FEVEROUS

该数据集旨在对维基百科中的句子和表格证据进行事实核实，采用了一种两阶段模型，该模型首先检索证据，然后基于主张进行推理。该数据集包含了79,856个合成样本，其任务是进行事实核实。

This dataset is designed for fact verification on sentence and tabular evidence from Wikipedia. It adopts a two-stage model which first retrieves relevant evidence and then performs reasoning based on the target claim. Comprising 79,856 synthetic samples, this dataset focuses on the task of fact verification.