ANDREA BENUCCI

I research the neural mechanisms involved in visual processing adopting a research framework capable of linking neural architectures to the underlying computations. My preferred approach involves integrating experimental methods for optically dissecting neuronal circuits (optical imaging and patterned optogenetics) with large-scale dynamic network models based on artificial neural networks (aNNs). Computation within artificial neural networks arises from connectivity principles among units, inspired by how biological networks perform computations. Therefore, aNNs serve as an effective ‘mechanistic’ modelling framework for integrating computational, algorithmic, and implementation levels of analysis. I employ this unified computational and experimental approach to uncover the principles of neural computation that underlie visual perception.