Towards Graph Foundation Models for Dynamics in Complex Networked Systems: Lessons from Super-Spreader Identification in Multilayer Networks

2026-06-06

AI Summary

This work addresses the limited generalizability of existing network dynamics models, which are predominantly transductive and fail to generalize to unseen graph structures. The study introduces ts-net, the first inductive model for modeling dynamics on complex multilayer networks grounded in the principles of graph foundation models, and establishes four key design principles. Trained solely on synthetic multilayer networks, ts-net achieves zero-shot generalization across graphs, scales, and numbers of layers, outperforming classical heuristic methods and transductive baselines on three of four evaluation metrics. The model demonstrates practical efficacy in identifying superspreaders in real-world multilayer networks. This research validates the feasibility of zero-shot cross-network dynamic modeling and outlines five critical challenges for future investigation.

Abstract

Network dynamics - including spreading, influence maximisation, and epidemic modelling - remain largely confined to the transductive paradigm, where models are trained on a single network and cannot be reused on unseen graphs without retraining. We argue that inductive cross-network generalisation is a necessary prerequisite for Graph Foundation Models (GFMs) in this domain and propose four design properties towards this goal. As a proof of concept, ts-net (TopSpreadersNetwork), trained solely on synthetic multilayer networks (MLNs), demonstrates zero-shot generalisation to real-world MLNs of varying size and layer count, outperforming classical heuristics and transductive baselines on three of four metrics. Based on ts-net's performance, we further outline five open challenges towards building GFMs for network dynamics: scale, many-layer generalisation, self-supervised pretraining, cross-task transfer, and node-attribute integration.

Problem

• network dynamics
• inductive generalization
• graph foundation models
• multilayer networks
• zero-shot transfer

Innovation

• Graph Foundation Models
• inductive generalization
• multilayer networks
• zero-shot transfer
• network dynamics