Yuanhe Zhang

backdoor attack (2) resource consumption (2) adversarial machine learning (1) model alignment (1) non-parametric statistics (1) multivariate prediction (1) adversarial defense (1) quasi-bayesian method (1) vine copula (1) black-box setting (1) predictive distribution (1) llm agent (1) security vulnerability (1) audio large language model (1) large language model (1) energy score (1) dynamic encryption (1) safety audit (1) multi-backdoor implantation (1) denial of service (1)