Newsfeeds

Journal of Physics A: Mathematical and Theoretical - latest papers

Latest articles for Journal of Physics A: Mathematical and Theoretical

IOPscience

  • Bifurcation analysis of figure-eight choreography in the three-body problem based on crystallographic point groups
    The bifurcation of figure-eight choreography is analyzed by its symmetry group based on the variational principle of the action. The irreducible representations determine the symmetry and the dimension of the Lyapunov–Schmidt reduced action, which yields four types of bifurcations in the sequence of the bifurcation cascade. Type 1 bifurcation, represented by trivial representation, bifurcates two solutions. Type 2, by non-trivial one-dimensional representation, bifurcates two congruent solutions. Type 3 and 4, by two-dimensional irreducible representations, bifurcate two sets of three and six congruent solutions, respectively. We analyze numerical bifurcation solutions previously published and four new ones: non-symmetric choreographic solution of type 2, non-planar solution of type 2, y-axis symmetric solution of type 3, and non-symmetric solution of type 4.

  • Wigner current in multidimensional quantum billiards
    In the present paper we derive the Wigner current of the particle in a multidimensional billiard—the compact region of space in which the particle moves freely. The calculation is based on proposed by us previously method of imposing boundary conditions by convolution of the free particle Wigner function with some time independent function, defined by the shape of the billiard. This method allowed to greatly simplify the general expression for the Wigner current, representing its p–component as a surface integral of the product of the shifted particles wave functions. The results are also connected to an alternative approach, which takes into account the boundary conditions by adding the term to the Hamiltonian. The latter is also generalized to the multidimensional case.

  • On the maximally symmetric vacua of generic Lovelock gravities *
    We survey elementary features of Lovelock gravity and its maximally symmetric vacuum solutions. The latter is solely determined by the real roots of a dimension-dependent polynomial. We also recover the static spherically symmetric (black hole) solutions of Lovelock gravity using Palais’ symmetric criticality principle. We show how to linearize the generic field equations of Lovelock models about a given maximally symmetric vacuum, which turns out to factorize into the product of yet another dimension-dependent polynomial and the linearized Einstein tensor about the relevant background. We also describe how to compute conserved charges using linearized field equations along with the relevant background Killing isometries. We further describe and discuss the special vacua which are defined by the simultaneous vanishing of the aforementioned polynomials.

  • Spectral analysis of metamaterials in curved manifolds
    Negative-index metamaterials possess a negative refractive index and thus present an interesting substance for designing uncommon optical effects such as invisibility cloaking. This paper deals with operators encountered in an operator-theoretic description of metamaterials. First, we introduce an indefinite Laplacian and consider it on a compact tubular neighbourhood in constantly curved compact two-dimensional Riemannian ambient manifolds, with Euclidean rectangle in being present as a special case. As this operator is not semi-bounded, standard form-theoretic methods cannot be applied. We show that this operator is (essentially) self-adjoint via separation of variables and find its spectral characteristics. We also provide a new method for obtaining alternative definition of the self-adjoint operator in non-critical case via a generalized form representation theorem. The main motivation is existence of essential spectrum in bounded domains.

  • Dissecting a small artificial neural network
    We investigate the loss landscape and backpropagation dynamics of convergence for the simplest possible artificial neural network representing the logical exclusive-OR gate. Cross-sections of the loss landscape in the nine-dimensional parameter space are found to exhibit distinct features, which help understand why backpropagation efficiently achieves convergence toward zero loss, whereas values of weights and biases keep drifting. Differences in shapes of cross-sections obtained by nonrandomized and randomized batches are discussed. In reference to statistical physics we introduce the microcanonical entropy as a unique quantity that allows to characterize the phase behavior of the network. Learning in neural networks can thus be thought of as an annealing process that experiences the analogue of phase transitions known from thermodynamic systems. It also reveals how the loss landscape simplifies as more hidden neurons are added to the network, eliminating entropic barriers caused by finite-size effects.