Many of the brain's most intriguing cognitive capacities are carried out by dynamic neural circuits which integrate information over timescales ranging from milliseconds to hours and days. One example is perceptual segmentation: our perceived experience of the world arrives in discrete 'chunks' though is derived from a continuous and never ending stream of sensory data. Chunking of complex motor tasks into combinations of pre-learned simpler units is another example. Tom studies the neural circuitry underlying these phenomena by, among other things, modelling them with recurrent neural networks.
Tom completed a BA/MSc in physics at the University of Cambridge before moving to Harvard for a year where he jumped ship to the greener pastures of computational neuroscience and machine learning. He joined the SWC in 2020 and now works primarily in the Akrami and Clopath labs.