MedicalHallucination/PubMedQA_Qwen14B_TG_1000

该数据集包含了问题、相关知识、正确答案以及五个生成的虚拟答案（及其解释和判别器欺骗情况）、难度等级信息。每个问题对应五个虚拟答案，每个答案都有是否欺骗了三个不同判别器的标记以及难度等级。此外，数据集还提供了与正确答案最不相似的答案和最终的问题难度等级。数据集包含训练集，共有1000个示例，文件大小为3258477字节。

The dataset includes questions, related knowledge, correct answers, and five generated hypothetical answers (with explanations and discriminator deception status), and difficulty levels. Each question corresponds to five hypothetical answers, each with a mark indicating whether it deceived three different discriminators and a difficulty level. In addition, the dataset also includes the least similar answer to the correct answer and the final question difficulty level. The dataset contains a training set with 1000 examples, totaling 3258477 bytes in size.