[1]
Y. Gong, “FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts”, AAAI, vol. 39, no. 22, pp. 23951–23959, Apr. 2025.