Recently Microsoft published a Paper on a new artificial neural network architecture whose weights are binary or ternary values (1-bit) and I would like to know how similar it is to the binary values of the weights in HTM
1 Like