Courses and awards

AI Safety Fundamentals: AI Alignment (LEAF / BlueDot Impact, Sep-Dec 2024)

Weekly readings and discussion groups working through the alignment curriculum. I spent most of my time on interpretability and scalable oversight, and came up with some project ideas around comparing circuits in small open-source models.

Dilemmas and Dangers in AI (LEAF, Jun-Aug 2024)

A broader course on AI risk and governance, with some hands-on work like forecasting and stress-testing models.

CS50x (Harvard, 2023)

Harvard's intro to computer science, done online: ten problem sets, nine labs and a final project. Certificate

Competitions

UK Cyber Leaders Challenge 2026 (Regional Round)

Four of us worked through a simulated national cyber emergency as a team, then briefed a panel of expert judges on the technical and strategic risks.

BPhO Computational Challenge 2024 (Gold Award)

I modelled projectile motion in code. Challenge page

Apart Research AI Capabilities and Risks Demo-Jam (Aug 2024)

A hackathon with two others from my LEAF cohort. We built a tool that compresses arXiv abstracts and uses another model to reconstruct them. Hackathon page