SearcharxivSearch

arXiv subjects

Long Meng

Publications and source records attributed to Long Meng.

At least 19 recordsLinked to original sources

Approximate eigenfunctions for some aperiodic crystals

In this paper, we consider Hamiltonians for aperiodic crystals of the form \begin{align*} H_\varepsilon:=T(-i\nabla_x+{\mathbf A}(x,\varepsilon x))+V(x,\varepsilon x),\qquad x\in {\mathbb R}^d \end{align*} where $T$ represents either a Dirac operators or a Schr\"odinger operator, and $x\mapsto {\mathbf A}(x,X)$ and $x\mapsto V(x,X)$ are $\mathbb L$-periodic with respect to some lattice $\mathbb L\subset{\mathbb R}^d$. Let \begin{align*} (k,X)\ni {\mathbb R}^d\times {\mathbb R}^d\mapsto h(k,X):=T(-i\nabla_x+k+{\mathbf A}(x,X))+V(x,X) \end{align*} be a family of operators acting on $L^2_{\rm per}(\mathbb{R}^d/\mathbb{L})$ with periodic boundary conditions. We show that, under some suitable assumptions on the family of operators $ (h(k,X))_{k,X}$ around an energy level $e_0\in {\mathbb R}$ and some points $(k_0,X_0)\in {\mathbb R}^d\times {\mathbb R}^d$, one can construct localized approximate eigenfunctions $\Phi_\varepsilon\in L^2({\mathbb R}^d)$ of the operator $H_\varepsilon$ such that for $\varepsilon$ small enough and for some $m\in \{1,2\}$ and $\mu\in {\mathbb R}$, \begin{align}\label{eq:abstract} \|(H_\varepsilon-e_0-\varepsilon^{\frac{m}{2}}\mu)\Phi_\varepsilon\|_{L^2({\mathbb R}^d)}={\mathcal O}(\varepsilon^{\frac{m}{2}+\frac{1}{4}}). \end{align} with \begin{align*} \|\Phi_\varepsilon\|_{L^2({\mathbb R}^d)}=\frac{1}{|{\mathbb R}^d/\mathbb L|^{1/2}}+{\mathcal O}(\sqrt{\varepsilon}). \end{align*}

math-ph

Eigenvalue asymptotics of M\"uller minimizers for atoms and molecules

We study the spectral properties of minimizers of the M\"uller functional for atoms and molecules with $N$ electrons and total nuclear charge $Z$. We prove that under some suitable assumptions on $Z$ and $N$, the $k$-th eigenvalue of a M\"uller minimizer $\gamma_*$ behaves as $A_* k^{-8/3}$ when $k\to \infty$, with a constant $A_*>0$ determined explicitly by the density of $\gamma_*$. In particular, in the atomic case $V=Z|x|^{-1}$ our assumption holds if $Z$ is sufficiently large and $N\le Z- C_0 Z^{1/3}$. While our proof is inspired by Sobolev's work on the asymptotic behavior of the one-particle density matrix of Schr\"odinger ground states, the analysis in M\"uller theory requires several new ingredients concerning both the singular behavior of the integral kernel of the minimizers near the diagonal and the decay properties at infinity.

math-ph

Numerical computation of the density of states of aperiodic multiscale Schr\"odinger operators

Computing the electronic structure of incommensurate materials is a central challenge in condensed matter physics, requiring efficient ways to approximate spectral quantities such as the density of states (DoS). In this paper, we numerically investigate two distinct approaches for approximating the DoS of incommensurate Hamiltonians for small values of the incommensurability parameters $\epsilon$ (e.g., small twist angle, or small lattice mismatch): the first employs a momentum-space decomposition, and the second exploits a semiclassical expansion with respect to $\epsilon$. In particular, we compare these two methods using a 1D toy model. We check their consistency by comparing the asymptotic expansion terms of the DoS, and it is shown that, for full DoS, the two methods exhibit good agreement in the small $\epsilon$ limit, while discrepancies arise for less small $\epsilon$, which indicates the importance of higher-order corrections in the semiclassical method for such regimes. We find these discrepancies to be caused by oscillations in the DoS at the semiclassical analogues of Van Hove singularities, which can be explained qualitatively, and quantitatively for $\epsilon$ small enough, by a semiclassical approach.

math-ph

Mixed regularity and sparse grid approximations of $N$-body Schr\"odinger evolution equation

In this paper, we present a mathematical analysis of time-dependent $N$-body electronic systems and establish mixed regularity for the corresponding wavefunctions. Based on this, we develop sparse grid approximations to reduce computational complexity, including a sparse grid Gaussian-type orbital (GTO) scheme. We validate the approach on the Helium atom (${\rm He}$) and Hydrogen molecule (${\rm H}_2$), showing that sparse grid GTOs offer an efficient alternative to full grid discretizations.

math.NA

Reduced Muscle Fatigue Using Continuous Subthreshold Kilohertz Stimulation of Peripheral Nerves

Functional electrical stimulation (FES) is a prevalent technique commonly used to activate muscles in individuals with neurological disorders. Traditional FES strategies predominantly utilize low-frequency (LF) stimulation, which evokes synchronous action potentials, leading to rapid muscle fatigue. To address these limitations, we introduced a subthreshold high-frequency (HF) stimulation method that employed continuous, charge-balanced subthreshold current pulses at kilohertz frequencies, designed to evoke motor unit (MU) activation similar to voluntary activation. We evaluated the effectiveness of HF stimulation on the reduction of muscle fatigue across different force levels (10 %, 25 %, and 40 % of maximum force). The HF stimulation utilized continuous charge-balanced, short pulses of 80 {\mu}s (at a 10 kHz frequency) targeted the ulnar/median nerve bundles. We compared the fatigue effects with conventional LF stimulation and voluntary muscle contractions. Our results indicated that HF stimulation maintained more sustained force outputs and muscle activation over a prolonged time compared with LF stimulation. The HF stimulation also evoked a more dispersed muscle activation pattern, similar to voluntary muscle contractions. These findings suggest that HF stimulation can significantly enhance the sustainability of muscle contractions and reduce muscle fatigue, potentially improving the efficacy and applicability of FES in clinical and home-based settings for individuals with neurological impairments.

eess.SY

On the relativistic effect in the Dirac--Fock theory

In this paper, we study the error bound between the Dirac--Fock ground-state energy and the Hartree--Fock ground-state energy, a quantity known as the relativistic effect in quantum mechanics. We confirm that the relativistic effect in the Dirac--Fock ground-state energy is of the order $\mathcal{O}(c^{-2})$ with $c$ being the speed of light. Furthermore, if the potential between electrons and nuclei is regular, we get the well-known leading order relativistic correction -- the Breit--Pauli term, which is the sum of the mass-velocity term, the Darwin term, and the spin-orbit term. As a consequence, we also show that the same relativistic effects and leading order relativistic correction also hold in a QED model introduced by Mittleman when the vacuum polarization -- a term of the order $\mathcal{O}(c^{-3})$ -- is neglected. To our knowledge, this is the first time in mathematics that the leading-order relativistic correction has been obtained from nonlinear Dirac ground-state energy problems.

math-ph

Enhancing Speech Large Language Models with Prompt-Aware Mixture of Audio Encoders

Connecting audio encoders with large language models (LLMs) allows the LLM to perform various audio understanding tasks, such as automatic speech recognition (ASR) and audio captioning (AC). Most research focuses on training an adapter layer to generate a unified audio feature for the LLM. However, different tasks may require distinct features that emphasize either semantic or acoustic aspects, making task-specific audio features more desirable. In this paper, we propose Prompt-aware Mixture (PaM) to enhance the Speech LLM that uses multiple audio encoders. Our approach involves using different experts to extract different features based on the prompt that indicates different tasks. Experiments demonstrate that with PaM, only one Speech LLM surpasses the best performances achieved by all single-encoder Speech LLMs on ASR, Speaker Number Verification, and AC tasks. PaM also outperforms other feature fusion baselines, such as concatenation and averaging. Our code would be available at: https://github.com/shanweiqiao/PaM

eess.AS

Early Exit Is a Natural Capability in Transformer-based Models: An Empirical Study on Early Exit without Joint Optimization

Large language models (LLMs) exhibit exceptional performance across various downstream tasks. However, they encounter limitations due to slow inference speeds stemming from their extensive parameters. The early exit (EE) is an approach that aims to accelerate auto-regressive decoding. EE generates outputs from intermediate layers instead of using the whole model, which offers a promising solution to this challenge. However, additional output layers and joint optimization used in conventional EE hinder the application of EE in LLMs. In this paper, we explore the possibility of LLMs EE without additional output layers and joint optimization. Our findings indicate that EE is a natural capability within transformer-based models. While joint optimization does not give model EE capability, it must be employed to address challenges by improving the accuracy of locating the optimal EE layer through gating functions. Additionally, our study reveals patterns in EE behavior from a sub-word perspective based on the LLaMA model and the potential possibility for EE based on sub-layers.

cs.CL

Massless Dirac equation on spinor bundles over real hyperbolic spaces

We prove a sharp-in-time dispersive estimate of the Dirac equation on spinor bundles over the real hyperbolic space. Compared with the Euclidean counterparts, our result shows that the dispersive estimate differs between short and long times, reflecting the intuitive influence of negative curvature on the dispersion. Moreover, the well-known equivalence between dispersive estimates for Dirac and wave propagators in the Euclidean setting no longer holds in this context. This finding suggests that spinor fields are affected by the geometry at infinity of the manifold. As a key application, we establish an improved global-in-time Strichartz estimate, in the sense that there is no loss of angular derivatives and the admissible set is larger than previously known results in other settings.

math.AP

The Ground State Energy of Heavy Atoms: Leading and Subleading Asymptotics

We study atomic ground state energies for neutral atoms as the nuclear charge $Z$ is large in the no-pair formalism. We show that for a large class of projections defining the underlying Dirac sea -- covering not only the physical reasonable cases but also ``weird'' ones -- the corresponding no-pair ground state energy does not exceed the one of the Furry energy up to subleading order. An essential tool is the use and extension of S\'er\'e's results on atomic Dirac-Hartree-Fock theory.

math-ph

Relativistic Exchange Bounds

We collect estimates of the exchange energy of the relativistic no-pair Hartree-Fock and M\"uller functional and use them to show the existence of a minimizer and stability of matter of the relativistic M\"uller functional in the free picture.

math-ph

Semiclassical analysis of two-scale electronic Hamiltonians for twisted bilayer graphene

This paper investigates the mathematical properties of independent-electron models for twisted bilayer graphene by examining the density-of-states of corresponding single-particle Hamiltonians using tools from semiclassical analysis. This study focuses on a specific atomic-scale Hamiltonian $H_{d,\theta}$ constructed from Density-Functional Theory, and a family of moir\'e-scale Hamiltonians $H_{d,K,\theta}^{\rm eff}$ containing the Bistritzer-MacDonald model. The parameter $d$ represents the interlayer distance, and $\theta$ the twist angle. It is shown that the density-of-states of $H_{d,\theta}$ and $H_{d,K,\theta}^{\rm eff}$ admit asymptotic expansions in the twist angle parameter $\epsilon:=\sin(\theta/2)$. The proof relies on a twisted version of the Weyl calculus and a trace formula for an exotic class of pseudodifferential operators suitable for the study of twisted 2D materials. We also show that the density-of-states of $H_{d,\theta}$ admits an asymptotic expansion in $\eta:=\tan(\theta/2)$ and comment on the differences between the expansions in $\epsilon$ and $\eta$.

math-ph

Properties of periodic Dirac--Fock functional and minimizers

Existence of minimizers for the Dirac--Fock model in crystals was recently proved by Paturel and S\'er\'e and the authors \cite{crystals} by a retraction technique due to S\'er\'e \cite{Ser09}. In this paper, inspired by Ghimenti and Lewin's result \cite{ghimenti2009properties} for the periodic Hartree--Fock model, we prove that the Fermi level of any periodic Dirac--Fock minimizer is either empty or totally filled when $\frac{\alpha}{c}\leq C_{\rm cri}$ and $\alpha>0$. Here $c$ is the speed of light, $\alpha$ is the fine structure constant, and $C_{\rm cri}$ is a constant only depending on the number of electrons and on the charge of nuclei per cell. More importantly, we provide an explicit upper bound for $C_{\rm cri}$. Our result implies that any minimizer of the periodic Dirac--Fock model is a projector when $\frac{\alpha}{c}\leq C_{\rm cri}$ and $\alpha>0$. In particular, the non-relativistic regime (i.e., $c\gg1$) and the weak coupling regime (i.e., $0<\alpha\ll1$) are covered. The proof is based on a delicate study of a second-order expansion of the periodic Dirac--Fock functional composed with the retraction used in \cite{crystals}.

math.AP

Strichartz estimates for the half wave/Klein-Gordon and Dirac Equations on compact manifolds without boundary

In this paper we study Strichartz estimates for the half wave, the half Klein-Gordon and the Dirac Equations on compact manifolds without boundary, proving in particular for each of these flows local in time estimates both for the wave and Schroedinger admissible couples (in this latter case with an additional loss of regularity). The strategy for the proof is based on a refined version of the WKB approximation.

math.AP

A rigorous justification of the Mittleman's approach to the Dirac--Fock model

In this paper, we study the relationship between the Dirac--Fock model and the electron-positron Hartree--Fock model. We justify the Dirac--Fock model as a variational approximation of QED when the vacuum polarization is neglected and when the fine structure constant $\alpha$ is small and the velocity of light $c$ is large. As a byproduct, we also prove, when $\alpha$ is small or $c$ is large, the no-unfilled shells theory in the Dirac--Fock theory for atoms and molecules. The proof is based on some new properties of the Dirac--Fock model.

math.AP

Existence of minimizers for the Dirac-Fock model of crystals

Whereas many different models exist in the mathematical and physics literature for ground states of non-relativistic crystals, the relativistic case has been much less studied and we are not aware of any mathematical result on a fully relativistic treatment of crystals. In this paper, we introduce a mean-field relativistic energy for crystals in terms of periodic density matrices. This model is inspired both from a recent definition of the Dirac-Fock ground state for atoms and molecules, due to one of us, and from the non-relativistic Hartree-Fock model for crystals. We prove the existence of a ground state when the number of electrons per cell is not too large.

math.AP

Strichartz estimates for the Dirac equation on asymptotically flat manifolds

In this paper we prove Strichartz estimates for the Dirac equation on asymptotically flat manifolds. The proof combines the weak dispersive estimates proved by the first two authors with the Strichartz and smoothing estimates for the wave and Klein-Gordon flows, exploiting some recent results in the same geometrical setting.

math.AP

Large time well posedness for a Dirac--Klein-Gordon system

In this paper we prove well posedness for a system coupling a nonlinear Dirac with a Klein-Gordon equation that represents a toy model for the Helium atom with relativistic corrections: the wave function of the electrons interacts with an electric field generated by a nucleus with a given charge density. One of the main ingredients we need is a new family of Strichartz estimates for time dependent perturbations of the Dirac equation: these represent a result of independent interest.

math.AP