saunak1994 / snnformnist Goto Github PK

I am working on Spiking Neural Nets and have gone through P Diehl's original paper and code and had some questions on his implementation as well. Since Brian2 and Spiking Neural Nets have less members and support forums, I am trying to get a better understanding of your implementation. Do you have the learned weights plot and does it resemble the original digits? In my experiments, I found overlaying with the PoissonGroup makes the SNN to start firing spikes even for the noise considering to wrongly as part of the pattern. Could you shed some clarity since yours is the only implementation which seems to adhere to the original papers Synapse connectivity namely the Excitatory to Inhibitory and vice versa. In my repo where I have experimented, I have posted your link as a reference in my code samples.