I found various resources that explain how encoders and SPs work from a high-level perspective. But I would like to know how the encoders and SPs actually do this from algorithmic perspective. I would like to implement my own encoder and SP for my usecase. So, can you please help me with describing the steps in the algorithm for encoder and SP separately or point me towards a resource if it already exists?
I’d highly recommend BAMI.
You may also read the source code of some encoders. Like these:
The rule of thumb is that encoders converts whatever into SDR, and values close in the input space is also close in the SDR.