Thomas Akam

I am interested in how brains generate flexible, adaptive, behaviour in the complex and richly structured environments we live in.  This capacity for flexible behaviour relies on models of the world, or cognitive maps, learned by our brains. These models enable us to plan action sequences to achieve our goals, generalise prior learning to new situations, and infer hidden structure in our environments.  My research aims to understand how these models are learned, represented by brain activity, and used to guide behaviour.

A major focus of my work has been developing new laboratory behavioural tasks which are rich enough to recruit these cognitive mechanisms, while generating large datasets ideal for quantifying brain activity and computational modelling.  I combine these behaviours with modern tools for recording and manipulating brain activity, including silicon probe electrophysiology, fiber photometry, and optogenetics.  I work primarily on frontal cortex and hippocampus – regions thought to be critical for learning and using internal models, and on the dopamine system – which plays a key role in learning the value of situations and actions.   My work is strongly influenced by theoretical neuroscience and machine learning, and I use computational modelling to understand behaviour and brain activity. 

I also like building things, and created two open source hardware/software tools for behavioural neuroscience; pyControl and pyPhotometry, that are used by labs worldwide.