[Paper Review] Network science Ising states of matter

Hanlin Sun, Rajat Kumar Panda+4 more|arXiv (Cornell University)|Aug 25, 2023

Complex Network Analysis Techniques90 references4 citations

TL;DR

This paper introduces a network science framework to characterize phases of matter using Ising snapshot networks (IsingNets) derived from 2D Ising model Monte Carlo simulations. By analyzing topological, spectral, and structural properties—such as percolation, persistent homology, degree distributions, and graph Laplacian spectra—it reveals distinct network signatures across the ferromagnetic, paramagnetic, and critical phases, demonstrating that IsingNets encode universal phase transition features beyond traditional order parameters.

ABSTRACT

Network science provides very powerful tools for extracting information from interacting data. Although recently the unsupervised detection of phases of matter using machine learning has raised significant interest, the full prediction power of network science has not yet been systematically explored in this context. Here we fill this gap by providing an in-depth statistical, combinatorial, geometrical and topological characterization of 2D Ising snapshot networks (IsingNets) extracted from Monte Carlo simulations of the $2$D Ising model at different temperatures, going across the phase transition. Our analysis reveals the complex organization properties of IsingNets in both the ferromagnetic and paramagnetic phases and demonstrates the significant deviations of the IsingNets with respect to randomized null models. In particular percolation properties of the IsingNets reflect the existence of the symmetry between configurations with opposite magnetization below the critical temperature and the very compact nature of the two emerging giant clusters revealed by our persistent homology analysis of the IsingNets. Moreover, the IsingNets display a very broad degree distribution and significant degree-degree correlations and weight-degree correlations demonstrating that they encode relevant information present in the configuration space of the $2$D Ising model. The geometrical organization of the critical IsingNets is reflected in their spectral properties deviating from the one of the null model. This work reveals the important insights that network science can bring to the characterization of phases of matter. The set of tools described hereby can be applied as well to numerical and experimental data.

Motivation & Objective

• To systematically apply network science tools to characterize phases of matter in the 2D Ising model using configuration-space snapshots.
• To identify unsupervised network-based indicators of phase transitions that go beyond conventional order parameters.
• To compare IsingNets with randomized null models to quantify their non-random structural, topological, and spectral properties.
• To explore the utility of persistent homology, centrality measures, and spectral graph theory in detecting critical phenomena in spin systems.
• To establish a generalizable framework for analyzing numerical and experimental many-body data through network-based topological data analysis.

Proposed method

• Construct IsingNets from Monte Carlo spin configurations by connecting spins within a distance threshold based on the 5th nearest neighbor average distance.
• Apply filtration-based percolation analysis to study the emergence of giant connected components across temperature regimes.
• Use persistent homology on clique complexes of IsingNets to quantify topological features such as Betti numbers and detect compact cluster formation.
• Compute graph Laplacian eigenvalue distributions and von Neumann entropy to probe spectral properties and intrinsic dimensionality of IsingNets.
• Perform UMAP and MST-based network embeddings to visualize structural changes across phases.
• Compare all network properties of IsingNets against randomized null models to isolate non-random, physically meaningful features.

Experimental results

Research questions

• RQ1 How do the topological and structural properties of IsingNets change across the ferromagnetic-to-paramagnetic phase transition in the 2D Ising model?
• RQ2 To what extent do persistent homology and Betti numbers of IsingNets reveal compact cluster formation and symmetry-breaking in the ordered phase?
• RQ3 Can spectral properties of IsingNets, such as eigenvalue distributions and von Neumann entropy, serve as unsupervised indicators of criticality?
• RQ4 How do degree-degree and weight-degree correlations in IsingNets reflect the underlying spin correlation structure across phases?
• RQ5 Do IsingNets display significant deviations from randomized null models in combinatorial, geometric, and topological features, and what do these deviations imply about the physical state?

Key findings

• IsingNets exhibit two distinct giant connected components in the ferromagnetic phase, corresponding to configurations with opposite magnetization, reflecting spontaneous symmetry breaking.
• Persistent homology analysis shows suppressed Betti numbers in IsingNets compared to null models, indicating compact, highly organized clusters in the spin configuration space.
• The eigenvalue distribution of the IsingNet graph Laplacian displays a power-law growth with exponent $\hat{d} = 0.78 \pm 0.04$ near the critical temperature, signaling critical scaling behavior.
• IsingNets have a broad degree distribution with strong degree-degree and weight-degree correlations, indicating heterogeneous, correlated network organization beyond random graphs.
• The von Neumann entropy of IsingNets peaks near the critical temperature $T_c$, serving as a robust unsupervised indicator of the phase transition.
• Spectral and topological features of IsingNets deviate significantly from randomized null models, confirming that they encode non-trivial physical information about the spin system’s phase structure.