mijabench/jailbreaking_prompts_answer_cls2
收藏Hugging Face2025-12-16 更新2025-12-20 收录
下载链接：
https://hf-mirror.com/datasets/mijabench/jailbreaking_prompts_answer_cls2
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含文本内容、语言、数据集来源、是否有害标记、场景类别、场景描述、jb策略、jb提示以及多个不同模型（如Llama-3、Qwen3、gemma等）的jb响应和分类结果。数据集主要用于评估不同模型在特定场景下的响应和分类能力，包含44000个训练样本。
This dataset includes text content, language, dataset source, harmful flag, scenario category, scenario description, jb strategy, jb prompts, and jb responses and classification results from various models (e.g., Llama-3, Qwen3, gemma, etc.). It is primarily used to evaluate the response and classification capabilities of different models in specific scenarios, containing 44,000 training examples.
提供机构：
mijabench
