Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · cond-mat/9802102

Self-Averaging, Distribution of Pseudo-Critical Temperatures and Finite Size Scaling in Critical Disordered Systems

S. Wiseman·E. Domany
View original source ↗Read source document ↗Export metadata ↓

Abstract

The distributions $P(X)$ of singular thermodynamic quantities in an ensemble of quenched random samples of linear size $l$ at the critical point $T_c$ are studied by Monte Carlo in two models. Our results confirm predictions of Aharony and Harris based on Renormalization group considerations. For an Ashkin-Teller model with strong but irrelevant bond randomness we find that the relative squared width, $R_X$, of $P(X)$ is weakly self averaging. $R_X\sim l^{α/ν}$, where $α$ is the specific heat exponent and $ν$ is the correlation length exponent of the pure model fixed point governing the transition. For the site dilute Ising model on a cubic lattice, known to be governed by a random fixed point, we find that $R_X$ tends to a universal constant independent of the amount of dilution (no self averaging). However this constant is different for canonical and grand canonical disorder. We study the distribution of the pseudo-critical temperatures $T_c(i,l)$ of the ensemble defined as the temperatures of the maximum susceptibility of each sample. We find that its variance scales as $(δT_c(l))^2 \sim l^{-2/ν}$ and NOT as $\sim l^{-d}. We find that $R_χ$ is reduced by a factor of $\sim 70$ with respect to $R_χ(T_c)$ by measuring $χ$ of each sample at $T_c(i,l)$. We analyze correlations between the magnetization at criticality $m_i(T_c,l)$ and the pseudo-critical temperature $T_c(i,l)$ in terms of a sample independent finite size scaling function of a sample dependent reduced temperature $(T-T_c(i,l))/T_c$. This function is found to be universal and to behave similarly to pure systems.

Explore related subjects

cond-mat.dis-nn

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

S. Wiseman, E. Domany. 1998-02-10. Self-Averaging, Distribution of Pseudo-Critical Temperatures and Finite Size Scaling in Critical Disordered Systems. https://doi.org/10.1103/physreve.58.2938

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Deep-layered machines have a built-in Occam's razor

Many complex systems map vast numbers of microscopic descriptions onto far fewer macroscopic outcomes. Surprisingly, these input-output maps strongly favor simple outputs. In an exactly solvable deep-layered machine, we prove that simplicity bias strengthens with every layer. The probability of an output function $F$ approaches $\Pr(F)\propto 2^{-K(F)}$, where $K$ is the Kolmogorov complexity. Our results show that hierarchical composition alone can generate an intrinsic Occam's razor, suggesting a general mechanism for simplicity bias in learning and evolution.

cond-mat.dis-nn↗
arXiv2026

Neural Networks with Exciton-Polariton Waveguides

We propose and numerically study a scheme for neural networks using exciton-polariton waveguide meshes. These enable arbitrary programmable interlayer couplings suited to a restricted Boltzmann machine architecture where polariton phases act as stochastic network variables. We show that such a network can be trained using unsupervised Hebbian and anti-Hebbian updates. This allows the implementation of an associative memory, where the network recovers learned patterns from imperfect inputs.

cond-mat.dis-nn↗
arXiv2026

Finite-width shift of the edge of chaos in deep neural networks

Deep neural networks train best when they are initialised at the edge of chaos, the critical line in the plane of weight and bias variances on which the distance between two inputs neither grows nor shrinks with depth. The line is known for infinitely wide networks; at finite width it moves, and how it moves has been disputed. We treat the logarithm of the distance between two inputs through a network of width $N$ as a random walk and compute its drift and variance to order $1/N$. The edge of chaos moves to a larger weight variance, by $Δσ_w^2=-a_1/(N\,\partial_{σ_w^2}\lnχ_1)$, about $8/N$ for $\tanh$ networks at a bias variance of $0.05$, confirmed by simulations of the exact finite-width recursion. The shift describes a typical network; the average over many networks is dominated by rare networks in which the distance grows, and its critical point moves the other way, which explains why calculations and simulations have disagreed. Backpropagation follows the same random walk, so the shift applies to gradients unchanged. Along the edge of chaos, the information about the similarity of two inputs is constant until the compression over the depth exceeds about one nat, which fixes where to sit on the line. The results give an initialisation rule, Crinit, that places a network on its finite-width edge of chaos without a search. On MNIST and Fashion-MNIST at width $64$ it is more accurate than the standard rules at their usual settings, by a small margin in shallow networks and a large one in deep networks, where most standard rules fail; it trains in fewer steps, and it matches or beats the standard rules with a tuned gain.

cond-mat.dis-nn↗