Implications of sparsity and high triangle density for graph representation learning

Open in new window