Gao, X. 2026. Safety Alignment of Large Language Models: The Offensive-Defensive Game Mechanism under Multi-round Adversarial Prompting. CPS Digital Library - Series of Conferences. 1, (Jul. 2026), 33–37.