Presenting Neural Networks via Coherent Functors

Matthew Pugh, Jo Grundy, Corina Cirstea +1

This paper develops a methodology for representing machine learning models as models of formal theories, grounded in the perspective that machine learning models are a form of data…