This repo provides a library for investigating coding theory applications using techniques from spiking neural networks (SNNs). This is based on the following features of SNNs:
- SNNs provide a differentiable framework for working with discrete data and can be seen as a bridge between the discrete and continuous domain.
- They are adapted better than conventional neural networks for coding theory applications since they work directly with the discrete data.
- SNNs use a sparse collection of spikes as a communication tool between their neurons and have a dramatically smaller power consumption. This make SNNs better suited for coding-for-communication applications.