Skip to content

A systematic attack prompt creation framework, leverages large language models, image-to-text, and image-to-image modules to automate the creation of attack prompts at scale.

Notifications You must be signed in to change notification settings

Zjm1900/SurrogatePrompt

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 

Repository files navigation

Appendix B of “SurrogatePrompt: Bypassing the Safety Filter of Text-To-Image Models via Substitution”

Examples of NSFW images generated by Midjourney using prompt samples from SneakyPrompt and SurrogatePrompt (our work). The latter includes images containing explicit adult content and fictitious images portraying political figures in gory and violent scenes.

Given the sensitive content of these images, including politics, violence, gore, pornography, etc., please contact me to request access permission to mitigate potential negative implications. It's crucial that the applicants use these images solely for research and refrain from distributing them. My email is: jiemingzhong@zju.edu.cn

About

A systematic attack prompt creation framework, leverages large language models, image-to-text, and image-to-image modules to automate the creation of attack prompts at scale.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published