Interpretable and Compositional Relation Learning by Joint Training with an Autoencoder

Open in new window