Search

We study the maximum-average submatrix problem, in which given an $N \times N$ matrix $J$ one needs to find the $k \times k$ submatrix with the largest average of entries. We study the problem for random matrices $J$ whose entries are i.i.d. random variables by mapping it to a variant of the Sherrington-Kirkpatrick spin-glass model at fixed magnetization. We characterize analytically the phase diagram of the model as a function of the submatrix average and the size of the submatrix $k$ in the limit $N\to\infty$. We consider submatrices of size $k = m N$ with $0 < m < 1$. We find a rich phase diagram, including dynamical, static one-step replica symmetry breaking and full-step replica symmetry breaking. In the limit of $m \to 0$, we find a simpler phase diagram featuring a frozen 1-RSB phase, where the Gibbs measure is composed of exponentially many pure states each with zero entropy. We discover an interesting phenomenon, reminiscent of the phenomenology of the binary perceptron: there exist efficient algorithms that provably work in the frozen 1-RSB phase.

While empirical shell model calculations have successfully described low-lying nuclear data for decades, only recently has significant effort been made to quantify the uncertainty in such calculations. Here we quantify the statistical error in effective parameters for transition operators in empirical calculations in the $sd$ ($1s_{1/2}$-$0d_{3/2}$-$0d_{5/2}$) valence space, specifically the quenching of Gamow-Teller transitions, effective charges for electric quadrupole (E2) transitions, and the effective orbital and spin couplings for magnetic dipole (M1) transitions. We find the quenching factor for Gamow-Teller transitions relative to free-space values is tightly constrained and that the isoscalar coupling of E2 is much more tightly constrained than the isovector coupling. For effective M1 couplings, we found isovector components more constrained than isoscalar, but that to get any sensible result we had to fix one of four couplings. This detailed quantification of uncertainties, while highly empirical, nonetheless is an important step towards interpretation of experiments.

Predictions of nuclear properties far from measured data are inherently imprecise because of uncertainties in our knowledge of nuclear forces and in our treatment of quantum many-body effects in strongly-interacting systems. While the model bias can be directly calculated when experimental data is available, only an estimate can be made in the absence of such measurements. Current approaches to compute the estimated bias quickly lose predictive power when input variables such as proton or neutron number are extrapolated, resulting in uncontrolled uncertainties in applications such as nucleosynthesis simulations. In this letter, we present a novel technique to identify the input variables of machine learning algorithms that can provide robust estimates of model bias. Our process is based on selecting input variables, or features, based on their probability distribution functions across the entire nuclear chart. We illustrate our approach on the problem of quantifying the model bias in nuclear binding energies calculated with Density Functional Theory (DFT). We show that feature selection can systematically improve theoretical predictions without increasing uncertainties., Comment: Manuscript: 6 pages. Supplemental material: 3 pages

The Cr2Ge2Te6 (CGT) compound is an intrinsic ferromagnetic material with a van der Waals layered structure that shows great promise in spintronics applications. In this work, we investigated the edge effect in the formation of CGT nanoribbons with different terminations as well as the change in electronic and magnetic properties considering spin-orbit effects on the calculations. We studied the thermodynamic stability of the nanoribbon employing the surface formation energy formalism. According to the calculations, in regions of Ge-rich, Te-rich, and Ge-poor, Te-poor of chemical potential the nanoribbon with TeCr edges is the most stable structure, while in the regions of Ge-rich, Te-poor and Ge-poor, Te-poor the nanoribbon with TeCr edges and Te vacancies is the most stable structure. Furthermore, calculations show that the nanoribbon is ferromagnetic with half metal characteristics.

The NN scattering problem is usually analyzed in terms of partial waves and the corresponding coupled channel phase-shifts and mixing angles, but the available experiments induce correlations amongst the corresponding channels with different quantum numbers. Based on the Granada-2013 database we analyze the meaning and impact of those correlations taking into account both the purely statistical ones reflecting the primary experimental data uncertainties as well as the systematic ones exhibiting the ambiguities in the form of the potential representing the unknown nuclear force for distances below 3fm. We also ponder on the implications of fixing nuclear potentials by direct fits to the phase-shifts stemming from a full fledged partial wave analysis (PWA) inferred from experimental data., Comment: 10 pages, 3 figures

This paper introduces the PettingZoo library and the accompanying Agent Environment Cycle ("AEC") games model. PettingZoo is a library of diverse sets of multi-agent environments with a universal, elegant Python API. PettingZoo was developed with the goal of accelerating research in Multi-Agent Reinforcement Learning ("MARL"), by making work more interchangeable, accessible and reproducible akin to what OpenAI's Gym library did for single-agent reinforcement learning. PettingZoo's API, while inheriting many features of Gym, is unique amongst MARL APIs in that it's based around the novel AEC games model. We argue, in part through case studies on major problems in popular MARL environments, that the popular game models are poor conceptual models of games commonly used in MARL and accordingly can promote confusing bugs that are hard to detect, and that the AEC games model addresses these problems.

It is well known that Gaussian polynomials (i.e., $q$-binomials) describe the distribution of the $area$ statistic on monotone paths in a rectangular grid. We introduce two new statistics, $corners$ and $cindex$; attach ``ornaments'' to the grid; and re-evaluate these statistics, in order to argue that all scrambled versions of the $cindex$ statistic are equidistributed with $area$. Our main result is a representation of the generating function for the bi-statistic $(cindex,corners)$ as a two-variable Vandermonde convolution of the original Gaussian polynomial. The proof relies on explicit bijections between differently ornated paths., Comment: 2 figures

We investigate the flexibility of the entropy (topological and metric) for the class of piecewise expanding unimodal maps. We show that the only restrictions for the values of the topological and metric entropies in this class are that both are positive, the topological entropy is at most $\log 2$, and by the Variational Principle, the metric entropy is not larger than the topological entropy. In order to have a better control on the metric entropy, we work mainly with topologically mixing piecewise expanding skew tent maps, for which there are only 2 different slopes. For those maps, there is an additional restriction that the topological entropy is larger than $\frac{1}{2}\log2$. We also make the interesting observation that for skew tent maps the sum of reciprocals of derivatives of all iterates of the map at the critical value is zero. It is a generalization and a different interpretation of the Milnor-Thurston formula connecting the topological entropy and the kneading determinant for unimodal maps., Comment: 13 pages, 6 figures

Consider two objects associated to the Iterated Function System (IFS) $\{1+\lambda z,-1+\lambda z\}$: the locus $\mathcal{M}$ of parameters $\lambda\in\mathbb{D}\setminus\{0\}$ for which the corresponding attractor is connected; and the locus $\mathcal{M}_0$ of parameters for which the related attractor contains $0$. The set $\mathcal{M}$ can also be characterized as the locus of parameters for which the attractor of the IFS $\{1+\lambda z, \lambda z, -1+\lambda z\}$ contains $\lambda^{-1}$. Exploiting the asymptotic similarity of $\mathcal{M}$ and $\mathcal{M}_0$ with the respective associated attractors, we give sufficient conditions on $\lambda\in\partial\mathcal{M}$ or $\partial\mathcal{M}_0$ to guarantee it is path accessible from the complement $\mathbb{D}\setminus\mathcal{M}$., Comment: 22 pages, 6 figures. Added an Appendix and updated references. It will appear on Experimental Mathematics

This white paper reports on the discussions of the 2018 Facility for Rare Isotope Beams Theory Alliance (FRIB-TA) topical program "From bound states to the continuum: Connecting bound state calculations with scattering and reaction theory". One of the biggest and most important frontiers in nuclear theory today is to construct better and stronger bridges between bound state calculations and calculations in the continuum, especially scattering and reaction theory, as well as teasing out the influence of the continuum on states near threshold. This is particularly challenging as many-body structure calculations typically use a bound state basis, while reaction calculations more commonly utilize few-body continuum approaches. The many-body bound state and few-body continuum methods use different language and emphasize different properties. To build better foundations for these bridges, we present an overview of several bound state and continuum methods and, where possible, point to current and possible future connections., Comment: 48 pages, 23 figures; paper from the FRIB Theory Alliance Workshop on "From Bound States to the Continuum", June 11-22, 2018

Ab initio calculations in Nuclear physics for atomic nuclei require a specific knowledge of the interactions among their constituents, protons and neutrons. In particular, NN interactions can be constrained down to scale resolutions of $\Delta r \sim 0.6 {\rm fm}$ from the study of phase shifts below the pion production threshold. However, this allows for ambiguities and uncertainties which have an impact on finite nuclei, nuclear- and neutron-matter properties. On the other hand the nuclear many body problem is intrinsically difficult and the computational cost increases with numerical precision and number of nucleons. However, it is unclear what the physical precision should be for these calculations. In this contribution we review much of the work done in Granada to encompass both the uncertainties stemming from the NN scattering database in light nuclei such as triton and alpha particle and the numerical precision required by the solution method., Comment: 40 pages, 2 figures. Small modifications and Refs added. To appear in Frontiers in Physics (Research topic) The Long-Lasting Quest for Nuclear Interactions: The Past, the Present and the Future

Recent investigations have emphasized the importance of uncertainty quantification (UQ) to describe errors in nuclear theory. We carry out UQ for configuration-interaction shell model calculations in the $1s$-$0d$ valence space, investigating the sensitivity of observables to perturbations in the 66 parameters (matrix elements) of a high-quality empirical interaction. The large parameter space makes computing the corresponding Hessian numerically costly, so we compare a cost-effective approximation, using the Feynman-Hellmann theorem, to the full Hessian and find it works well. Diagonalizing the Hessian yields the principal components of the interaction: linear combinations of parameters ordered by sensitivity. This approximately decoupled distribution of parameters facilitates theoretical error propagation onto structure observables: electromagnetic transitions, Gamow-Teller decays, and dark matter-nucleus scattering matrix elements., Comment: 30 pages, 11 figures, 2 tables

Are chiral theories at present describing experimental NN scattering data satisfactorily ?. Will the chiral approach offer a framework where fitting and selecting the existing np and pp data can be done without theoretical bias ?. While predictive power in theoretical nuclear physics has been a major concern in the study of nuclear structure and reactions, the Effective Field Theory (EFT) based on chiral expansions has emerged after Weinberg as a model independent hierarchy for many body forces and much progress has been achieved over the last decades. We review some of the issues involved which point to being close to the solution, but also that work remains still to be done to validate the theory. We analyze several examples including zero energy NN scattering and perturbative counterterm -- free peripheral scattering where one would expect these methods to work best and unveil relevant systematic discrepancies when a fair comparison to the Granada-2013 NN-database and partial wave analysis (PWA) based on coarse graining the interaction is undertaken., Comment: Prepared for the special issue of "The tower of effective (field) theories and the emergence of nuclear phenomena"; 20 pages, 7 figures

We present the first example of a rigorous uncertainty quantification on elastic Nucleus-Nucleus scattering at energies near the Coulomb barrier. Experimental data has been analyzed using an energy dependent effective optical model potential with physical constraints imposed. We confirm the compatibility of these uncertainties with the well known Coulomb threshold anomaly, explained in terms of a dispersive relation, and contrast our results with previous analyses that suggest otherwise., Comment: 6 pages, 3 figures

The main purpose of scattering experiments is to unveil the underlying structure of the colliding particles and their interaction. Typically one measures scattering observables (cross sections and polarizations) at discrete angles and energies and mutually consistent data may validate or falsify proposed theories or models. However, the accumulation of data from different laboratories while potentially improves the statistical significance it may sometimes generate mutually inconsistent data as a side-effect. Thus, some decision has to be made on what are the maximal amount of data which are mutually compatible. We show elastic $\pi N $ and $NN$ scattering as prominent examples where this selection is called for. We discuss how it can be done in a self-consistent manner invoking a principle of maximal consensus of the database and with the help of a sufficiently flexible model involving a minimal number of theoretical assumptions. In the NN case this has become possible with a combination of long distance field theoretical constraints at the hadronic level such as pion exchanges and electromagnetic effects and a coarse graining of the unknown interaction over the shortest de Broglie wavelength being probed in the scattering process., Comment: 7 pages, 5 figures, talk presented by ERA at XVII International Conference on Hadron Spectroscopy and Structure - Hadron2017, Salamanca, 25-29 September 2017

We present a minimal nuclear energy density functional (NEDF) called "SeaLL1" that has the smallest number of possible phenomenological parameters to date. SeaLL1 is defined by 7 significant phenomenological parameters, each related to a specific nuclear property. It describes the nuclear masses of even-even nuclei with a mean energy error of 0.97 MeV and a standard deviation 1.46 MeV, two-neutron and two-proton separation energies with r.m.s. errors of 0.69 MeV and 0.59 MeV respectively, and the charge radii of 345 even-even nuclei with a mean error $\epsilon_r=$0.022 fm and a standard deviation $\sigma_r=$0.025 fm. SeaLL1 incorporates constraints on the EoS of pure neutron matter from quantum Monte Carlo calculations with chiral effective field theory two-body (NN) interactions at N3LO level and three-body (NNN) interactions at the N2LO level. Two of the seven parameters are related to the saturation density and the energy per particle of the homogeneous symmetric nuclear matter, one is related to the nuclear surface tension, two are related to the symmetry energy and its density dependence, one is related to the strength of the spin-orbit interaction, and one is the coupling constant of the pairing interaction. We identify additional phenomenological parameters that have little effect on ground-state properties, but can be used to fine-tune features such as the Thomas-Reiche-Kuhn sum rule, the excitation energy of the giant dipole and Gamow-Teller resonances, the static dipole electric polarizability, and the neutron skin thickness., Comment: Updated to match published version. This article supersedes arXiv:1506.09195. Fit data provided in source files *.txt. (35 pages, 23 figures.)