I am a PhD student in the mathematics department jointly supervised by Nick Jones and Tom Ouldridge. In general, I am interested in the physical costs of chemical inference. 

Recently, I have been thinking about minimal chemical devices for exploiting structured environments. I have devised machines based on a biochemical implementation of the Szilard engine that can extract work from correlated sequences of two-state molecules in ‘Biochemical Szilard engines for memory-limited inference’. We have also used these biochemical Szilard engines to elucidate the fundamental thermodynamics of information in ‘The power of being explicit: demystifying work, heat, and free energy in the physics of computation’. Now I am working on more realistic autonomous machines with a single chemical buffer that can extract work from a fluctuating environment. 

Previously, I have published a paper, ‘What we learn from the learning rate’, in which we explore the interpretation of the ‘learning rate’, which is a quantity defined on bipartite Markov chains. This quantity has previously been used as measure of sensing between an upstream and downstream system, however we find that this is not necessarily a good interpretation.

Before my PhD, my MPhys project was on the theory of bouncing water drop droplets on superhydrophobic surfaces. It was supervised by Julia Yeomans at the University of Oxford.