One Prompt Word is Enough to Boost Adversarial Robustness for Pre-trained Vision-Language Models

Devs

One Prompt Word is Enough to Boost Adversarial Robustness for Pre-trained Vision-Language Models | Read Paper on Bytez