Temperature chaos as a logical consequence of reentrant transition in spin glasses
Hidetoshi Nishimori, Masayuki Ohzeki, Manaka Okuyama · 2025
Read Temperature chaos as a logical consequence of reentrant transition in spin glasses by Hidetoshi Nishimori, Masayuki Ohzeki, Manaka Okuyama (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-00276.txt.
Introduction
Hidetoshi Nishimori, Masayuki Ohzeki, and Manaka Okuyama proved a link in twenty twenty-five between two odd effects in spin glasses. Spin glasses are magnetic alloys with frozen, disordered spins. In temperature chaos, a tiny temperature shift completely rearranges the spin state. In the reentrant transition, cooling drives the system from an ordered phase back into a less ordered one.
Working in the finite-dimensional Edwards-Anderson model, the authors add correlations among the disorder variables. Assuming a spin glass phase exists at finite temperature, they show temperature chaos follows logically from reentrance. The two effects, long treated as unrelated curiosities, turn out to be two faces of one structure. The result offers a new view of spin glass physics beyond the standard mean-field picture.
Catalog record
| CREATOR | Hidetoshi Nishimori, Masayuki Ohzeki, Manaka Okuyama |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-00276
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Temperature chaos as a logical consequence of reentrant transition in spin glasses Hidetoshi Nishimori, Masayuki Ohzeki, Manaka Okuyama Abstract Temperature chaos is a striking phenomenon in spin glasses, where even slight changes in temperature lead to a complete reconfiguration of the spin state. Another intriguing effect is the reentrant transition, in which lowering the temperature drives the system from a ferromagnetic phase into a less ordered spin-glass or paramagnetic phase. In the present paper, we reveal an unexpected connection between these seemingly unrelated phenomena in the finite-dimensional Edwards-Anderson model of spin glasses by introducing a generalized formulation that incorporates correlations among disorder variables. Assuming the existence of a spin glass phase at finite temperature, we establish that temperature chaos arises as a logical consequence of reentrance in the Edwards-Anderson model. Our findings uncover a previously hidden mathematical structure relating reentrance and temperature chaos, offering a new perspective on the physics of spin glasses beyond the mean-field theory. I Introduction Statistical-mechanical studies of spin glasses have uncovered a number of striking phenomena, including replica symmetry breaking in the mean-field Sherrington–Kirkpatrick model Sherrington and Kirkpatrick 1975 ; Parisi 1980 ; Talagrand 2006 , as well as deep connections with fields beyond traditional physics, such as computer science Mézard and Montanari 2009 , inference problems Zdeborová and Krzakala 0116 , and a number of other domains Nishimori 2001 ; Charbonneau et al. 2023 . Many of these theoretical developments concern mean-field-type models. By contrast, much less is known on firm theoretical grounds about the finite-dimensional Edwards–Anderson model Edwards and Anderson 1975 , especially in two and three dimensions, which are most relevant to experiments Mydosh 1993 . For these cases, numerical simulations remain the primary investigative tool. Numerical studies suggest the existence of two counterintuitive phenomena in the finite-dimensional Edwards–Anderson model, reentrant transitions and temperature chaos. A reentrant transition, where the lower-temperature phase is less ordered than the higher-temperature phase, has been observed in two, three, and four dimensions, near the boundary between the ferromagnetic and non-ferromagnetic (spin glass or paramagnetic) phases Nobre 2001 ; Wang et al. 2003 ; Amoruso and Hartmann 2004 ; Hasenbusch et al. 2007 ; Toldin et al. 2009 ; Hasenbusch et al. 2011 ; Thomas and Katzgraber 2011 ; Liu et al. 2025 . Another notable phenomenon is temperature chaos, in which small changes in temperature within the spin glass phase lead to a complete reorganization of the spin configuration Bray and Moore 1987 ; Banavar and Bray 1987 ; Fisher and Huse 1986 ; Fisher and Huse 1988 ; Kondor 1989 ; Ney-Nifle and Young 1997 ; Ney-Nifle 1998 ; Parisi and Rizzo 2010 ; Mathieu et al. 2001 ; Bouchaud et al. 2001 ; Aspelmeier et al. 2002 ; Rizzo and Crisanti 2003 ; Houdayer and Hartmann 2004 ; Katzgraber and Krza¸kała 2007 ; Fernandez et al. 2013 ; Wang et al. 2015 ; Billoire et al. 2018 ; Baity-Jesi et al. 2021 . It is worth noting for later discussions that temperature chaos has been studied predominantly in the case with symmetric distributions of disorder, where positive and negative interactions occur with equal probability, far from the region of the phase diagram where reentrant transitions have been observed. As a result, little attention has been given to the possibility of a connection between the two. One of the notable recent developments in spin glass theory is the discovery of the significant impact of spatial correlations in disorder variables Nishimori 2024 ; Nishimori 2025 ; Braunstein et al. 2025 . Spatially correlated disorder has been shown to induce anomalous behaviors, including straight, non-reentrant, phase boundaries between the ferromagnetic and non-ferromagnetic phases, under the condition that temperature chaos is absent. If temperature chaos is present, the ferromagnetic phase is confined to a single line, the Nishimori line (NL) Nishimori 1981 ; Nishimori 1980 ; Nishimori 1982 ; Nishimori 2001 , and is surrounded by the spin glass phase in the model with strong correlations in disorder variables. Furthermore, assuming replica symmetry breaking in the Edwards–Anderson model, the distribution of the magnetization on the NL exactly matches the distribution of the replica overlap in the spin glass phase, with support on a finite interval. This result implies that the magnetization fluctuates from one measurement to another or from one disorder realization to another, which is an unusual property for a macroscopic quantity. However, these surprising conclusions pertain specifically to models with strongly correlated disorder and cannot be directly extended to the standard Edwards–Anderson model with uncorrelated disorder. In the present paper, we extend the theoretical framework for correlated disorder proposed in Ref. Nishimori 2024 by introducing an additional parameter that controls the level of frustration. Analysis of the resulting model reveals an unexpected relation between temperature chaos and the reentrant transition in the Edwards–Anderson model with uncorrelated disorder. Specifically, we demonstrate that, under the assumption of the existence of a spin glass phase, the phase boundary between the ferromagnetic and spin glass phases is non-reentrant if temperature chaos is absent. The contrapositive of this statement is that temperature chaos exists if the phase boundary is reentrant. This establishes a highly nontrivial mathematical relationship between two seemingly unrelated physical phenomena observed in different regions of the phase diagram in the Edwards–Anderson model, which is derived through the analysis of a model with correlations in disorder variables beyond the Edwards-Anderson model. The correlated disorder model introduced in this work is closely related to the model proposed by Kitatani Kitatani 1992 , who used it to argue against the existence of a reentrant transition in the Edwards–Anderson model. Our model can be viewed as a reformulation of his model, by incorporating a more natural parametrization of the control variables. This formulation allows for a more direct and modern analysis, leading to conclusions that do not necessarily align with Kitatani’s. The structure of this paper is as follows. We begin by introducing a correlated disorder model and show how it reduces to known models in several special cases in Sec. II . Then, in Sec. III , we first analyze several limiting cases to construct 2D cross sections of a 3D phase diagram. 1 1 1 To avoid confusion, we write “3D” for the phase diagram with three axes and “three dimensions” for the spatial dimensionality of the lattice. In the later part of Sec. III , we integrate these results to establish a logical relation between reentrance and temperature chaos from the 3D phase diagram, under the assumption of the existence of a spin glass phase. We conclude the paper in the final section. II Formulation of the problem Let us first introduce the model and define several physical quantities that are central to our discussion. II.1 Problem Definition We consider the \pm J Ising spin glass with the dimensionless Hamiltonian \displaystyle H=-\beta\sum_{\langle ij\rangle}\tau_{ij}S_{i}S_{j}\quad(S_{i}=\pm 1), (1) where \beta is the inverse temperature (coupling constant), \tau_{ij}(=\pm 1) denotes the quenched disorder variable associated with the bond \langle ij\rangle , and S_{i} is the Ising spin at site i . The summation runs over all interacting spin pairs on a given lattice, with no restriction on the dimensionality or the structure of the lattice. Consequently, our theory applies also to the all-to-all