About
I'm a computer scientist and AI safety researcher focused on ensuring beneficial outcomes from advancing AI capabilities. After spending four years in Fintech post-graduation, I pursued an MPhil in Machine Learning at the University of Cambridge, where my dissertation explored the inductive biases of shallow neural networks. My research interests lie at the intersection of machine learning and technical AI safety, with a focus on evaluation methodologies and neural network behavior. I investigate how networks' inductive biases affect alignment challenges and develop interpretability methods for detecting potentially harmful systems. My recent work has centered on evaluation frameworks and metrics for assessing AI systems' capabilities and safety properties.
Badges
