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

Shuoyuan Wang, Yixuan Li, Hongxin Wei

Browse the full ICML paper archive.