Seanie-lee/ThinkSafe-R1-Distill-8B-v3
收藏Hugging Face2025-10-31 更新2025-11-15 收录
下载链接:
https://hf-mirror.com/datasets/Seanie-lee/ThinkSafe-R1-Distill-8B-v3
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含了模型对有害提示的拒绝统计信息。在40000个示例中,模型拒绝了4654个有害提示,并生成了13234个拒绝的有害提示,同时识别出22112个良性提示。经过LlamaGuard过滤后,共有39750个示例被保存在文件中。
The dataset contains statistics on the models refusals of harmful prompts. Out of 40000 examples, the model refused 4654 harmful prompts and generated 13234 refusals for harmful prompts, while identifying 22112 benign prompts. After LlamaGuard filtering, a total of 39750 examples were saved to a file.
提供机构:
Seanie-lee



