Possible LIF Refractory Period Bug

UCI-CARL / CARLsim6

CARLsim is an efficient, easy-to-use, GPU-accelerated software framework for simulating large-scale spiking neural network (SNN) models with a high degree of biological detail.

MIT License

41 stars 16 forks source link

Possible LIF Refractory Period Bug #26

Open nmsutton opened 1 year ago

nmsutton commented 1 year ago

Code was added to snn_gpu_module.cu's kernel_findFiring() involving INT_MAX to have the refractory period (RP) correctly processed with the Izhikevich model in the imfix branch. It is suspected that this code may also be needed for correct Leaky-Integrate-and-Fire neuron model (LIF) RP processing with different RP times. Without the code, LIF RP may have a bug.

larsnm commented 1 year ago

This seems to be an issue for LIF specific to the branch. The unit test for LIF of the master branch did not show this issue. It might be addressed when merging the PR for the imfix.

nmsutton commented 1 year ago

Thanks @larsnm for checking this. @jkopsick has told me that he thinks somehow this does not appear to be an issue with LIF. I requested that he describe his evidence for that in this issue post and I will close this issue. This will help document this if anyone has the same idea in the future. He has agreed to do that, but is just busy with other work at the current time, and will do it later.