Yuanpu Cao

large language model (4) adversarial attack (3) ai safety (2) model alignment (2) multimodal large language model (2) policy learning (1) preference optimization (1) model security (1) harmful content (1) jailbreak attack (1) safety alignment (1) chain-of-thought reasoning (1) backdoor attack (1) model safety (1) hallucination mitigation (1) language model alignment (1) jailbreaking attack (1) safety evaluation (1) knowledge editing (1) adversarial robustness (1)