Answers are grounded in this paper and cite their sources.
The Axelrod model is an agent based model composed of N F-dimensional vectors with each value taken from a range of 0 to q−1. Agents are allowed to interact with their topological neighbors on a given network based on homophily-driven dynamical rules. This work studies the evolution of link densities of a toy model (F = 2) on a complete graph and a regular lattice. I wrote mean-field master equations and find that a pairwise link dynamics is inherently dependent on higher-order triadic correlations, and the difference in topologies leads to a difference in local geometry and triad configurations. On a complete graph, the triad forms a interconnected triangle, perfectly closing the master equations, and all higher-order correlations canceled out due to symmetry at F = 2. However, on a lattice, the triad forms a L-shape with a missing third link, forcing a closure approximation under mean-field assumptions, which introduces non-linear terms in the equations. Simulation results show a perfect alignment with theoretical predictions on complete graphs but a significant discrepancy on lattices. Further visualizations highlight that the dynamics on lattices are governed by a long-term coarsening process of cultural domains due a path length L ∝ N 1/2 , whereas on complete graphs information spread globally due to L = 1, causing it to behave nearly as a mean-field system. (This version fixes the typos in previous ones. I'm a highschool student, and I'm opened to any criticisms and pieces of advice regarding this work) Contact: Nianzheng Mao (26nmao@cacegypt.org; maon@wustl.edu)
No takes yet. Share an insight, caveat, or question.
Nianzheng(Nate) Mao (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: