arXiv subjects

J. W. Clark

Publications and source records attributed to J. W. Clark.

At least 73 records · Page 4Linked to original sources

arXiv2005

Thermodynamic properties of Fermi systems with flat single-particle spectra

The behavior of strongly correlated Fermi systems is investigated beyond the onset of a phase transition where the single-particle spectrum $ξ({\bf p})$ becomes flat. The Landau-Migdal quasiparticle picture is shown to remain applicable on the ordered side of this transition. Nevertheless, low-temperature properties evaluated within this picture show profound changes relative to results of Landau theory, as a direct consequence of the flattening of $ξ({\bf p})$. Stability conditions for this class of systems are examined, and the nature of antiferromagnetic quantum phase transitions is elucidated.

cond-mat.str-el↗

arXiv2004

Anomalous Behavior of Thermodynamic Properties of Strongly Correlated Fermi Systems

Thermodynamic characteristics of Fermi systems are investigated in the vicinity of a phase transition where the effective mass diverges and the single-particle spectrum becomes flat. It is demonstrated that at extremely low temperatures $T$, the flattening of the spectrum is reflected in non-Fermi-liquid behavior of the inverse susceptibility $χ^{-1}(T) \sim T^α$ and the specific heat $C(T)/T\sim T^{-α}$, with the critical index $α=2/3$. In the presence of an external static magnetic field $H$, both these quantities are found to exhibit a scaling behavior, e.g. $χ^{-1}(T,H)=χ^{-1}(T,0)+ T^{2/3}F(H/T)$, in agreement with available experimental data.

cond-mat.str-el↗

arXiv2004

Nuclear mass systematics using neural networks

New global statistical models of nuclidic (atomic) masses based on multilayered feedforward networks are developed. One goal of such studies is to determine how well the existing data, and only the data, determines the mapping from the proton and neutron numbers to the mass of the nuclear ground state. Another is to provide reliable predictive models that can be used to forecast mass values away from the valley of stability. Our study focuses mainly on the former goal and achieves substantial improvement over previous neural-network models of the mass table by using improved schemes for coding and training. The results suggest that with further development this approach may provide a valuable complement to conventional global models.

nucl-th↗

arXiv2004

Phase Transitions in Nucleonic Matter and Neutron-Star Cooling

A new scenario for neutron-star cooling is proposed, based on the correspondence between pion condensation, occurring in neutron matter due to critical spin-isospin fluctuations, and the metal-insulator phase transition in a two-dimensional electron gas. Beyond the threshold density for pion condensation, where neutron-star matter loses its spatial homogeneity, the neutron single-particle spectrum acquires an insulating gap that quenches neutron contributions to neutrino-production reactions and to the star's specific heat. In the liquid phase at densities below the transition point, spin-isospin fluctuations are found to play dual roles. On the one hand, they lead to a multi-sheeted neutron Fermi surface that extends to low momenta, thereby activating the normally forbidden direct-Urca cooling mechanism; on the other, they amplify the nodeless $P$-wave neutron superfluid gap while suppressing $S$-wave pairing. In this picture, lighter stars without a pion-condensed core experience slow cooling, while enhanced cooling occurs in heavier stars through direct-Urca emission from a narrow shell of the interior.

astro-ph↗

arXiv2004

Neural Networks Processing Mean Values of Random Variables

We introduce a class of neural networks derived from probabilistic models in the form of Bayesian belief networks. By imposing additional assumptions about the nature of the probabilistic models represented in the belief networks, we derive neural networks with standard dynamics that require no training to determine the synaptic weights, that can pool multiple sources of evidence, and that deal cleanly and consistently with inconsistent or contradictory evidence. The presented neural networks capture many properties of Bayesian belief networks, providing distributed versions of probabilistic models.

cond-mat.dis-nn↗

arXiv2004

Anomalous Behavior of the Spin Susceptibility of Strongly Correlated Fermi Systems

The spin susceptibility $χ(T)$ of strongly correlated Fermi systems is investigated in the density region where Fermi-liquid theory fails. We attribute this failure to a specific quantum phase transition associated with a rearrangement of the Landau state at low temperatures $T$, retaining the assumption that the Landau quasiparticle picture survives in a generic sense. Taking into account the resulting modification of the quasiparticle istribution function, the spin susceptibility $χ(T)$ is shown to contain a Curie-Weiss component $χ_{\rm CW}(T)\sim (T-Θ_{\rm W})^{-1}$, with the Weiss temperature $Θ_{\rm W}$ vanishing at the critical density for the transition.

cond-mat.str-el↗

arXiv2004

Mechanisms driving alteration of the Landau state in the vicinity of a second-order phase transition

The rearrangement of the Fermi surface of a homogeneous Fermi system upon approach to a second-order phase transition is studied at zero temperature. The analysis begins with an investigation of solutions of the equation $ε(p)=μ$, a condition that ordinarily has the Fermi momentum $p_F$ as a single root. The emergence of a bifurcation point in this equation is found to trigger a qualitative alteration of the Landau state, well before the collapse of the collective degree of freedom that is responsible for the second-order transition. The competition between mechanisms that drive rearrangement of the Landau quasiparticle distribution is explored, taking into account the feedback of the rearrangement on the spectrum of critical fluctuations. It is demonstrated that the transformation of the Landau state to a new ground state may be viewed as a first-order phase transition.

cond-mat.str-el↗

arXiv2003

Kirkwood Phase Transition for Boson and Fermion Hard-Sphere Systems

The London ground-state energy formula as a function of number density $ρ$ for a system of boson hard spheres of diameter $c$ at zero temperature (corrected for the reduced mass of a pair of particles in a ``sphere-of-influence'' picture) generalized to describe fermion hard-sphere systems with four and two intrinsic degrees of freedom such as helium-three or neutron matter and symmetric nuclear matter, respectively, is proposed as the crystalline energy branch for hard-sphere systems. For the fluid branch we use the well-known, exact, low-density equation-of-state expansions for many-boson and many-fermion systems, appropriately extrapolated to physical densities. Here, via a double-tangent construction the crystallization and melting densities for boson and fermion hard spheres are determined. They agree well with variational Monte Carlo, density-functional, and Green Function Monte Carlo calculations.

cond-mat.stat-mech↗

arXiv2003

$^3P_2$-$^3F_2$ Pairing in Dense Neutron Matter: The Spectrum of Solutions

The $^3P_2$-$^3F_2$ pairing model is generally considered to provide an adequate description of the superfluid states of neutron matter at densities some 2-3 times that of saturated symmetrical nuclear matter. The problem of solving the system of BCS gap equations expressing the $^3P_2$-$^3F_2$ model is attacked with the aid of the separation approach. This method, developed originally for quantitative study of S-wave pairing in the presence of strong short-range repulsions, serves effectively to reduce the coupled, singular, nonlinear BCS integral equations to a set of coupled algebraic equations. For the first time, sufficient precision becomes accessible to resolve small energy splittings between the different pairing states. Adopting a perturbative strategy, we are able to identify and characterize the full repertoire of real solutions of the $^3P_2$-$^3F_2$ pairing model, in the limiting regime of small tensor-coupling strength. The P-F channel coupling is seen to lift the striking parametric degeneracies revealed by a earlier separation treatment of the pure, uncoupled $^3P_2$ pairing problem. Remarkably, incisive and robust results are obtained solely on the basis of analytic arguments. Unlike the traditional Ginzburg-Landau approach, the analysis is not restricted to the immediate vicinity of the critical temperature, but is equally reliable at zero temperature. Interesting connections and contrasts are drawn between triplet pairing in dense neutron matter and triplet pairing in liquid $^3$He.

nucl-th↗

arXiv2002

Superfluid Phases of Triplet Pairing

The phase transitions in dense neutron matter with triplet pairing have been investigated. The spectrum of phases of the $^3P_2$-$^3F_2$ model, which adequately describes pairing in this system, is identified and classified with the aid of a separation method for solving BCS gap equations associated with pairing states of arbitrary angular-momentum content. In addition to usual solutions involving a single value of the magnetic quantum number (and its negative), there exist ten real multicomponent solutions. Five of the corresponding angle-dependent order parameters have nodes, and five do not. In contrast to the case of superfluid $^3$He, transitions occur between phases with nodeless order parameters. Results for the temperature dependence of the competition between the various phases have been obtained.

nucl-th↗

arXiv2002

Final-state interactions in the response of nuclear matter

Final-state interactions in the response of a many-body system to an external probe delivering large momentum are normally described using the eikonal approximation, for the trajectory of the struck particle, and the frozen approximation, for the positions of the spectators. We propose a generalization of this scheme, in which the initial momentum of the struck particle is explicitly taken into account. Numerical calculations of the nuclear matter response at 1 $< |{\bf q}| <$ 2 GeV/c show that the inclusion of this momentum dependence leads to a sizable effect in the low energy tail. Possible implications for the analysis of existing electron-nucleus scattering data are discussed.

nucl-th↗

arXiv2001

Non-BCS pairing in anisotropic strongly correlated electron systems in solids

The problem of pairing in anisotropic electron systems possessing patches of fermion condensate in the vicinity of the van Hove points is analyzed. Attention is directed to opportunities for the occurrence of non-BCS pairing correlations between the states belonging to the fermion condensate. It is shown that the physical emergence of such pairing correlations would drastically alter the behavior of the single-particle Green function, the canonical pole of Fermi-liquid theory being replaced by a branch point.

cond-mat.str-el↗

arXiv2001

Statistical Modeling of Nuclear Systematics

Statistical modeling of data sets by neural-network techniques is offered as an alternative to traditional semiempirical approaches to global modeling of nuclear properties. New results are presented to support the position that such novel techniques can rival conventional theory in predictive power, if not in economy of description. Examples include the statistical inference of atomic masses and beta-decay halflives based on the information contained in existing databases. Neural network modeling, as well as other statistical strategies based on new algorithms for artificial intelligence, may prove to be a useful asset in the further exploration of nuclear phenomena far from stability.

nucl-th↗

arXiv2001

Abnormal Occupation REvisited

It is demonstrated the Fermi surface of dense neutron matter may experience a rearrangement near the onset of pion condensation, due to strong momentum dependence of the effective interaction induced by spin-isospin fluctuations. In particular, the Fermi surface may take the form of a partially hollow sphere having a spherical hole in its center (a ``Lifschitz bubble''). Thus, a second (inner) Fermi surface may form as high-momentum single-particle states are filled and low-momentum states are vacated. The influence of this phenomenon on the superfluid transition temperature of the Fermi system is characterized with the help of a separation transformation of the BCS gap equation. This work may be viewed as a revival of the search for physical realizations of abnormal occupation in infinite, homogeneous Fermi systems -- plausible instances in which the quasiparticle distribution differs from that of an ideal Fermi gas and Fermi liquid theory breaks down.

cond-mat↗

arXiv2001

Bose-Einstein Condensation and The Lambda Transition in Liquid Helium

Integrating seminal ideas of London, Feynman, and Feenberg, this paper continues the development of an ab initio theory of the lambda transition in liquid 4He. The theory is based on variational determination of a trial density matrix suitable for description of both normal and superfluid phases, which includes the effects both of temperature-dependent dynamical correlations and of statistical correlations corresponding to thermal phonon-roton and quasiparticle/quasihole excitations - all at the level of two-point descriptors. A full functional minimization of the associated free energy is performed through solution of a coupled set of Euler-Lagrange equations. Violation of particle-hole exchange symmetry emerges as a crucial aspect of the lambda transition, along with broken gauge symmetry. In conjunction with a semi-phenomenological study in which renormalized effective masses are introduced for quasiparticle/quasihole excitations, the results suggest that a quantitative description can be achieved once the trial density matrix is modified to include backflow effects and allow for ab initio treatment of important variations in effective masses.

cond-mat↗

arXiv2001

Higher-Order Neutral Networks, Polya Polynomials, and Fermi Cluster Diagrams

The problem of controlling higher-order interactions in neural networks is addressed with techniques commonly applied in the cluster analysis of quantum many-particle systems. For multi-neuron synaptic weights chosen according to a straightforward extension of the standard Hebbian learning rule, we show that higher-order contributions to the stimulus felt by a given neuron can be readily evaluated via Polyà's combinatoric group-theoretical approach or equivalently by exploiting a precise formal analogy with fermion diagrammatics.

cond-mat.dis-nn↗

arXiv2001

Neural Propagation of Beliefs

We continue to explore the hypothesis that neuronal populations represent and process analog variables in terms of probability density functions (PDFs). A neural assembly encoding the joint probability density over relevant analog variables can in principle answer any meaningful question about these variables by implementing the Bayesian rules of inference. Aided by an intermediate representation of the probability density based on orthogonal functions spanning an underlying low-dimensional function space, we show how neural circuits may be generated from Bayesian belief networks. The ideas and the formalism of this PDF approach are illustrated and tested with several elementary examples, and in particular through a problem in which model-driven top-down information flow influences the processing of bottom-up sensory input.

cond-mat.dis-nn↗

arXiv2001

Neural Representation of Probabilistic Information

It has been proposed that populations of neurons process information in terms of probability density functions (PDFs) of analog variables. Such analog variables range, for example, from target luminance and depth on the sensory interface to eye position and joint angles on the motor output side. The requirement that analog variables must be processed leads inevitably to a probabilistic description, while the limited precision and lifetime of the neuronal processing units leads naturally to a population representation of information. We show how a time-dependent probability density, residing in a specified function space of dimension D, may be decoded from the neuronal activities in a population as a linear combination of certain decoding functions, with coefficients given by the N firing rates (generally with D << N). We show how the neuronal encoding process may be described by projecting a set of complementary encoding functions on the probability density, and passing the result through a rectifying nonlinear activation function. We show how both encoders and decoders may be determined by minimizing cost functions that quantify the inaccuracy of the representation. Expressing a given computation in terms of manipulation and transformation of probabilities, we show how this representation leads to a neural circuit that can carry out the required computation within a consistent Bayesian framework, with the synaptic weights being explicitly generated in terms of encoders, decoders, conditional probabilities, and priors.

cond-mat.dis-nn↗