معرفی
Hongyang Zhang is a tenure-track Assistant Professor at the University of Waterloo's David R. Cheriton School of Computer Science and Faculty of Mathematics, affiliated with the Vector Institute for AI. His research bridges theoretical and applied aspects of machine learning, including world modeling, AI inference acceleration, and security. He leads the SafeAI Lab and holds memberships in the AI Institute and Cybersecurity and Privacy Institute.
Dr. Zhang completed his Ph.D. at Carnegie Mellon University's Machine Learning Department and conducted postdoctoral research at Toyota Technological Institute at Chicago. His educational background includes a degree from Peking University.
Research interests focus on:
- Developing generative world models for robotics and autonomous systems
- Creating efficient algorithms for accelerating LLM inference
- Building System-2 LLMs for enhanced reasoning
- Advancing AI security against adversarial threats
Publications demonstrate strong focus on efficient AI systems and security, with recent work on EAGLE series acceleration techniques and The Matrix world generation framework dominating top conferences like ICML, NeurIPS, and CVPR.
Awards and honors:
- 1st place in CVPR 2021 Security AI Challenger
- Multiple NeurIPS challenge wins in adversarial vision
- AAAI New Faculty Highlights awardee
- IEEE Senior Member
Leads SafeAI Lab focusing on trustworthy AI systems and regularly serves as area chair for top machine learning conferences including NeurIPS, ICML, and ICLR.
