Reproduce “Multi-Depth Temporal Fusion for Feedforward, Locally Trained 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
Aidin Attar, Eleonora Cicciarella, Michele Rossi
The authors introduce MDTF, a routing method for spiking neural networks that preserves shallow time-to-first-spike evidence, adds sparse intermediate events, and admits deep events only when timing agrees with intermediate evidence. Shared-backbone tests favor preserving features over serial replacement; sparse refinements help more on tasks than MNIST. Fashion-MNIST accuracy is 86.3% versus 68.2% for a baseline, a whole-system comparison.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.