ImgTrojan: Jailbreaking Vision-Language Models with ONE Image

Xijia Tao, Shuai Zhong, Lei Li, Qi Liu, Lingpeng Kong
March, 2024
PDF
Type
Conference paper
Publication
In arXiv