GloVe

该数据集是由GloVe模型生成的非上下文嵌入，被用作在探测任务中作为比较的基准。此外，该数据集还用于确立每种信息类型是如何被单词的非上下文表示所捕捉的。在此基础上，开展了非上下文嵌入的基准探测任务。This dataset comprises non-contextual embeddings generated by the GloVe model, which acts as a comparative baseline for probing tasks. Moreover, it is used to elucidate how each type of information is captured by non-contextual word representations. Building on this foundation, benchmark probing tasks for non-contextual embeddings were conducted.