English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Spiking Neural Networks (SNNs) are being explored for their potential energyefficiency benefits due to sparse, event-driven computation. Non-spikingartificial neural networks are typically trained with stochastic gradientdescent using backpropagation. The calculation of true gradients forbackpropagation in spiking neural networks is impeded by the non-differentiablefiring events of spiking neurons. On the other hand, using approximategradients is effective, but computationally expensive over many time steps. Onecommon technique, then, for training a spiking neural network is to train atopologically-equivalent non-spiking network, and then convert it to an spikingnetwork, replacing real-valued inputs with proportionally rate-encoded Poissonspike trains. Converted SNNs function sufficiently well because the meanpre-firing membrane potential of a spiking neuron is proportional to the dotproduct of the input rate vector and the neuron weight vector, similar to thefunctionality of a non-spiking network. However, this conversion only considersthe mean and not the temporal variance of the membrane potential. As thestandard deviation of the pre-firing membrane potential is proportional to theL4-norm of the neuron weight vector, we propose a weight adjustment based onthe L4-norm during the conversion process in order to improve classificationaccuracy of the converted network.

L4-Norm Weight Adjustments for Converted Spiking Neural Networks