Zhengyu Zhao

Seeing It or Not? Interpretable Vision-aware Latent Steering to Mitigate Object Hallucinations

AdvFLYP: Adversarially Finetune Like You Pretrain for Zero-shot Robustness of CLIP

Pay Less Attention to Function Words for Free Robustness of Vision-Language Models

Enhancing Query-Free Jailbreaks on Text-to-Image Models with Bimodal Guidance

Prediction Inconsistency Helps Generalizable Detection of Adversarial Examples

Robustness Over Time: Understanding Adversarial Examples’ Effectiveness on Longitudinal Versions of Large Language Models

Breaking Semantic Artifacts for Generalized AI-generated Image Detection

Composite Backdoor Attacks Against Large Language Models

Prompt Backdoors in Visual Prompt Learning