I'm a first year DPhil student, co-supervised by Yee Whye Teh and Tom Rainforth. Before this I completed an integrated masters in Information and Computer Engineering at Cambridge, with my final year project on improving neural transducer-based speech recognition system algorithms.

I am currently working on improving understanding of when and why training fails in reinforcement learning. Broadly I'm interested in probabilistic machine learning, meta reinforcement learning, robust causal representation learning, and in general understanding what models are learning, how they learn, and why.