security attack

large language model (4) adversarial attack (3) llm security (2) prompt injection (2) text embedding inversion (1) embedding privacy (1) cross-lingual attack (1) attack technique (1) model retraining (1) copyright infringement (1)