A deep learning library for spiking neural networks which is based on PyTorch, focuses on fast training and supports inference on neuromorphic hardware.
The problem is in this line, which registers a spiking hook to any StatefulLayer, without checking whether it spikes. This, for instance, may include an ExpLeak layer, which can result in negative "firing rates" and other unexpected data being recorded.
The problem is in this line, which registers a spiking hook to any
StatefulLayer, without checking whether it spikes. This, for instance, may include anExpLeaklayer, which can result in negative "firing rates" and other unexpected data being recorded.