Spin Glasses on Thin Graphs

TL;DR

This paper investigates spin glass and ferromagnetic transitions on thin random graphs, using saddle point calculations and simulations to compare with mean-field and Bethe lattice results, across various models.

Contribution

It extends saddle point analysis to multiple models on thin graphs and compares analytical results with numerical simulations, highlighting similarities with infinite-range spin glasses.

Findings

Transition temperatures agree with Bethe lattice predictions.

Spin glass transition is independent of bond fraction in Ising models.

Overlap distributions show characteristic spin glass behavior.

Abstract

In a recent paper we found strong evidence from simulations that the Isingantiferromagnet on ``thin'' random graphs - Feynman diagrams - displayed amean-field spin glass transition. The intrinsic interest of considering such random graphs is that they give mean field results without long range interactions or the drawbacks, arising from boundary problems, of the Bethe lattice. In this paper we reprise the saddle point calculations for the Ising and Potts ferromagnet, antiferromagnet and spin glass on Feynman diagrams. We use standard results from bifurcation theory that enable us to treat an arbitrary number of replicas and any quenched bond distribution. We note the agreement between the ferromagnetic and spin glass transition temperatures thus calculated and those derived by analogy with the Bethe lattice, or in previous replica calculations. We then investigate numerically spin glasses with a plus or minus J bond distribution for the Ising and Q=3,4,10,50 state Potts models, paying particular attention to the independence of the spin glass transition from the fraction of positive and negative bonds in the Ising case and the qualitative form of the overlap distribution in all the models. The parallels with infinite range spin glass models in both the analytical calculations and simulations are pointed out.