AI RESEARCH
Training Non-Differentiable Networks via Optimal Transport
arXiv CS.LG
•
ArXi:2605.01928v1 Announce Type: new Neural networks increasingly embed non-differentiable components (spiking neurons, quantized layers, discrete routing, blackbox simulators, etc.) where backpropagation is inapplicable and surrogate gradients PolyStep trains genuinely non-differentiable models where existing gradient-free methods collapse to near-random accuracy. On hard-LIF spiking networks we reach 93.4% test accuracy, outperforming all gradient-free baselines by over 60~pp and closing to within 4.4~pp of a surrogate-gradient Adam ceiling.