Yige Li

Privacy (1)

large language model (3) backdoor attack (3) vision-language model (2) adversarial learning (2) adversarial attack (2) backdoor defense (2) safety alignment (1) model safety (1) feature extraction (1) model behavior (1) model unlearning (1) machine unlearning (1) neural network security (1) model alignment (1) multimodal learning (1) empirical evaluation (1) approximation error (1) self-supervised learning (1) deep neural network (1) model security (1)