wangbing1416/LonsRex-Misinfomation-SFT

该数据集是一个英语文本生成数据集，专注于错误信息、假新闻、可解释性、推理、知识和思维相关任务。数据规模在10万到100万条之间，包含两个配置版本：ours和Qwen3-Next。每个配置都定义了输入、输出和标签等特征字段，其中Qwen3-Next版本额外包含ID和来源字段。数据集旨在支持文本生成模型在相关领域的训练和评估。This dataset is an English text-generation dataset focused on tasks related to misinformation, fake news, explainability, reasoning, knowledge, and thinking. It has a size between 100,000 and 1,000,000 entries and includes two configurations: ours and Qwen3-Next. Each configuration specifies features such as input, output, and label, with the Qwen3-Next version additionally including ID and source fields. The dataset is designed to support the training and evaluation of text generation models in these domains.