This is version 2 of the LogiQA dataset, first released as a multi-choice reading comprehension dataset. The dataset is collected from the Chinese Civil Service Entrance Examination. The dataset is both in Chinese and English (by translation)