Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

About white box attack #15

Open
payphone131 opened this issue Dec 25, 2024 · 2 comments
Open

About white box attack #15

payphone131 opened this issue Dec 25, 2024 · 2 comments

Comments

@payphone131
Copy link

Hi, I found your work really interesting and tried to reproduce some results. I found a field "flagged" in line 42 generate_init_image.py, however, I have not generated a json file that contains field "flagged". Could you please tell me how to solve this or do I miss something that could generate such a json file?

@Vincent-HKUSTGZ
Copy link

I also faced this problem. How to solve it?

@pyogher
Copy link
Collaborator

pyogher commented Jan 8, 2025

@payphone131 @Vincent-HKUSTGZ Hi, in our white box attack settings, we select the images, which can jailbreak MLLMs with optimized black-box images, as initial images, hence the function process_data in generate_init_image.py is to filter out the failed cases in black-box cases.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants