We gratefully acknowledge support from
the Simons Foundation and member institutions.

Disordered Systems and Neural Networks

New submissions

[ total of 13 entries: 1-13 ]
[ showing up to 2000 entries per page: fewer | more ]

New submissions for Thu, 9 Jul 20

[1]  arXiv:2007.03738 [pdf, other]
Title: Skin effect and winding number in disordered non-Hermitian systems
Comments: 6 pages, 3 pages supplement
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn); Quantum Physics (quant-ph)

Unlike their Hermitian counterparts, non-Hermitian (NH) systems may display an exponential sensitivity to boundary conditions and an extensive number of edge-localized states in systems with open boundaries, a phenomena dubbed the "non-Hermitian skin effect." The NH skin effect is one of the primary challenges to defining a topological theory of NH Hamiltonians, as the sensitivity to boundary conditions invalidates the traditional bulk-boundary correspondence. The NH skin effect has recently been connected to the winding number, a topological invariant unique to NH systems. In this paper, we extend the definition of the winding number to disordered NH systems by generalizing established results on disordered Hermitian topological insulators. Our real-space winding number is self-averaging, continuous as a function of the parameters in the problem, and remains quantized even in the presence of strong disorder. We verify that our real-space formula still predicts the NH skin effect, allowing for the possibility of predicting and observing the NH skin effect in strongly disordered NH systems. As an application we apply our results to predict a NH Anderson skin effect where a skin effect is developed as disorder is added to a clean system, and to explain recent results in optical funnels.

[2]  arXiv:2007.03837 [pdf, ps, other]
Title: Many-body localization in a non-Hermitian quasi-periodic system
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn)

In present study, the interplay among interaction, topology, quasi-periodicity, and non-Hermiticity is studied. The hard-core bosons model on a one-dimensional lattice with an asymmetry hopping and a quasi-periodic onsite potential is selected. This model, which preserves time-reversal symmetry, will exhibit three types of phase transitions: real-complex eigenenergies transition, topological phase transition and many-body localization phase transition. In the real-complex eigenenergies transition, it is found that the imaginary parts of the eigeneneriges are always suppressed by the many-body localization. Moreover, by calculating the winding number, a topological phase transition can be revealed with the increase of potential amplitude, and we find that the behavior is quite different from the single-particle systems. Based on our numerical results, we conjecture that these three types of transition occur at the same point in the thermodynamic limit, and the many-body localization transition of quasi-periodic system and disorder system should belong to different universality class. Finally, we demonstrate that these phase transition can profoundly affect the dynamics of the non-Hermitian many-body system.

[3]  arXiv:2007.04053 [pdf, other]
Title: A fast approach to Anderson localization for even-$N$ Dyson insulator
Comments: 7 pages 7 figures
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn)

Dyson insulators with random hoppings in a lattice approach localization faster compared to the usual Anderson insulators with site disorder. For even-$N$ lattice sites the Dyson insulators mimic topological insulators with a pseudo-gap at the band center and the energy-level statistics obtained via the $P(S)$ distribution is of an intermediate type close to the Anderson localized Poisson limit. For odd-$N$ level-repulsion and Wigner statistics appears as in the quasi-metallic regime of $2D$ Anderson insulators, plus a single $E=0$ mode protected by chiral symmetry. The distribution of the participation ratio and the multifractal dimensions of the midband state are computed. In $1D$ the Dyson state is localized and in $2D$ is fractal. Our results might be relevant for recent experimental studies of chiral localization in photonic waveguide arrays.

Cross-lists for Thu, 9 Jul 20

[4]  arXiv:2007.03719 (cross-list from physics.comp-ph) [pdf, other]
Title: Reproducible gray-box neural network for predicting the fragility index and the temperature-dependency of viscosity
Authors: Daniel R. Cassar
Comments: 11 pages, 6 figures
Subjects: Computational Physics (physics.comp-ph); Disordered Systems and Neural Networks (cond-mat.dis-nn); Soft Condensed Matter (cond-mat.soft)

The temperature-dependence of the viscosity of a liquid is relevant in many scientific and technological fields, for example, it is critical to adjust process variables for glass making. The current trend in glass science is building reliable models for property prediction to accelerate glass development. Recently, Tandia and co-authors developed a gray-box neural network model with high performance; they connected the pattern recognition of neural networks with a physical model, the MYEGA equation. Similarly, the aim of this work was to use the SciGlass database to build an open-source gray-box model to predict viscosity. The viscosity dataset used had about 130,000 examples with 28 different chemical elements. This new gray-box model included a pre-processing unit that extracts and scales chemical features before feeding them to the neural network. The best model (after hyperparameter tuning) had a coefficient of determination ($R^{2}$) of 0.987 and root mean squared error (RMSE) of 0.59, both computed for the holdout dataset, which was not used for training. In addition to the temperature-dependence of viscosity, the fragility index of the liquid can also be computed by the gray-box model. The hope is that this free and open framework for property prediction can be used and improved by the community to accelerate the development of new materials.

[5]  arXiv:2007.03822 (cross-list from quant-ph) [pdf, other]
Title: Statistical Mechanics of Quantum Error-Correcting Codes
Comments: 22 pages, 8 figures
Subjects: Quantum Physics (quant-ph); Disordered Systems and Neural Networks (cond-mat.dis-nn); Statistical Mechanics (cond-mat.stat-mech)

We study stabilizer quantum error-correcting codes (QECC) generated under hybrid dynamics of local Clifford unitaries and local Pauli measurements in one dimension. Building upon 1) a general formula relating the error-susceptibility of a subregion to its entanglement properties, and 2) a previously established mapping between entanglement entropies and domain wall free energies of an underlying spin model, we propose a statistical mechanical description of the QECC in terms of "entanglement domain walls". Free energies of such domain walls generically feature a leading volume law term, coming from its "surface energy", and a sub-volume law correction coming from thermodynamic entropies of its transverse fluctuations. These are most easily accounted for by capillary-wave theory of liquid-gas interfaces, which we use as an illustrative tool. We show that the information-theoretic decoupling criterion corresponds to a geometric decoupling of domain walls, which further leads to the identification of the "contiguous code distance" of the QECC as the crossover length scale at which the energy and entropy of the domain wall are comparable. The contiguous code distance thus diverges with the system size as the subleading entropic term of the free energy, protecting a finite code rate against local undetectable errors. We support these correspondences with numerical evidence, where we find capillary-wave theory describes many qualitative features of the QECC; we also discuss when and why it fails to do so.

[6]  arXiv:2007.03871 (cross-list from cond-mat.stat-mech) [pdf, other]
Title: A scaling law describes the spin-glass response in theory, experiments and simulations
Comments: 6 pages, 4 figures
Subjects: Statistical Mechanics (cond-mat.stat-mech); Disordered Systems and Neural Networks (cond-mat.dis-nn)

The correlation length $\xi$, a key quantity in glassy dynamics, can now be precisely measured for spin glasses both in experiments and in simulations. However, known analysis methods lead to discrepancies either for large external fields or close to the glass temperature. We solve this problem by introducing a scaling law that takes into account both the magnetic field and the time-dependent spin-glass correlation length. The scaling law is successfully tested against experimental measurements in a CuMn single crystal and against large-scale simulations on the Janus~II dedicated computer.

Replacements for Thu, 9 Jul 20

[7]  arXiv:1910.06024 (replaced) [pdf, other]
Title: Direct measurement of non-local interactions in the many-body localized phase
Comments: 5+28 pages, 5+22 figures, updated version
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn); Statistical Mechanics (cond-mat.stat-mech); Strongly Correlated Electrons (cond-mat.str-el); Quantum Physics (quant-ph)
[8]  arXiv:2004.03256 (replaced) [pdf, ps, other]
Title: Inflation versus projection sets in aperiodic systems: The role of the window in averaging and diffraction
Comments: Minor adjustments, including slight change of title. 13 pages, many figures. Topical review linked to invited keynote lecture at the 2020 IUCr congress (which has been postponed to 2021)
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn); Metric Geometry (math.MG)
[9]  arXiv:2004.06329 (replaced) [pdf, ps, other]
Title: Replica analysis of overfitting in generalized linear models
Comments: 45 pages, 7 figures, accepted for publication in Journal of Physics A
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn); Statistics Theory (math.ST)
[10]  arXiv:1710.11056 (replaced) [pdf, other]
Title: Noise-tolerant quantum speedups in quantum annealing without fine tuning
Comments: 21 pages, 7 figures
Subjects: Quantum Physics (quant-ph); Disordered Systems and Neural Networks (cond-mat.dis-nn); Superconductivity (cond-mat.supr-con)
[11]  arXiv:1909.08460 (replaced) [pdf, other]
Title: Protection of helical transport in Quantum Spin Hall samples: the role of symmetries on edges
Comments: Substantially re-written version: improved explanation of the setup, Abstract, and Conclusions; more detailed proof of the main result; re-arranged and substantially extended examples and Supplemental Materials; extended Bibliography; 11 pages, 2 figures
Subjects: Mesoscale and Nanoscale Physics (cond-mat.mes-hall); Disordered Systems and Neural Networks (cond-mat.dis-nn)
[12]  arXiv:1910.11359 (replaced) [pdf, other]
Title: Time evolution of many-body localized systems in two spatial dimensions
Comments: 11 pages, 8 figures, revised version, new plots and phase diagram added
Subjects: Strongly Correlated Electrons (cond-mat.str-el); Disordered Systems and Neural Networks (cond-mat.dis-nn); Quantum Physics (quant-ph)
[13]  arXiv:2006.16811 (replaced) [pdf, other]
Title: Path Integral Based Convolution and Pooling for Graph Neural Networks
Comments: 15 pages, 4 figures, 6 tables. arXiv admin note: text overlap with arXiv:1904.10996
Subjects: Machine Learning (cs.LG); Disordered Systems and Neural Networks (cond-mat.dis-nn); Networking and Internet Architecture (cs.NI); Data Analysis, Statistics and Probability (physics.data-an); Machine Learning (stat.ML)
[ total of 13 entries: 1-13 ]
[ showing up to 2000 entries per page: fewer | more ]

Disable MathJax (What is MathJax?)

Links to: arXiv, form interface, find, cond-mat, recent, 2007, contact, help  (Access key information)