verazuo/jailbreak_llms

[CCS'24] A dataset consists of 15,140 ChatGPT prompts from Reddit, Discord, websites, and open-source datasets (including 1,405 jailbreak prompts).

语言：Jupyter Notebook · GitHub 星标：3,867 · Fork：339

话题：chatgpt · jailbreak · jailbreaking · large-language-model · llm · llm-security · prompt

在 GitHub 上查看原仓库