Reproduce “Supervised Learning with First-to-Spike Decoding in Multilayer Spiking Neural Networks” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Brian Gardner, André Grüning
The authors present a feedforward spiking neural network that classifies handwritten MNIST digits by the first output neuron to spike. Scanline encoding turns image paths into timed spike sequences, reducing input neurons; stochastic hidden neurons may fire multiple times, while deterministic outputs decide. On 10,000 test images, delayed scanline networks reached 87 ± 2% accuracy, versus 76 ± 6% without delays, in the tested setup.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.