Spikes are cool again! What's next?

Institute for Neuroinformatics, University of Zurich
2025-03-06
 
Goodman DFM
 

Abstract

Invited talk. I discussed my history with spiking neural networks, from writing the Brian simulator, sound localisation, surrogate gradient descent and SNUFA. I talked about why spikes, particularly in a neuromorphic context. My argument included nonlinearity and temporal dynamics (including coincidence detection, multimodal processing in time, and delay networks), heterogeneity, sparsity and modularity.

Slides

Links

Organisation for neuroscientists interested in spiking neural networks.

Brian

A Python simulator for spiking neural networks.

Categories