Fig. 5: Oscillation-based integrator model exhibits more robustness to changes in the recurrent feedback weight than a traditional non-oscillation-based model.
From: An oscillatory mechanism for multi-level storage in short-term memory
a Responses of oscillation-based model to a sequence of positive and negative input pulses. Red and yellow traces show conditions in which the recurrent feedback strength has been detuned by ±5%, respectively. The activity levels remain persistent following detuning. b Steady-state firing rates as a function of synaptic feedback activation that is held at steady values; mistuning the autapse weight value by ±5% has no effect on the existence and ___location of the stable fixed points (intersections of black lines and horizontal stairs). c Responses of a traditional, approximately linear, conductance-based model of persistent neural activity (adapted from model of ref. 36). Detuning the recurrent feedback strength by 5% (orange and red traces) causes spiking activity to decay to 0 (orange, decreased feedback strength) or run off to high rates (red, increased feedback strength). d Small weight changes cause systematic loss of fixed points in the traditional model.
