探究预训练语言模型中价值观的跨文化差异
Probing Pre-Trained Language Models for Cross-Cultural Differences in Values
AI总结：
本文提出用探针方法检测PTLM中的跨文化价值观，发现其能捕捉文化差异但与既有价值观调查仅微弱一致，并讨论了对齐方法。
摘要：
语言蕴含着人们所持社会、文化和政治价值观的信息。先前工作已探究预训练语言模型（PTLM）中编码的社会偏见及潜在有害偏见。然而，尚无系统性研究考察这些模型中嵌入的价值观如何随文化而变化。本文引入探针方法，研究这些模型嵌入了哪些跨文化价值观，以及它们是否与现有理论和跨文化价值观调查相一致。我们发现，PTLM能够捕捉跨文化价值观差异，但这些差异与既有价值观调查仅存在微弱一致性。本文讨论了在跨文化场景中使用不一致模型的影响，以及使PTLM与价值观调查对齐的方法。
英文摘要：
Language embeds information about social, cultural, and political values people hold. Prior work has explored social and potentially harmful biases encoded in Pre-Trained Language models (PTLMs). However, there has been no systematic study investigating how values embedded in these models vary across cultures. In this paper, we introduce probes to study which values across cultures are embedded in these models, and whether they align with existing theories and cross-cultural value surveys. We find that PTLMs capture differences in values across cultures, but those only weakly align with established value surveys. We discuss implications of using mis-aligned models in cross-cultural settings, as well as ways of aligning PTLMs with value surveys.
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。
