معرفی
David Duvenaud is an Associate Professor at the University of Toronto, affiliated with the Faculty of Arts and Science. He is a founding member of the Vector Institute and holds a Schwartz Reisman Chair in Technology and Society. His research focuses on AGI governance, existential risk mitigation, and probabilistic machine learning. Previously, he conducted postdoctoral research at Harvard and earned his Ph.D. from the University of Cambridge under advisors Carl Rasmussen and Zoubin Ghahramani.
- Education: Ph.D. (University of Cambridge), M.Sc. (University of British Columbia)
- Research Interests: Neural ODEs, AI ethics, probabilistic modeling, and scalable generative systems.
His work bridges foundational machine learning theory with real-world applications, including chemical design and alignment science. Key contributions include LLM Processes, sabotage evaluations for frontier models, and neural ODE frameworks. He teaches advanced machine learning courses (e.g., STA414, CSC2547) and advises students like James Requeima and Raymond Douglas. His awards include a Sloan Fellowship and distinguished paper recognitions.
Current projects address systemic risks posed by incremental AI progress, emphasizing long-term societal impacts and governance mechanisms.
