RussianNLP/coat

CoAT（人工文本语料库）是一个俄语的大型语料库，包含246k个人类编写的文本和13个神经模型生成的文本。这个语料库旨在用于人工文本检测和作者归属两种任务。数据集包括训练集、验证集和测试集。CoAT (Corpus of Artificial Texts) is a large-scale Russian corpus consisting of 246k human-written texts and texts generated by 13 neural models. It is designed for two tasks: detection of artificial texts and authorship attribution. The dataset includes training, validation, and test sets.