Kaike Zhang

large language model (2) neural network security (1) adversarial training (1) instruction following (1) poisoning attack (1) evaluation benchmark (1) recommender system (1) pretrained language model (1) backdoor mitigation (1) backdoor attack detection (1) coarse-to-fine generation (1) textual backdoor attack (1) poison detection (1) reinforcement learning with verifiable reward (1) lexical instruction (1) programmatic evaluation (1) adversarial collaborative filtering (1) rubric generation (1) model security (1) rejection sampling fine-tuning (1)