Mint: A Simple Test-Time Adaptation of Vision-Language Models against Common Corruptions

Open in new window