Maya-Nidra P1: A Neuromorphic Spiking Neural Network for Sleep-Phase Memory Consolidation with Bhaya-Gated Quiescence

We present Maya-Nidra, a dual-cortex neuromorphic spiking neural network (SNN) architecture thatmodels biological sleep-phase memory consolidation through event-driven sharp-wave ripple (cSPW-R)replay and Bhaya-gated quiescence. Two cortical subsystems are introduced: NexusSpikeNet (114classes, 14,762 samples, Gate 1) achieving 39.71% accuracy on N-Caltech101 and CIFAR10-DVS; andMayaEEGNet (5 AASM sleep stages, Gate 2) trained on Sleep-EDF Expanded, achieving 64.9% overallwith N2=92.5% and N3=96.5% — exceeding PicoSleepNet (IEEE 2026) on the two most clinicallysignificant stages by +9.5% and +11.5% respectively.Parametric ablation across 36 configurations confirms the Bhaya Quiescence Law: 0.32%…