Modeling Transitivity and Cyclicity in Directed Graphs via Binary Code Box Embeddings

Modeling directed graphs with differentiable representations is a fundamental requirement for performing machine learning on graph-structured data.