Yueqi Xie

Privacy (1)

large language model (2) federated learning (2) adversarial learning (1) image generation (1) distributed learning (1) privacy preservation (1) harmful content (1) non-iid datum (1) safety alignment (1) mutual information (1) adversarial attack (1) hidden state (1) adversarial defense (1) client heterogeneity (1) multimodal large language model (1) model aggregation (1) retrieval-augmented generation (1) model poisoning (1) harmful content detection (1) adversarial robustness (1)