RIVAL：信息价值驱动的大语言模型多轮越狱攻击方法
彭奕杰, 李欣, 李明锋, 吴昊博, 刘明帅, 陈东先, 赵晓凡
RIVAL: An Information-Value-Driven Multi-Turn Jailbreak Attack Method for Large Language Models
Peng Yijie, Li Xin, Li Mingfeng, Wu Haobo, Liu Mingshuai, Chen Dongxian, Zhao Xiaofan
网络空间安全科学学报
.
2026, (): 0
.
DOI: 10.20172/j.issn.2097-3136.260737
