Most examinations of neural networks’ learned latent spaces typically employ dimensionality reduction techniques such as t-distributed stochastic neighbor embedding (t-SNE) or uniform manifold ...