A Categorical Framework For Deep Learning Models

“This paper introduces a categorical framework for deep learning models that formalizes broadcasting through the novel axis-stride and array-broadcasted categories.”