Research

My research aims to understand perception by combining Artificial Neural Networks (ANNs) with neuroimaging and behavioural data. In a recent collaborative review , we gave a detailed account of this approach, which we call neuroconnectionism. ANNs offer much more than black box predictors: they provide an innovative and progressive research programme for cognitive neuroscience. This is possible because ANNs live in a goldilocks zone of computational complexity: they are abstract enough to be computationally tractable and applicable to naturalistic cognitive tasks, while also remaining sufficiently close to biology to enable hypothesis-driven research into the biological basis of human cognition.