often reflect Western-centric biases, limiting their effectiveness in diverse cultural contexts. Although some work has explored cultural alignment, the potential for cross-cultural transfer, using alignment in one culture to improve performance in others, remains underexplored. This paper investigates cross-cultural transfer of commonsense reasoning in the Arab world, where linguistic and historical similarities coexist with local cultural differences. Using a culturally grounded commonsense reasoning dataset covering 13 Arab countries, we evaluate lightweight alignment methods such as in-context learning and demonstration-based reinforcement (DITTO), alongside baselines like supervised fine-tuning and direct preference optimization. Our results show that merely 12 culture-specific examples from one country can improve performance in others by 10\% on average, within multilingual models. In addition, we demonstrate that out-of-culture demonstrations from Indonesia and US contexts can match or surpass in-culture alignment for MCQ reasoning, highlighting cultural commonsense transferability beyond the Arab world. These findings demonstrate that efficient cross-cultural alignment is possible and offer a promising approach to adapt LLMs to low-resource cultural settings.
format Preprint
id arxiv_https___arxiv_org_abs_2509_19265
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Cross-Cultural Transfer of Commonsense Reasoning in LLMs: Evidence from the Arab World
Almheiri, Saeed
Hossam, Rania
Attia, Mena
Wang, Chenxi
Nakov, Preslav
Baldwin, Timothy
Koto, Fajri
Artificial Intelligence
Computation and Language
Large language models (LLMs) often reflect Western-centric biases, limiting their effectiveness in diverse cultural contexts. Although some work has explored cultural alignment, the potential for cross-cultural transfer, using alignment in one culture to improve performance in others, remains underexplored. This paper investigates cross-cultural transfer of commonsense reasoning in the Arab world, where linguistic and historical similarities coexist with local cultural differences. Using a culturally grounded commonsense reasoning dataset covering 13 Arab countries, we evaluate lightweight alignment methods such as in-context learning and demonstration-based reinforcement (DITTO), alongside baselines like supervised fine-tuning and direct preference optimization. Our results show that merely 12 culture-specific examples from one country can improve performance in others by 10\% on average, within multilingual models. In addition, we demonstrate that out-of-culture demonstrations from Indonesia and US contexts can match or surpass in-culture alignment for MCQ reasoning, highlighting cultural commonsense transferability beyond the Arab world. These findings demonstrate that efficient cross-cultural alignment is possible and offer a promising approach to adapt LLMs to low-resource cultural settings.
title Cross-Cultural Transfer of Commonsense Reasoning in LLMs: Evidence from the Arab World
topic Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2509.19265