Cultural alignment of Language Models and the effects of prompt language and cultural prompting

ACL ARR 2025 May Submission4304 Authors

Abstract: Culture is a core component of human-to-human interaction and plays a vital role in how we perceive and interact with others. Advancements in the effectiveness of Large Language Models (LLMs) in generating human-sounding text have greatly increased the amount of human-to-computer interaction. As this field grows, the cultural alignment of these human-like text agents becomes an important field of study. Our work uses Hofstede's VSM13 international surveys and prompts LLMs to answer questions to understand the cultural alignment of these models. We use a combination of prompt language and cultural prompting, a strategy that uses a system prompt to inform the model of the desired country, to shift the model's alignment to specific cultures. Our results show that DeepSeek-V3 exhibits a close alignment with the survey responses of the United States, and does not shift its alignment even when using cultural prompts from other cultures or changing the prompt language from English. We also find that GPT-4 exhibits an alignment closer to China when prompted in English, but cultural prompting is effective in shifting this alignment closer to the United States. Other low-cost models, GPT-4o and GPT-4.1, respond to the prompt language used (i.e., English or simplified Chinese) and cultural prompting strategies to create strong alignments with either the United States or China.

Paper Type: Long

Research Area: Language Modeling

Research Area Keywords: language/cultural bias analysis, values and culture, LLM/AI agents, prompting, safety and alignment

Contribution Types: Model analysis & interpretability, Data analysis

Languages Studied: English, Chinese

Submission Number: 4304