TheoreticallyProvableSpikingNeuralNetworks