I have some problem with reproducing results of Non-hebbian learning algorithm presented in the paper, from the https://www.sciencedirect.com/science/article/pii/S15684946040010
