Universal Adversarial Perturbations for Vision-Language Pre-trained Models Paper • 2405.05524 • Published May 9, 2024 • 1