Abstract:We consider the graphical representations of the Ising model on tree-like graphs. We construct a class of graphs on which the loop $\mathrm{O}(1)$ model and then single random current exhibit a non-unique phase transition with respect to the inverse temperature, highlighting the non-monotonicity of both models. It follows from the construction that there exist infinite graphs $\mathbb{G}\subseteq \mathbb{G}'$ such that the uniform even subgraph of $\mathbb{G}'$ percolates and the uniform even subgraph of $\mathbb{G}$ does not. We also show that on the wired $d$-regular tree, the phase transitions of the loop $\mathrm{O}(1)$, the single random current, and the random-cluster models are all unique and coincide.

What problem does this paper attempt to address?

The problems that this paper attempts to solve mainly focus on the non - uniqueness of phase transitions in the graphical representation of the Ising Model on tree - like graphs. Specifically, the paper explores whether the phase transitions of the loop O(1) model and the single random current model with respect to the inverse temperature are unique on tree - like graphs. By constructing a certain type of graph, the author proves that the phase transitions of these models are non - unique and that these phase transitions are not monotonic. ### Main research questions: 1. **Non - unique phase transitions**: One of the main objectives of the paper is to prove that on graphs with certain specific structures, the phase transitions of the loop O(1) model and the single random current model are non - unique. This means that at different inverse temperatures, the system may have multiple phase transition points instead of a single critical point. 2. **Non - monotonicity**: Another important issue is to prove that the phase transition probabilities of these models with respect to the inverse temperature are non - monotonic. That is, as the inverse temperature increases, the probability of phase transition may first increase and then decrease, or vice versa. ### Specific achievements: - **Theorem 1.1**: There exists a graph $M$ such that $x\mapsto\ell_0^{x,M}[C_{\infty}]$ is not monotonic. Here $C_{\infty}$ represents the event of the existence of an infinite cluster. - **Corollary 1.2**: There exist two graphs $G'\subset G$ such that $UEG^G[C_{\infty}] = 0$ while $UEG^{G'}[C_{\infty}]= 1$. This shows that the percolation properties of the uniform even subgraph may exhibit different behaviors on different graphs. - **Theorem 1.3**: For any $d\geq2$ and $n\geq1$, on the $d$-regular tree $T_d$ and on the graph $C_d^n$ obtained by replacing each edge with a cycle of length $2n$, the phase transition points of the loop O(1) model, the single random current model and the random cluster model are unique, and their phase transition points satisfy the following relationships: \[ x_c(\ell_1^{T_d})=x_c(P_1^{T_d}) = x_c(P_1^{T_d}\cup P_1^{T_d})=x_c(\phi_1^{T_d}) \] \[ x_c(\ell_0^{T_d})>x_c(P_0^{T_d})>x_c(P_0^{T_d}\cup P_0^{T_d})>x_c(\phi_0^{T_d}) \] ### Methods and techniques: - **Cut - point factorisation**: The paper utilizes the cut - point factorisation technique to decompose complex graphs into simpler sub - graphs, thereby simplifying the analysis. - **Bernoulli percolation**: By relating the models to Bernoulli percolation, the author is able to use the known results of percolation theory to prove their conclusions. - **Critical parameter calculation**: For specific graph structures, such as the $d$-regular tree and $C_d^n$, the paper obtains the critical parameters of each model through direct calculation. ### Conclusion: By constructing specific graph structures, the paper proves that in some cases, the phase transitions of the graphical representations of the Ising Model (such as the loop O(1) model and the single random current model) are non - unique and non - monotonic. This finding provides a new perspective for understanding the phase transition behavior in complex networks and lays the foundation for further research on the behavior of these models on different graph structures.

Non-uniqueness of phase transitions for graphical representations of the Ising model on tree-like graphs

Ising-like Phase Transition in the Fully Frustrated XY Model with Weak Disorder

The Uniform Even Subgraph and Its Connection to Phase Transitions of Graphical Representations of the Ising Model

Cardinality Of Phase-Transition Of Ising-Models On Closed Cayley Trees

Non-reversible stationary states for majority voter and Ising dynamics on trees

Phase Transition for the Ising Model on the Critical Lorentzian Triangulation

Bifurcation in correlation length of the Ising model on a "Toblerone" lattice

Bifurcation in correlation length of the Ising model on a 'Toblerone' lattice

Sharp phase transition for the random-cluster and Potts models via decision trees

On Gibbs measures of the Ising model on [math]-ary trees

The weak limit of Ising models on locally tree-like graphs

Phase Transitions in the Simplicial Ising Model on Hypergraphs

Phase Transitions in single species Ising Models with Non-Reciprocal couplings

Ising Model on Locally Tree-like Graphs: Uniqueness of Solutions to Cavity Equations

Non-linear conductances of Galton-Watson trees and application to the (near) critical random cluster model

Non-triviality of the phase transition for percolation on finite transitive graphs

Phase transition in loop percolation

A combinatorial approach to phase transitions in random graph isomorphism problems

A Bound On The Phase-Transition Region For Ising-Models On Closed Cayley Trees

Phase diagram and metastability of the Ising model on two coupled networks

Double phase transition of the Ising model in core-periphery networks