Higher-Order Graphon Neural Networks: Approximation and Cut Distance
This work addresses the challenge of making powerful higher-order GNNs transferable across graph sizes, which is crucial for applications in machine learning on graphs, though it builds incrementally on prior graphon-based methods.
The paper tackles the problem of ensuring size transferability for higher-order graph neural networks (GNNs) by extending them to graphon limits, showing that Invariant Graphon Networks (IWNs) of order k are at least as powerful as the k-WL test and achieve universal approximation in L^p distances.
Graph limit models, like graphons for limits of dense graphs, have recently been used to study size transferability of graph neural networks (GNNs). While most literature focuses on message passing GNNs (MPNNs), in this work we attend to the more powerful higher-order GNNs. First, we extend the $k$-WL test for graphons (Böker, 2023) to the graphon-signal space and introduce signal-weighted homomorphism densities as a key tool. As an exemplary focus, we generalize Invariant Graph Networks (IGNs) to graphons, proposing Invariant Graphon Networks (IWNs) defined via a subset of the IGN basis corresponding to bounded linear operators. Even with this restricted basis, we show that IWNs of order $k$ are at least as powerful as the $k$-WL test, and we establish universal approximation results for graphon-signals in $L^p$ distances. This significantly extends the prior work of Cai & Wang (2022), showing that IWNs--a subset of their IGN-small--retain effectively the same expressivity as the full IGN basis in the limit. In contrast to their approach, our blueprint of IWNs also aligns better with the geometry of graphon space, for example facilitating comparability to MPNNs. We highlight that, while typical higher-order GNNs are discontinuous w.r.t. cut distance--which causes their lack of convergence and is inherently tied to the definition of $k$-WL--transferability remains achievable.