X. Gao, “Safety Alignment of Large Language Models: The Offensive-Defensive Game Mechanism under Multi-round Adversarial Prompting”, SCJ, vol. 1, pp. 33–37, Jul. 2026.