A u t o T r u s t

A Comprehensive Benchmark of Trustworthiness in VLMs for AD