I'm a postdoc at Mila supervised by Yoshua Bengio. I'm currently the scientific lead of the first International Scientific Report on the Safety of Advanced AI.
My interests in machine learning include LLM honesty, health applications, risk management, and data selection for large-scale deep learning. Across these areas, my publications as leading author have been covered in TV and newspapers like The Guardian, Time, etc, and have been discussed by ministers and in a national bill.
Before joining Mila, I was a PhD student at the University of Oxford under Yarin Gal, and I worked on learning human preferences and game-theoretical machine learning with David Duvenaud and Roger Grosse at Toronto’s Vector Institute and UC Berkeley, and with the Governance of AI group at Oxford. I studied machine learning (UCL), maths (Amsterdam) and Future Planet Studies (Amsterdam).