Ganqu Cui

large language model (6) backdoor attack (3) preference optimization (3) pre-trained language model (3) adversarial defense (3) preference learning (2) few-shot learning (2) reward model (2) reinforcement learning from human feedback (2) multimodal large language model (2) reinforcement learning (2) instruction tuning (2) natural language processing (2) text classification (2) adversarial attack (2) hallucination reduction (2) benchmark evaluation (1) model security (1) zero-shot learning (1) contrastive learning (1)