Wiki Knowledge (Unpopular)

该数据集名为Wiki Knowledge，包含了较少被查看的维基百科文章，既有全文也有截短版本。此外，数据集还包含了针对回应长度的控制实验。规模上，我们抽取了500个实例进行研究。该数据集的任务是生成知识密集型的回应。

The dataset, named Wiki Knowledge, consists of under-viewed Wikipedia articles available in both full-text and truncated versions. Additionally, it includes controlled experiments targeting response length. For this research, we sampled 500 instances. The core task of this dataset is to generate knowledge-intensive responses.