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.

翻译：我们提出了一种名为AgentNet的新型图神经网络，专为图级任务设计。AgentNet受亚线性算法启发，其计算复杂度与图规模无关。AgentNet的架构从根本上区别于传统图神经网络。在AgentNet中，经过训练的神经智能体智能地遍历图，并共同决定输出。我们对AgentNet进行了广泛的理论分析：证明了智能体可以学习系统性地探索其邻域，且AgentNet能够区分某些甚至2-WL无法区分的结构。此外，AgentNet可以分离任何在子图层面存在足够差异的两个图。我们通过在难以区分的图上的合成实验和真实世界图分类任务中验证了这些理论结果。在这两种情况下，我们不仅优于标准GNN，也优于计算成本更高的GNN扩展模型。