The order parameter for spin glasses: a function on the interval 0-1

Renormalization-group computation of the critical exponents of hierarchical spin glasses: large-scale behavior and divergence of the correlation length

In a recent work [M. Castellana and G. Parisi, Phys. Rev. E 82, 040105(R) (2010)], the large-scale behavior of the simplest non-mean-field spin-glass system has been analyzed, and the critical exponent related to the divergence of the correlation length has been computed at two loops within the ε-expansion technique by two independent methods. By performing the explicit calculation of the critical exponents at two loops, one obtains that the two methods yield the same result. This shows that the underlying renormalization group ideas apply consistently in this disordered model, in such a way that an ε-expansion can be set up. The question of the extension to high orders of this ε-expansion is particularly interesting from the physical point of view. Indeed, once high orders of the series in ε for the critical exponents are known, one could check the convergence properties of the series, and find out if the ordinary series resummation techniques, yielding very accurate predictions for the Ising model, work also for this model. If this is the case, a consistent and predictive non-mean-field theory for such a disordered system could be established. In that regard, in this work we expose the underlying techniques of such a two-loop computation. We show with an explicit example that such a computation could be quite easily automatized, i.e., performed by a computer program, in order to compute high orders of the ε-expansion, and so eventually make this theory physically predictive. Moreover, all the underlying renormalization group ideas implemented in such a computation are widely discussed and exposed.

Equilibrium state of the mean-field Potts glass

We investigate the low-temperature equilibrium state of the p-state mean-field Potts glass. It is the presently believed that there is a temperature interval T(2) < T < T(c) within which the equilibrium state for p > p(*) approximately 2.82 is described by the cavity method corresponding to the first level of replica-symmetry breaking. We demonstrate that this locally stable solution is globally unstable and that true equilibrium in the region of instability of the replica-symmetric solution (T < T(c)), marked by the highest free energy, is a state displaying a continuous replica-symmetry breaking.

de Almeida-Thouless line in vector spin glasses

We consider the infinite-range spin glass in which the spins have m>1 components (a vector spin glass). Applying a magnetic field which is random in direction, there is a de Almeida-Thouless (AT) line below which the "replica symmetric" solution is unstable, just as for the Ising (m=1) case. We calculate the location of this AT line for Gaussian random fields for arbitrary m and verify our results by numerical simulations for m=3 .

Physics and complexity

This paper is concerned with complex macroscopic behaviour arising in many-body systems through the combinations of competitive interactions and disorder, even with simple ingredients at the microscopic level. It attempts to indicate and illustrate the richness that has arisen, in conceptual understanding, in methodology and in application, across a large range of scientific disciplines, together with a hint of some of the further opportunities that remain to be tapped. In doing so, it takes the perspective of physics and tries to show, albeit rather briefly, how physics has contributed and been stimulated.

Spatial dimension and the dynamics of supercooled liquids

Inspired by recent theories that apply ideas from critical phenomena to the glass transition, we have simulated an atomistic model of a supercooled liquid in three and four spatial dimensions. At the appropriate temperatures and density, dynamic density correlation functions in three and four spatial dimensions correspond nearly exactly. Dynamic heterogeneity, quantified through the breakdown of the Stokes-Einstein relationship, is weaker in four dimensions than in three. We discuss this in the context of recent theories for dynamical heterogeneity. Because dimensionality is a crucially important variable, our work adds a stringent test for emerging theories of glassy dynamics.

Study of the de Almeida-Thouless line using power-law diluted one-dimensional Ising spin glasses

We test for the existence of a spin-glass phase transition, the de Almeida-Thouless line, in an externally applied (random) magnetic field by performing Monte Carlo simulations on a power-law diluted one-dimensional Ising spin glass for very large system sizes. We find that a de Almeida-Thouless line occurs only in the mean-field regime, which corresponds, for a short-range spin glass, to dimension d larger than 6.

Dynamical self-arrest in symmetric and asymmetric diblock copolymer melts using a replica approach within a local theory

We investigate dynamical self-arrest in a diblock copolymer melt using a replica approach within a self-consistent local method based on dynamical mean-field theory (DMFT). The local replica approach effectively predicts (chiN)_{A} for dynamical self-arrest in a block copolymer melt for symmetric and asymmetric cases. We discuss the competition of the cubic and quartic interactions in the Landau free energy for a block copolymer melt in stabilizing a glassy state depending on the chain length. Our local replica theory provides a universal value for the dynamical self-arrest in block copolymer melts with (chiN)_{A} approximately 10.5+64N;{-3/10} for the symmetric case.

Measuring overlaps in mesoscopic spin glasses via conductance fluctuations

We consider the electronic transport in a mesoscopic metallic spin glass. We show that the distribution of overlaps between spin configurations can be inferred from the reduction of the conductance fluctuations by the magnetic impurities. Using this property, we propose new experimental protocols to probe spin glasses directly through their overlaps.

Effect of selection on ancestry: an exactly soluble case and its phenomenological generalization

We consider a family of models describing the evolution under selection of a population whose dynamics can be related to the propagation of noisy traveling waves. For one particular model that we shall call the exponential model, the properties of the traveling wave front can be calculated exactly, as well as the statistics of the genealogy of the population. One striking result is that, for this particular model, the genealogical trees have the same statistics as the trees of replicas in the Parisi mean-field theory of spin glasses. We also find that in the exponential model, the coalescence times along these trees grow like the logarithm of the population size. A phenomenological picture of the propagation of wave fronts that we introduced in a previous work, as well as our numerical data, suggest that these statistics remain valid for a larger class of models, while the coalescence times grow like the cube of the logarithm of the population size.

Double criticality of the Sherrington-Kirkpatrick model at T=0

Numerical results up to the 42nd order of replica-symmetry breaking (RSB) are used to predict the singular structure of the Sherrington-Kirkpatrick spin glass at T=0. We confirm predominant single parameter scaling and derive corrections for the T=0 order function q(a), related to a Langevin equation with pseudotime 1/a. a=0 and a=infinity are shown to be two critical points for infinity-RSB, associated with two discrete spectra of Parisi block size ratios, attached to a continuous spectrum. Finite-RSB-size scaling, associated exponents, and T=0-energy are obtained with unprecedented accuracy.

Low-temperature solution of the Sherrington-Kirkpatrick model

I propose a simple scaling ansatz for the full replica symmetry breaking solution of the Sherrington-Kirkpatrick model in the low energy sector. This solution is shown to become exact in the limit x --> 0, Bx --> infinity of the Parisi replica symmetry breaking scheme parameter . The distribution function of the frozen fields has been known to develop a linear gap at zero temperature. The scaling equations are integrated to find an exact numerical value for the slope of the gap thetaP(x,y)/delta|(y --> 0) = 0.301 046.... I also use the scaling solution to devise an inexpensive numerical procedure for computing finite time scale (x =1) quantities. The entropy, the zero field cooled susceptibility, and the local field distribution function are computed in the low-temperature limit with high precision, barely achievable by currently available methods.

Inverse melting and inverse freezing: a spin model

Systems of highly degenerate ordered or frozen state may exhibit inverse melting (reversible crystallization upon heating) or inverse freezing (reversible glass transition upon heating). This phenomenon is reviewed, and a list of experimental demonstrations and theoretical models is presented. A simple spin model for inverse melting is introduced and solved analytically for infinite range, constant paramagnetic exchange interaction. The random exchange analogue of this model yields inverse freezing, as implied by the analytic solution based on the replica trick. The qualitative features of this system (generalized Blume-Capel spin model) are shown to resemble a large class of inverse melting phenomena. The appearance of inverse melting is related to an exact rescaling of one of the interaction parameters that measures the entropy of the system. For the case of almost degenerate spin states, perturbative expansion is presented, and the first three terms correspond to the empiric formula for the Flory-Huggins chi parameter in the theory of polymer melts. The possible microscopic origin of this chi parameter and the limitations of the Flory-Huggins theory where the state degeneracy is associated with the different conformations of a single polymer or with the spatial structures of two interacting molecules are discussed.

Defect energy of infinite-component vector spin glasses

We compute numerically the zero-temperature defect energy DeltaE of the vector spin glass in the limit of an infinite number of spin components m , for a range of dimensions 2< or d < or =5 . Fitting to DeltaE approximately L(theta) , where L is the system size, we obtain: theta similar to-1.54 (d=2) , theta similar to-1.04 (d=3) , theta similar to -0.67 (d=4) , and theta similar to -0.37 (d=5) . These results show that the lower critical dimension dl (the dimension where theta changes sign) is significantly higher for m=infinity than for finite m (where 2< dl <3 ).

Spin glasses in the limit of an infinite number of spin components

We consider spin glass models in which the number of spin components m is infinite. In the formulation of the problem appropriate for numerical calculations proposed by several authors, we show that the order parameter defined by the long-distance limit of the correlation functions is actually zero and there is only "quasi-long-range order" below the transition temperature. Nonetheless, there can be a finite temperature phase transition where the decay of correlations changes from exponential to power law. We also show that the spin glass transition temperature is zero in three dimensions so power-law behavior only occurs at T=0 in this case. We also argue that the order of limits, m-->infinity and N-->infinity is important, where N is the number of spins.

Phase transition in vector spin glasses

We first give an experimental and theoretical introduction to spin glasses, and then discuss the nature of the phase transition in spin glasses with vector spins. Results of Monte Carlo simulations of the Heisenberg spin glass model in three dimensions are presented. A finite size scaling analysis of the correlation length of the spins and chiralities shows that there is a single, finite-temperature transition at which both spins and chiralities order.

Clustering analysis of the ground-state structure of the vertex-cover problem

Vertex cover is one of the classical NP-complete problems in theoretical computer science. A vertex cover of a graph is a subset of vertices such that for each edge at least one of the two endpoints is contained in the subset. When studied on Erdo s-Re nyi random graphs (with connectivity c) one observes a threshold behavior: In the thermodynamic limit the size of the minimal vertex cover is independent of the specific graph. Recent analytical studies show that on the phase boundary, for small connectivities c<e , the system is replica symmetric, while for larger connectivities replica symmetry breaking occurs. This change coincides with a change of the typical running time of algorithms from polynomial to exponential. To understand the reasons for this behavior and to compare with the analytical results, we numerically analyze the structure of the solution landscape. For this purpose, we have also developed an algorithm, which allows the calculation of the backbone, without the need to enumerate all solutions. We study exact solutions found with a branch-and-bound algorithm as well as configurations obtained via a Monte Carlo simulation. We analyze the cluster structure of the solution landscape by direct clustering of the states, by analyzing the eigenvalue spectrum of correlation matrices and by using a hierarchical clustering method. All results are compatible with a change at c=e . For small connectivities, the solutions are collected in a finite small number of clusters, while the number of clusters diverges slowly with system size for larger connectivities and replica symmetry breaking, but not one-step replica symmetry breaking (1-RSB) occurs.

Absence of an Almeida-Thouless line in three-dimensional spin glasses

We present results of Monte Carlo simulations of the three-dimensional Edwards-Anderson Ising spin glass in the presence of a (random) field. A finite-size scaling analysis of the correlation length shows no indication of a transition, in contrast with the zero-field case. This suggests that there is no Almeida-Thouless line for short-range Ising spin glasses.

Numerical study of the Sherrington-Kirkpatrick model in a magnetic field

We study numerically the Sherrington-Kirkpatrick model as a function of the magnetic field h, with fixed temperature T=0.6T(c). We investigate the finite size scaling behavior of several quantities, such as the spin-glass susceptibility, searching for numerical evidences of the transition on the de Almeida-Thouless line. We find strong corrections to scaling which make difficult to locate the transition point. This shows, in a simple case, the extreme difficulties of spin-glass simulations in a nonzero magnetic field. Next, we study various sum rules (consequences of stochastic stability) involving overlaps between three and four replicas, which appear to be numerically well satisfied, and in a nontrivial way. Finally, we present data on P(q) for a large lattice size (N=3200) at low temperature T=0.4T(c), where the shape predicted by the replica symmetry breaking solution of the model for a nonzero magnetic field is visible.

Free energy fluctuations in Ising spin glasses

The sample-to-sample fluctuations of the free energy in finite-dimensional Ising spin glasses are calculated, using the replica method, from higher order terms in the replica number n. It is shown that the Parisi symmetry breaking scheme does not give the correct answers for these higher order terms. A modified symmetry breaking scheme with the same stability is shown to resolve the problem.

Role of the interaction matrix in mean-field spin glass models

Mean-field models of two-spin Ising spin glasses with interaction matrices taken from ensembles that are invariant under O(N) transformations are studied. A general study shows that the nature of the spin glass transition can be deduced from the eigenvalue spectrum of the interaction matrix. A simple replica approach is derived to carry out the average over the O(N) disorder. The analytic results are confirmed by the extensive Monte Carlo simulations for large system sizes and by the exact enumeration for small system sizes.

Corrections to scaling are large for droplets in two-dimensional spin glasses

The energy of a droplet of linear extent l in the droplet theory of spin glasses goes as l(theta) for large l. It is shown by numerical studies of large droplets in two-dimensional systems that this formula needs to be modified by the addition of a scaling correction l(-omega) in order to accurately describe droplet energies at the length scales currently probed in numerical simulations. Using this simple modification, it is now possible to explain many results which have been found in simulations of three-dimensional Ising spin glasses with the droplet model.

Magnetic field chaos in the Sherrington-Kirkpatrick model

We study the Sherrington-Kirkpatrick model, both above and below the de Almeida-Thouless line, by using a modified version of the Parallel Tempering algorithm in which the system is allowed to move between different values of the magnetic field h. The behavior of the probability distribution of the overlap between two replicas at different values of the magnetic field h(0) and h(1) gives clear evidence for the presence of magnetic field chaos already for moderate system sizes, in contrast to the case of temperature chaos, which is not visible on system sizes that can currently be thermalized.

Fragility of the free-energy landscape of a directed polymer in random media

We examine the sensitiveness of the free-energy landscape of a directed polymer in random media with respect to various kinds of infinitesimally weak perturbation including the intriguing case of temperature chaos. To this end, we combine the replica Bethe Ansatz approach outlined by Sales and Yoshino (e-print cond-mat/0112384), the mapping to a modified Sinai model, and numerically exact calculations by the transfer-matrix method. Our results imply that for all the perturbations under study there is a slow crossover from a weakly perturbed regime, where rare events take place, to a strongly perturbed regime at larger length scales beyond the so-called overlap length, where typical events take place leading to chaos, i.e., a complete reshuffling of the free-energy landscape. Within the replica space, the evidence for chaos is found in the factorization of the replicated partition function induced by infinitesimal perturbations. This is the reflex of explicit replica-symmetry breaking.

Macroscopically frustrated Ising model

A disordered spin-glass model in which both static and dynamical properties depend on macroscopic magnetizations is presented. These magnetizations interact via random couplings and, therefore, the typical quenched realization of the system exhibits a macroscopic frustration. The model is solved by using a revisited replica approach, and the broken symmetry solution turns out to coincide with the symmetric solution. Some dynamical aspects of the model are also discussed, showing how it could be a useful tool for describing some properties of real systems such as, for example, natural ecosystems or human social systems.

Macroscopic dynamics in separable neural networks

Parallel dynamics of neural networks with separable coupling is given starting from Coolen-Sherrington theory. Away from saturation, it is shown that this parallel retrieve dynamics is equivalent to the sequential dynamics for finite temperature. But the finite-size effects were found to be governed by a homogeneous Markov process, not by a time-dependent Ornstein-Uhlenbeck process in sequential dynamics.

Ground-state clusters of two-, three-, and four-dimensional +/-J Ising spin glasses

A huge number of independent true ground-state configurations is calculated for two-, three- and four-dimensional +/-J spin-glass models. Using the genetic cluster-exact approximation method, system sizes up to N=20(2),8(3),6(4) spins are treated. A "ballistic-search" algorithm is applied, which allows even for large system sizes to identify clusters of ground states that are connected by chains of zero-energy flips of spins. The number of clusters n(C) diverges with N going to infinity. For all dimensions considered here, an exponential increase of n(C) appears to be more likely than a growth with a power of N. The number of different ground states is found to grow clearly exponentially with N. A zero-temperature entropy per spin of s(0)=0.078(5)k(B) (2D), s(0)=0.051(3)k(B) (3D), respectively, s(0)=0.027(5)k(B) (4D) is obtained.