About
Zewei Shi is a PhD Candidate at the University of Melbourne and CSIRO Data61. His thesis, Multimodal LLMs for Digital Trust, explores how LLMs can support digital trust among end users, developers, researchers, and regulators. His research focuses on deceptive patterns, machine learning unlearning, and adversarial attacks and defenses in time-series data within AI safety and security. Zewei's work are fully supported by CSIRO, Google and OpenAI.
AcademiaEarly career / PhD (0-5 years)Security/cybersecurityComputer science