Boundary Degree as a Node-level Feature for Epidemic Scenario Identification in Agent-based Cascade Simulations

This document is an arXiv feed containing multiple papers across epidemic modeling, graph/network methods, multi-agent AI systems, and governance analysis. Notable security-relevant items: (1) “Blackknife” describes hard-label, query-limited, structure-limited black-box attacks against heterogeneous graph neural networks (HGNNs), a practical threat to deployed HGNN-based services (fraud detection, recommendation, malware analysis). (2) Several network/graph papers (community-detection instability, edge-deletion mitigation for SIS epidemics, phase-boundary mapping) present methods that affect: