Singular Value Few-shot Adaptation of Vision-Language Models Paper • 2509.03740 • Published Sep 3 • 2 • 2