arXivDaily arXiv每日学术速递 周一至周五更新
​通知:联邦节假日调休:哥伦布日 / 原住民日(Columbus Day / Indigenous Peoples' Day),今日(13日)无更新。
arXiv 2512.12488cs.CL

机器中的美国幽灵:语言模型如何实现文化对齐以及文化提示的影响

The American Ghost in the Machine: How language models align culturally and the effects of cultural prompting

School of Data Science University of Virginia(数据科学学院 多伦多大学)

James Luther, Donald Brown

AI总结:

本研究通过文化提示测试了多种LLM对不同文化适应性,发现多数模型默认倾向美国,但对日本和中国文化对齐存在困难。

摘要:

文化是人类互动的基础;它决定了我们如何感知和回应日常互动。随着生成式大语言模型(LLMs)的兴起,人机交互领域不断扩大,这些模型的文化对齐性成为研究的重要领域。本研究使用VSM13国际调查和霍夫斯泰德的文化维度,识别了流行LLMs(DeepSeek-V3、V3.1、GPT-5、GPT-4.1、GPT-4、Claude Opus 4、Llama 3.1和Mistral Large)的文化对齐情况。然后,我们使用文化提示,即通过系统提示将模型的文化对齐转向目标国家,来测试这些模型适应其他文化的能力,即中国、法国、印度、伊朗、日本和美国。我们发现,测试的八款LLMs在未指定文化时,大多数倾向于美国,当提示其他文化时结果各异。当使用文化提示时,七款模型更接近预期文化。我们发现,尽管有两款测试模型源自中国公司DeepSeek,但模型在对齐日本和中国时仍存在困难。

英文摘要:

Culture is the bedrock of human interaction; it dictates how we perceive and respond to everyday interactions. As the field of human-computer interaction grows via the rise of generative Large Language Models (LLMs), the cultural alignment of these models become an important field of study. This work, using the VSM13 International Survey and Hofstede's cultural dimensions, identifies the cultural alignment of popular LLMs (DeepSeek-V3, V3.1, GPT-5, GPT-4.1, GPT-4, Claude Opus 4, Llama 3.1, and Mistral Large). We then use cultural prompting, or using system prompts to shift the cultural alignment of a model to a desired country, to test the adaptability of these models to other cultures, namely China, France, India, Iran, Japan, and the United States. We find that the majority of the eight LLMs tested favor the United States when the culture is not specified, with varying results when prompted for other cultures. When using cultural prompting, seven of the eight models shifted closer to the expected culture. We find that models had trouble aligning with Japan and China, despite two of the models tested originating with the Chinese company DeepSeek.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑