Implementation of the paper The Kanerva Machine: A Generative Distributed Memory in tensorflow and keras.
The Kanerva Machine can be viewed as an extension of a Variational Autoencoder with a memory. https://arxiv.org/abs/1804.01756
This model is inspired by Kanerva's Sparse Distributed Memory http://i.stanford.edu/pub/cstr/reports/csl/tr/89/400/CSL-TR-89-400.pdf
The code is not ready for testing yet, what's left to do before testing : -create the MLP layer -implement the loss function -implement derivation for online update rule