ABSTRACTWe here propose a neural network model to explore how neural oscillations might regulate the replay of memory traces. We simulate the encoding and retrieval of a series of events, where temporal sequences are uniquely identifiable by analysing population activity, as several recent EEG/MEG studies have shown. Our model comprises three parts, each considering distinct hypotheses. A cortical region actively represents sequences through the disruption of an intrinsically generated alpha rhythm, where a desynchronisation marks information-rich operations as the literature predicts. A binding region converts each event into a discrete index, enabling repetitions through a sparse encoding of events. We also instantiate a temporal region, where an oscillatory “ticking-clock” made up of hierarchical synfire chains discretely indexes a moment in time. By encoding the absolute timing between events, we show how one can use cortical desynchronisations to dynamically detect unique temporal signatures as they are replayed in the brain.
Learning Long Temporal Sequences in Spiking Networks by Multiplexing Neural Oscillations
