Text-Guided Attention is All You Need for Zero-Shot Robustness in Vision-Language Models Lu Y u

Open in new window