Dr4kl3s/global_mmlu_lite_core_grid_seed17

该数据集是一个多语言多项选择题的集合，包含多种语言（阿拉伯语、孟加拉语、德语、英语、西班牙语、法语、印地语、印尼语、意大利语、葡萄牙语、中文）的训练和测试分割。每个样本包含样本ID、主题、主题类别、问题、四个选项、答案、所需知识、时间敏感性、参考、文化、地区、国家、文化敏感性标签和是否已标注等特征。

This dataset is a multilingual collection of multiple-choice questions, including train and test splits for various languages (Arabic, Bengali, German, English, Spanish, French, Hindi, Indonesian, Italian, Portuguese, Chinese). Each sample contains features such as sample ID, subject, subject category, question, four options, answer, required knowledge, time sensitivity, reference, culture, region, country, cultural sensitivity label, and whether it is annotated.