Understanding and Mitigating Miscalibration in Prompt Tuning for Vision-Language Models

Open in new window