Bio-inspired plastic spiking neural networks engaged in learning and sleep cycles

https://agenda.infn.it/event/37812/

Seminari INFN, Monday 11th November 2023, Aula Conversi, Dip. FIsica Sapienza University Roma Italy

Abstract

Sleep passed the evolutionary siege in all studied animal species, notwithstanding its apparent unproductivity (lower reactivity to external dangers, no feeding, no mating). In humans, the time spent in sleep is higher in younger individuals, precisely when learning is faster. Another element to be considered is that, thanks to an evolutionary history that spanned hundreds of millions of years and selected among countless individuals, the inter-areal and local connectome captures the priors necessary to optimize the flow and combination of internal hypotheses and sensorial evidence. 

At the cellular level, optimal combination of contextual information and local computation is provided by the apical amplification principle, active during wakefulness. Deep-sleep (NREM) and REM sleep are characterized in mammals by pyramidal neurons changing to a different management of apical signals, namely apical-isolation and apical-drive.
The cognitive and energetic functions of sleep and its relations with awake performance have beeninvestigated by INFN in spiking models, engaged in learning and sleep cycles, that will be presented in this seminar. Also, preliminar information about a next generation of neural models supporting apical mechanisms will be presented.