Research

Research Interests

My research sits at the intersection of machine learning and hardware security, with three primary thrusts:

1. Secure & Privacy-Preserving ML — Protecting ML models from extraction, probing, and misuse
2. Side-Channel Analysis — Detecting adversarial attacks and hardware vulnerabilities via EM/power signals
3. ML-Hardware Co-Design — Building trustworthy computing foundations through joint design

These efforts advance two critical domains: Responsible AI (ensuring AI systems are secure and ethical) and Hardware Security (building trustworthy computing foundations). Now I am actively working on security of agentic AI, including vibe coding, openclaw, etc.