shalanova/benchmark-1-chinese-gt
收藏Hugging Face2026-04-30 更新2026-05-03 收录
下载链接:
https://hf-mirror.com/datasets/shalanova/benchmark-1-chinese-gt
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是通过Google Translate翻译成中文的,主要包含prompt-injection和典型的jailbreak-style指令,具有相对同质的攻击模式。数据集大小为1,000个提示,其中500个为安全提示,500个为不安全提示。数据集包含以下列:text(原始提示)、label(0表示安全,1表示不安全)、translation(通过Google Translate翻译的中文提示)和score_zh_google(与codebook的余弦相似度得分)。更多详细信息可参考提供的论文链接。
This dataset is translated into Chinese by Google Translate and primarily contains prompt-injection and canonical jailbreak-style instructions with relatively homogeneous attack patterns. The dataset size is 1,000 prompts (500 safe / 500 unsafe). The columns include: text (original prompt), label (0: safe, 1: unsafe), translation (prompt in Chinese translated by Google Translate), and score_zh_google (cosine similarity score with codebook). More information can be found in the provided paper link.
提供机构:
shalanova



