shalanova/benchmark-2-chinese-gt
收藏资源简介:
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。数据集包含四个列:text(原始提示)、label(0表示安全,1表示不安全)、translation(通过Google Translate翻译成中文的提示)、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, primarily containing prompt-injection and canonical jailbreak-style instructions with relatively homogeneous attack patterns. The dataset size is 1,000 prompts (500 safe / 500 unsafe). It includes four columns: `text` (original prompt), `label` (`0`: safe, `1`: unsafe), `translation` (prompt translated into Chinese by Google Translate), and `score_zh_google` (cosine similarity score with codebook). More information can be found in the provided paper link.




