Neural net interpretability with Bayesian interpolants - View it on GitHub