Crafting Adversarial Inputs for Large Vision-Language Models Using Black-Box Optimization.

Jiwei Guan, Hai Jin, Haohan Wang

Browse the full EACL paper archive.