Neural vectors or grandmother cells : are mental representations localized or distributed ?

In the space of a few decades, the recording of neurons in the animal and human brain has been revolutionized by the emergence of massively parallel electrophysiology and optical imaging techniques, which capture the responses of hundreds or even thousands of neurons simultaneously. The lecture will examine how we have moved from a notion of coding by single neurons (the famous " neurons grandma ") to a vector approach to neural code (a code distributed over a vast population of neurons). In the macaque monkey, with the pioneering work of Apostolos Georgopoulos in the 1980s, but also in very recent work in the Drosophila fly and the bat, vector coding by " phasors " explains how animals can make geometric mental calculations, for example to determine the direction of a target.