NeuroTrain: Surveying Local Learning Rules for Spiking Neural Networks with an Open Benchmarking Framework

• Post by: admin
• May 18, 2026
• No Comment

Training spiking neural networks is hard — and fairly comparing training algorithms under consistent conditions is even harder. 🧠

Our paper “NeuroTrain: Surveying Local Learning Rules for Spiking Neural Networks with an Open Benchmarking Framework” addresses these challenges and is now available on arXiv!

We surveyed the landscape of SNN training algorithms, organizing them along a taxonomy that puts locality at the center, and built NeuroTrain: an open, snnTorch-based benchmarking framework designed to evaluate training rules under shared, reproducible conditions across models and datasets.

We see NeuroTrain as a living project, we’re committed to growing it over time with more trainers, models, and datasets, and we’d love for the community to be part of that.

📄 Paper: arxiv.org/abs/2605.15058
💻 Code: https://github.com/smilies-polito/neurotrain

A huge thank you to my coauthors Filippo Marostica, Roberta Bardini, Alessandro Savino, and Stefano Di Carlo, and to the whole SMILIES Polito Research Group at Politecnico di Torino for the support throughout this work.

Shared by: Alessio Caviglia

View original post on LinkedIn