English

Agent-based Graph Neural Networks

Machine Learning 2023-03-01 v2 Artificial Intelligence Machine Learning

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 2-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.

Keywords

Cite

@article{arxiv.2206.11010,
  title  = {Agent-based Graph Neural Networks},
  author = {Karolis Martinkus and Pál András Papp and Benedikt Schesch and Roger Wattenhofer},
  journal= {arXiv preprint arXiv:2206.11010},
  year   = {2023}
}

Comments

32 pages, 6 figures, ICLR 2023

R2 v1 2026-06-24T11:59:57.134Z