CulturePark Cultural Samples

该数据集由41,000个文化样本组成，这些样本是通过一个由大型语言模型（LLM）驱动的多代理通信框架生成的，该框架模拟了跨文化的人类交流。该数据集包含了旨在体现人类信仰、规范和习俗的跨文化对话，并用于对八个特定文化的大型语言模型进行微调。这些模型的任务包括内容审核、文化对齐和文化教育。This dataset consists of 41,000 cultural samples, which are generated via a multi-agent communication framework powered by Large Language Models (LLMs) that simulates cross-cultural human communication. This dataset contains cross-cultural dialogues designed to embody human beliefs, norms and customs, and is used for fine-tuning Large Language Models for eight specific cultures. The tasks of these models include content moderation, cultural alignment and cultural education.