kai-xu/BCT-MMLU-Test-V2

该数据集包含三个主要特征：prompt_unbiased、prompt_biased和answer。每个prompt特征是一个包含content和role字段的列表，两者均为字符串类型。answer特征是一个单独的字符串。数据集有一个名为train的分割，包含2000个示例，总大小为2585483字节。下载大小为1221974字节。配置指定了train分割的数据文件。

The dataset contains three main features: prompt_unbiased, prompt_biased, and answer. Each prompt feature is a list with content and role fields, both of string type. The answer feature is a single string. The dataset has a single split named train with 2000 examples and a total size of 2585483 bytes. The download size is 1221974 bytes. The configuration specifies the data files for the train split.