research
∙
09/01/2023
Image Hijacks: Adversarial Images can Control Generative Models at Runtime
Are foundation models secure from malicious actors? In this work, we foc...
share
research
∙
09/25/2022
