STREAM (ChemBio): A Standard for Transparently Reporting Evaluations in AI Model Reports
Marginal Risk Relative to What? Distinguishing Baselines in AI Risk Management
Frontier AI Safety Frameworks: three open research areas
A Grading Rubric for AI Safety Frameworks
Coordinated Pausing: An Evaluations-Based Coordination Scheme for Frontier AI Developers
Research Experience
Member of Policy Staff at METR; Affiliate at the Centre for the Governance of AI and former Research Scholar; Fellowships and summer programs at the Center on Longterm Risk and the Institute for Law and AI; Co-founder of Trajectory Labs
Education
Degree in political science and philosophy from the University of Toronto
Background
An AI governance researcher based in London, UK. Research interests include strengthening risk management policies that address national security risks from advanced AI.
Miscellany
Pledged to donate at least 10% of annual income to charities that most effectively improve the lives of others