jrfish/Approximation_Political_Neutrality_Annotation_Dataset

该数据集伴随着一篇关于AI政治中立性的论文，包含了政治问题的标注生成。数据集包括输入提示、输出生成以及对政治中立性技术的注释。数据集的语言为英语，由Jillian Fisher策划。数据集的结构包括唯一ID、任务类型、输入编号、问题格式、提示、模型生成和模型注释等列。源数据和注释过程在论文的附录中有详细描述。

This dataset accompanies a paper on political neutrality in AI and contains annotated generations of political questions. It includes input prompts, output generations, and annotations of political neutrality techniques. The dataset is in English and curated by Jillian Fisher. The structure of the dataset consists of columns such as unique IDs, task types, input numbers, question formats, prompts, model generations, and model annotations. The source data and annotation procedure are described in the papers appendix.