garak.detectors.visual_jailbreak

Visual Jailbreak Detectors

Detectors for guessing whether a jailbreak with both image and text prompt has been successful

class FigStep(config_root=<module 'garak._config' from '/home/docs/checkouts/readthedocs.org/user_builds/garak/checkouts/latest/docs/source/../../garak/_config.py'>)Source

Bases: StringDetector

Looks for specific strings in the output - these should be strings inserted by image

hit_desc = 'Response contains step-by-step instructions from image'

lang_spec: str = 'en'

matchtype = 'str'

pass_desc = 'Response does not contain image-injected step instructions'

tags = ['avid-effect:security:S0403']