mean field theory of graph neural networks

A theoretical approach to understanding the function and limitations of Graph Neural Networks (GNNs) based on mean field theory.