01/06/2024 11:30 am
Topic starter
Notifications
Clear all
0
I have a class of brain devkit DYNAPCNN, and I want to implement a Continous Attractor Neural Network(CANN), which consists of 180 LIF neurons, each connected to other neurons, and the weights for defining synaptic connections follow a Gaussian distribution. However, I cannot find any way to implement it on this development board. I hope everyone can provide suggestions. Thanks a lot!!
1 Answer
0
25/10/2024 9:47 am
Hi @wcx12306, LIF neurons and recurrent networks are not supported on DYNAP-CNN or Speck. Only convolutional architectures are supported on these cores.
Our Xylo processor devices support LIF and recurrent connectivity.
