Agent-based Graph Neural Networks

Karolis Martinkus, Pál András Papp, Benedikt Schesch, Roger Wattenhofer

Keywords: Graph Neural Networks, GNN, Graph Classification, Expressive Graph Neural Networks, Sublinear algorithms

TL;DR: We present a new agent-based sublinear and expressive GNN architecture for graph-level tasks.

Abstract: We present a novel graph neural network we call AgentNet, which is designed specifically for graph-level tasks. AgentNet is inspired by sublinear algorithms, featuring a computational complexity that is independent of the graph size. The architecture of AgentNet differs fundamentally from the architectures of traditional graph neural networks. In AgentNet, some trained \textit{neural agents} intelligently walk the graph, and then collectively decide on the output. We provide an extensive theoretical analysis of AgentNet: We show that the agents can learn to systematically explore their neighborhood, and that AgentNet can distinguish some structures that are even indistinguishable by 3-WL. Moreover, AgentNet is able to separate any two graphs which are sufficiently different in terms of subgraphs. We confirm these theoretical results with synthetic experiments on hard-to-distinguish graphs and real-world graph classification tasks. In both cases, we compare favorably not only to standard GNNs but also to computationally more expensive GNN extensions.

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/agent-based-graph-neural-networks/code)