Kai He

FidLIF: Parallel Training of High-Fidelity Spiking Neurons with Exact BPTT

FidLIF is presented, a novel framework that reconciles the incremental update and hard-reset behavior of standard spiking LIF neurons with parallel forward propagation and BPTT-equivalent backward propagation and enables high-fidelity spiking dynamics to benefit from parallel hardware without compromising incremental i...