Disordered Systems and Neural Networks
New submissions
[ showing up to 2000 entries per page: fewer  more ]
New submissions for Tue, 25 Jan 22
 [1] arXiv:2201.09498 [pdf, ps, other]

Title: Phase transitions in XY models with randomly oriented crystal fieldsComments: 14 pages, 9 figuresSubjects: Statistical Mechanics (condmat.statmech); Disordered Systems and Neural Networks (condmat.disnn)
We obtain a representation of the free energy of an XY model on a fully connected graph with spins subjected to a random crystal field of strength $D$ and with random orientation $\alpha$. Results are obtained for an arbitrary probability distribution of the disorder using large deviation theory, for any $D$. We show that the critical temperature is insensitive to the nature and strength of the distribution $p(\alpha)$, for a large family of distributions which includes quadriperiodic distributions, with $p(\alpha)=p(\alpha+\frac{\pi}{2})$, which includes the uniform and symmetric bimodal distributions. The specific heat vanishes as temperature $T \rightarrow 0$ if $D$ is infinite, but approaches a constant if $D$ is finite. We also studied the effect of asymmetry on a bimodal distribution of the orientation of the random crystal field and obtained the phase diagram comprising four phases: a mixed phase (in which spins are canted at angles which depend on the degree of disorder), an $x$Ising phase, a $y$Ising phase and a paramagnetic phase, all of which meet at a tetracritical point. The canted mixed phase is present for all finite $D$, but vanishes when $D \rightarrow \infty$.
 [2] arXiv:2201.09723 [pdf, other]

Title: Predicting the failure of twodimensional silica glassesAuthors: Francesc FontClos, Marco Zanchi, Stefan Hiemer, Silvia Bonfanti, Roberto Guerra, Michael Zaiser, Stefano ZapperiComments: 41 pages 13 figuresSubjects: Materials Science (condmat.mtrlsci); Disordered Systems and Neural Networks (condmat.disnn); Data Analysis, Statistics and Probability (physics.dataan)
Being able to predict the failure of materials based on structural information is a fundamental issue with enormous practical and industrial relevance for the monitoring of devices and components. Thanks to recent advances in deep learning, accurate failure predictions are becoming possible even for strongly disordered solids, but the sheer number of parameters used in the process renders a physical interpretation of the results impossible. Here we address this issue and use machine learning methods to predict the failure of simulated two dimensional silica glasses from their initial undeformed structure. We then exploit Gradientweighted Class Activation Mapping (GradCAM) to build attention maps associated with the predictions, and we demonstrate that these maps are amenable to physical interpretation in terms of topological defects and local potential energies. We show that our predictions can be transferred to samples with different shape or size than those used in training, as well as to experimental images. Our strategy illustrates how artificial neural networks trained with numerical simulation results can provide interpretable predictions of the behavior of experimentally measured structures.
 [3] arXiv:2201.09847 [pdf, other]

Title: Disorderinduced dynamical Griffiths singularities after certain quantum quenchesComments: Main text: 6 pages, 4 figures. Supplemental Material: 6 pages, 2 figuresSubjects: Statistical Mechanics (condmat.statmech); Disordered Systems and Neural Networks (condmat.disnn); Strongly Correlated Electrons (condmat.strel)
The properties of matter in thermal equilibrium can be largely affected by atypical and rare spatial disorder configurations due to inhomogeneities. They induce YangLeeFisher zeros in the partition function which are arbitrarily close to the realtemperature (or field) axis. In simpler words, the partition function becomes a nonanalytical function: a phenomenon known as Griffiths singularities. We demonstrate that those same rare regions produce a dynamical analogue of this phenomenon on closed quantum systems. As a consequence, certain observables become nonanalytical functions of time after certain quantum quenches. We show that the nonanalyticities occur already at the earlytime stages of the unitary evolution, and are due to quantumquenchinduced excitations localized inside those rare regions. We finally discuss on the strength of the singularities associated to the nonanalyticities, and on their numerical and experimental signatures.
Replacements for Tue, 25 Jan 22
 [4] arXiv:2105.04591 (replaced) [pdf, other]

Title: Anomalous mobility edges in onedimensional quasiperiodic modelsJournalref: SciPost Phys. 12, 027 (2022)Subjects: Disordered Systems and Neural Networks (condmat.disnn); Quantum Gases (condmat.quantgas); Optics (physics.optics); Quantum Physics (quantph)
 [5] arXiv:2107.07710 (replaced) [pdf, other]

Title: Stability and Dynamics of ManyBody Localized Systems Coupled to Small BathSubjects: Disordered Systems and Neural Networks (condmat.disnn); Quantum Physics (quantph)
 [6] arXiv:2108.07893 (replaced) [pdf, other]

Title: Failure of the simultaneous block diagonalization technique applied to complete and cluster synchronization of random networksComments: Accepted for publication in PRESubjects: Disordered Systems and Neural Networks (condmat.disnn); Systems and Control (eess.SY)
 [7] arXiv:2104.06664 (replaced) [pdf, ps, other]

Title: Solution of the random field $XY$ magnet on a fully connected graphComments: replaced with the version accepted for publication in J. Phys. A : Math. TheorSubjects: Statistical Mechanics (condmat.statmech); Disordered Systems and Neural Networks (condmat.disnn)
 [8] arXiv:2112.11578 (replaced) [pdf, other]

Title: Mean field theory of yielding under oscillatory shearComments: 23 pages, 20 figures (reference added)Subjects: Soft Condensed Matter (condmat.soft); Disordered Systems and Neural Networks (condmat.disnn)
[ showing up to 2000 entries per page: fewer  more ]
Disable MathJax (What is MathJax?)
Links to: arXiv, form interface, find, condmat, recent, 2201, contact, help (Access key information)