Why Is Prompt Tuning for Vision-Language Models Robust to Noisy Labels? | IEEE Conference Publication | IEEE Xplore