arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.10436cs.CL
收藏

我不是他们:大型语言模型中的流动身份与持续的外群体偏见

I Am Not Them: Fluid Identities and Persistent Out-group Bias in Large Language Models

KAIST(韩国科学技术院);Institute for Basic Science(基础科学研究院);Max Planck Institute for Security and Privacy(马克斯·普朗克安全与隐私研究所)

Wenchao Dong, Assem Zhunis, Hyojin Chin, Jiyoung Han, Meeyoung Cha

AI总结:

本研究探究了 ChatGPT 在多语言下的文化偏见,发现模型在获得特定社会身份时会区分内外群体,拥抱内群体价值观而摒弃外群体价值观,且外群体负面偏见强于内群体正面倾向,该现象在政治领域及其他模型中均表现出一致性。

摘要:

我们探究了 ChatGPT 在三种西方语言(即英语、德语和法语)与三种东方语言(即中文、日语和韩语)中的文化偏见——即个人主义与集体主义。当 ChatGPT 在西方语言中采用个人主义人格时,其集体主义得分(即外群体价值观)呈现出更明显的负面趋势,甚至超过了其对个人主义(即内群体价值观)的正面倾向。相反,当在东方语言中为 ChatGPT 分配集体主义人格时,出现了类似模式,相较于集体主义(即内群体价值观),对个人主义(即外群体价值观)的负面反应更多。结果表明,当被赋予特定社会身份时,ChatGPT 能够区分内群体与外群体,拥抱内群体价值观而摒弃外群体价值观。值得注意的是,对外群体的负面态度(偏见与歧视由此产生)超过了对内群体的正面态度。该实验在政治领域进行了重复,结果保持一致。此外,此次重复实验揭示了大型语言模型(LLMs)固有的民主党偏见,这与早期发现相一致,并为通过提示工程缓解此类偏见提供了重要见解。使用不同的超参数和人格设置方法,在有或无社会身份标签的情况下,对其他流行语言模型进行了广泛的鲁棒性检验。

英文摘要:

We explored cultural biases-individualism vs. collectivism-in ChatGPT across three Western languages (i.e., English, German, and French) and three Eastern languages (i.e., Chinese, Japanese, and Korean). When ChatGPT adopted an individualistic persona in Western languages, its collectivism scores (i.e., out-group values) exhibited a more negative trend, surpassing their positive orientation towards individualism (i.e., in-group values). Conversely, when a collectivistic persona was assigned to ChatGPT in Eastern languages, a similar pattern emerged with more negative responses toward individualism (i.e., out-group values) as compared to collectivism (i.e., in-group values). The results indicate that when imbued with a particular social identity, ChatGPT discerns in-group and out-group, embracing in-group values while eschewing out-group values. Notably, the negativity towards the out-group, from which prejudices and discrimination arise, exceeded the positivity towards the in-group. The experiment was replicated in the political domain, and the results remained consistent. Furthermore, this replication unveiled an intrinsic Democratic bias in Large Language Models (LLMs), aligning with earlier findings and providing integral insights into mitigating such bias through prompt engineering. Extensive robustness checks were performed using varying hyperparameter and persona setup methods, with or without social identity labels, across other popular language models.

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。

↑