Fact-checked by Grok 2 weeks ago

Spectral gap

The is a fundamental quantity in , referring to the separation between an isolated eigenvalue (or part of the spectrum) and the remainder of the spectrum of a , often specifically the difference between the leading extremal eigenvalues. This concept arises prominently in diverse fields, including , where it characterizes the decay rates of semigroups generated by the operator, and in applied contexts like , probability, and . The presence and size of a provide insights into the structural properties and dynamical behavior of the underlying system, such as , to , or stability. In , the spectral gap typically denotes the second-smallest eigenvalue \lambda_2 of the graph Laplacian matrix L = D - A (where D is the and A the ), since the smallest eigenvalue \lambda_1 = 0 corresponds to constant eigenvectors for connected graphs. A larger \lambda_2 indicates stronger connectivity and expansion properties of the graph, linking to the Cheeger constant, which bounds the edge expansion and is crucial for algorithms in network analysis, random walks, and . For regular graphs, it is often expressed as d - \lambda_2 for the adjacency matrix eigenvalues, where d is the degree, highlighting expander graphs with robust mixing and . In the study of Markov chains, the spectral gap is defined as $1 - |\lambda_2|, where |\lambda_2| is the largest modulus among the eigenvalues of the transition matrix P strictly less than 1, for a reversible, ergodic chain, measuring the rate at which the chain converges to its stationary distribution \pi. This gap governs the exponential decay of total variation distance to equilibrium, with mixing time \tau(\epsilon) bounded above by \frac{1}{1 - |\lambda_2|} \ln\left(\frac{1}{\pi_{\min} \epsilon}\right), where \pi_{\min} is the minimum stationary probability; larger gaps imply faster mixing, essential for Monte Carlo simulations and sampling algorithms. Equivalently, it can be expressed via the Dirichlet form as \lambda = \min \frac{\mathcal{E}(f,f)}{\mathrm{Var}_\pi(f)} for non-constant functions f, connecting to conductance and bottleneck ratios in chain geometry. In quantum many-body physics, the spectral gap \Delta is the energy difference between the ground-state eigenvalue E_0 and the first excited-state eigenvalue E_1 of a H, i.e., \Delta = E_1 - E_0 > 0. A gapped implies stability of the against perturbations, enabling phenomena like and protection against decoherence in ; for instance, an inverse-polynomial gap \Delta = \Omega(1/\mathrm{poly}(n)) reduces the complexity of ground-state energy estimation from to PP-complete in certain precision regimes. This gap is pivotal in condensed matter systems, distinguishing insulators (gapped) from metals (gapless) and influencing phase transitions.

Definition

General Concept

In , the spectrum of a bounded linear T on a is defined as the set of all complex numbers \lambda such that T - \lambda I is not invertible. For operators on s or symmetric matrices, the spectrum consists entirely of real eigenvalues. The is a fundamental concept in , referring to the distance between an isolated part of the spectrum (often an extremal eigenvalue) and the remainder of the spectrum of the operator. In , for a on a , it is the length of an open in the real line containing no points of the spectrum except possibly at the endpoints. Intuitively, a larger spectral gap indicates faster to in dynamical systems associated with the , as the is dominated by the leading eigenvalue while contributions from others decay exponentially at a rate governed by the gap. In broader applications, the spectral gap influences the rate of diffusion in graph-based processes and the energy separation between ground and excited states in .

Variations Across Fields

In , particularly in , the spectral gap of the Laplacian operator on a is commonly defined as the difference between the smallest positive eigenvalue λ₂ and the trivial eigenvalue λ₁ = 0 of the normalized , providing a measure of the and expansion properties. This definition emphasizes the gap from zero to the next eigenvalue, which quantifies how well the mixes or expands. In , for reversible Markov chains, the spectral gap is defined as the of the between the largest eigenvalue 1 (corresponding to the ) and the second-largest eigenvalue modulus of the , often denoted as γ = 1 - |λ₂|, which governs the to . In physics, especially quantum many-body systems, the spectral gap refers to the ΔE between the E₀ and the first E₁ of the operator, also known as the excitation gap, which distinguishes gapped phases from gapless ones and influences stability and response properties. The term "spectral gap" gained prominence in the and through developments in expander graphs in and , as well as in quantum many-body , building on foundational established by in the early 1900s for operators on Hilbert spaces.
FieldOperator TypeTypical Gap Formula
MathematicsNormalized Laplacianλ₂ (with λ₁ = 0)
Probability|1 - λ₂|
PhysicsΔE = E₁ - E₀

Mathematical Contexts

Functional Analysis and Operators

In , the spectral gap of a L on a , often positive and unbounded, is defined as \gamma = \dist(0, \sigma(L) \setminus \{0\}), where \sigma(L) denotes the of L. This measures the separation between the of L (or the origin) and the remainder of the , capturing stability properties. For operators with , the gap may more generally refer to \inf\{|\lambda - \mu| : \lambda \in \sigma_{\disc}(L), \mu \in \sigma_{\ess}(L)\}, where \sigma_{\disc} and \sigma_{\ess} are the and spectra, respectively; this is particularly relevant for elliptic operators on manifolds, where the arises from behavior at or boundaries. The for unbounded operators, which decomposes the into a direct integral over the via a spectral measure, plays a central role in analyzing such gaps. If a spectral gap exists away from the essential spectrum, it ensures that the resolvent (L - zI)^{-1} is compact for z in the gap, implying that the spectrum in that region consists of isolated eigenvalues of finite multiplicity accumulating only at the essential spectrum boundaries. This compactness facilitates and eigenvalue isolation, as seen in second-order elliptic operators with suitable boundary conditions. A example is the Dirichlet Laplacian -\Delta on a bounded \Omega \subset \mathbb{R}^n with L^2(\Omega) as the . Here, the is purely discrete, starting with the first eigenvalue \lambda_1 > 0, so the spectral gap is \gamma = \lambda_1. This gap directly relates to the Poincaré constant in the inequality \|u\|_{L^2(\Omega)}^2 \leq \lambda_1^{-1} \int_\Omega |\nabla u|^2 \, dx for u \in H_0^1(\Omega), providing control on function oscillations essential for Sobolev embeddings and elliptic regularity. More broadly, \gamma governs the decay of the associated analytic e^{-tL}, yielding \|e^{-tL} f\| \leq e^{-\gamma t} \|f\| for f orthogonal to the (though trivial for Dirichlet conditions), which underpins exponential stabilization in parabolic equations. An advanced application arises in unitary representations of Lie groups, where a uniform spectral gap refers to a positive lower bound on the distance from zero to the spectrum across a family of representations, excluding the trivial one. For compact simple Lie groups, this property holds for dense subgroups generated by elements with algebraic entries, implying rates in the associated unitary dynamics, such as in random walks or ergodic actions. This uniform gap strengthens Kazhdan's property (T), ensuring non-amenability and rapid mixing in representation-theoretic settings.

Graph Theory and Combinatorics

In , the spectral gap of a is defined in terms of the eigenvalues of its L = D - A, where A is the and D is the diagonal . For a connected , the eigenvalues of L satisfy $0 = \lambda_1 < \lambda_2 \leq \cdots \leq \lambda_n, and the spectral gap is given by \lambda_2, known as the algebraic connectivity of the . This value measures the 's connectivity in a spectral sense, with larger \lambda_2 indicating stronger overall connectivity compared to traditional vertex or edge connectivity measures. For the normalized Laplacian \mathcal{L} = I - D^{-1/2} A D^{-1/2}, the eigenvalues are $0 = \mu_1 < \mu_2 \leq \cdots \leq \mu_n \leq 2, and the is \mu_2. This gap relates to the h(G), which quantifies the graph's edge expansion via \min_{S \subseteq V, 0 < |S| \leq |V|/2} \frac{|E(S, V \setminus S)|}{\min(\deg(S), \deg(V \setminus S))}, through : \frac{\mu_2}{2} \leq h(G) \leq \sqrt{2 \mu_2}. This inequality provides a spectral bound on combinatorial expansion properties, linking eigenvalues to cuts in the graph. Expander graphs are families of d-regular graphs where the spectral gap remains bounded below by a fixed \gamma > 0 as the number of vertices n \to \infty. These graphs exhibit strong properties and are applied in constructing error-correcting codes due to their ability to mix information efficiently. Explicit constructions of optimal expanders include Ramanujan graphs, which for d-regular graphs achieve the largest possible spectral gap of at least d - 2\sqrt{d-1} for the (corresponding to \lambda_2 \geq d - 2\sqrt{d-1} for the Laplacian). In the Erdős–Rényi random graph model G(n, p) with p = (1 + \epsilon) \frac{\log n}{n} for fixed \epsilon > 0 above the connectivity threshold, the spectral gap of the normalized Laplacian is \Theta(1) with high probability. This constant gap ensures robust expansion in random settings. Combinatorially, a positive spectral gap \gamma implies vertex expansion: for any set S \subseteq V with |S| \leq n/2, the neighborhood satisfies |N(S)| \geq (1 + \frac{\gamma}{2}) |S| in d-regular graphs. This property underpins applications in derandomization and network design, where expansion guarantees balanced cuts and efficient routing.

Probabilistic and Dynamical Systems

Markov Chains and Mixing Times

In the context of irreducible Markov chains, the spectral gap quantifies the rate at which the chain converges to its . For a P with eigenvalues $1 = \lambda_1 > |\lambda_2| \geq \cdots \geq |\lambda_n|, the spectral gap \gamma is defined as \gamma = 1 - |\lambda_2|. This gap arises from the of P, where the eigenvalue 1 corresponds to the , and the second-largest eigenvalue in modulus determines the decay rate of transient behaviors. The spectral gap directly relates to the mixing time \tau_{\text{mix}}(\epsilon), the number of steps required for the total variation distance between the distribution after t steps and the stationary distribution \pi to be at most \epsilon. Specifically, \tau_{\text{mix}}(\epsilon) \leq \frac{1}{\gamma} \log\left(\frac{1}{\epsilon \pi_{\min}}\right), where \pi_{\min} = \min_i \pi(i). This upper bound is derived by bounding the total variation distance using the L^2 norm: for a reversible lazy chain starting from \mu_0, \| \mu_t - \pi \|_{\text{TV}} \leq \frac{1}{2} \| \mu_t - \pi \|_2 \leq \frac{1}{2} \sqrt{\frac{1}{\pi_{\min}}} e^{-t \gamma}, leading to the logarithmic factor upon solving for t. Lower bounds also scale as \Omega(1/\gamma \log(1/\epsilon)), establishing that the inverse gap captures the asymptotic mixing rate up to logarithmic factors. For reversible Markov chains, where \pi(x) P(x,y) = \pi(y) P(y,x) holds, the spectral gap admits a variational characterization via the Dirichlet form. The gap equals \gamma = \inf_f \frac{\mathcal{E}(f,f)}{\text{Var}_\pi(f)}, where the infimum is over non-constant functions f with \mathbb{E}_\pi = 0, \mathcal{E}(f,f) = \frac{1}{2} \sum_{x,y} \pi(x) P(x,y) (f(x) - f(y))^2 is the Dirichlet form, and \text{Var}_\pi(f) = \mathbb{E}_\pi[f^2] - (\mathbb{E}_\pi)^2. This expression links the gap to conductance \Phi^*, the minimum escape probability from sets of small stationary mass, via Cheeger's inequality: \gamma \geq \Phi^{*2}/2 and \gamma \leq 2 \Phi^*, providing a geometric interpretation for mixing rates. A representative example is the lazy on the C_n of n vertices, where at each step the chain stays put with probability $1/2 or moves to a neighbor with equal probability $1/4. The eigenvalues of the are \lambda_k = \frac{1}{2} (1 + \cos(2\pi k / n)) for k = 0, \dots, n-1, yielding \gamma = 1 - \lambda_1 = 1 - \cos(2\pi / n) = \Theta(1/n^2). The mixing time is then \tau_{\text{mix}}(\epsilon) = \Theta(n^2 \log(n / \epsilon)), illustrating how a small gap leads to quadratic scaling in n. For non-reversible chains, where fails, the standard spectral gap may not tightly bound mixing due to complex eigenvalue dynamics. A relaxation known as the pseudo-spectral gap \gamma_{\text{ps}} addresses this by considering the maximum spectral gap over reversible dilations of the chain, such as Metropolis-Hastings reversibilizations. This enables mixing time bounds analogous to the reversible case, with \tau_{\text{mix}}(\epsilon) \lesssim 1/\gamma_{\text{ps}} \log(1/(\epsilon \pi_{\min})), facilitating analysis of faster-mixing non-reversible processes in applications like optimization.

Random Walks on Graphs

In the context of s on undirected graphs, the simple random walk is governed by the P = D^{-1} A, where A is the and D is the of degrees. The spectral gap of this walk, denoted \gamma, is defined as \gamma = 1 - \lambda_2(P), with \lambda_2(P) being the second-largest eigenvalue of P. For d-regular graphs, this gap equals \lambda_2(L)/d, where L = D - A is the unnormalized and \lambda_2(L) is its second-smallest eigenvalue. In irregular graphs, the gap approximates \lambda_2(L)/\bar{d}, with \bar{d} = 2m/n the average degree, m the number of edges, and n the number of vertices. The spectral gap provides bounds on hitting times, which measure the expected steps for a walk starting at u to reach v. Larger gaps accelerate traversal, with estimates highlighting the graph's scale and connectivity constraints. For regular graphs, the of the is uniform over the vertices, as each has equal and the chain is reversible. The spectral gap \gamma governs the to this distribution in the L^\infty norm, with the error decaying exponentially as O((1 - \gamma)^t) after t steps, ensuring rapid mixing relative to the uniform measure. A representative example is the n-dimensional hypercube graph, which is n-regular with $2^n vertices. Its spectral gap is $2/n, stemming from the eigenvalues of the adjacency matrix n - 2k for k = 0, \dots, n, yielding \lambda_2(P) = 1 - 2/n. This gap implies a mixing time of O(n \log n), reflecting efficient diffusion across the boolean lattice. In directed graphs, the spectral gap of the relates to strong , ensuring the existence of a unique and quantifying the walk's mixing rate only if every is reachable from every other. Unlike undirected cases, the gap here depends on the directed structure, with small gaps indicating potential bottlenecks in reachability despite overall .

Physical Contexts

Quantum Many-Body Systems

In quantum many-body systems, the spectral gap of a H is defined as the difference \Delta = \inf \{ E - E_0 : E > E_0 \} between the E_0 and the lowest E of the excited spectrum above it. This gap characterizes the low-energy excitations of interacting quantum particles and plays a central role in determining the stability and correlation structure of the . For local Hamiltonians describing systems like spin chains or lattices, the presence of a finite \Delta > 0 distinguishes gapped phases from gapless ones, where \Delta = 0 implies a of low-energy states. A finite spectral gap implies short-range correlations in the ground state, with connected correlation functions decaying exponentially with distance, typically on a length scale \xi \sim v / \Delta, where v is the Lieb-Robinson velocity bounding the speed of information propagation. In contrast, gapless phases exhibit power-law or algebraic decay of correlations, signaling critical behavior or long-range order. A representative example is the one-dimensional , H = -J \sum_i \sigma_i^z \sigma_{i+1}^z - h \sum_i \sigma_i^x, where the gap closes at the h = J, marking the transition from a gapped ferromagnetic phase (h < J) to a gapped paramagnetic phase (h > J) through a gapless critical point. The spectral gap also governs locality in the time evolution generated by H. Lieb-Robinson bounds establish that a finite \Delta enforces a light-cone structure for operator spreading, with the speed of information propagation effectively limited by v / \Delta, ensuring that distant regions remain approximately decoupled over short times. This locality underpins simulations of many-body dynamics and proofs of area laws for entanglement entropy in gapped systems. Determining the existence of a spectral gap for local Hamiltonians is fundamentally challenging. In one dimension, the problem of deciding whether a translationally invariant nearest-neighbor Hamiltonian has a gap is undecidable, meaning no algorithm can solve it for all instances; this was initially shown for two dimensions and extended to one dimension in subsequent work. Under small perturbations, the spectral gap exhibits stability quantified by Weyl's inequalities for Hermitian operators: for a perturbation V with \|V\| \ll \Delta(H), the perturbed gap satisfies \Delta(H + V) \geq \Delta(H) - 2 \|V\|, ensuring the system remains gapped for sufficiently weak local modifications. This robustness is crucial for understanding phase stability in realistic .

In , the spectral gap refers to the energy difference between the top of the band and the bottom of the conduction band in the of solids, denoted as E_g = \min E_c - \max E_v, where E_c and E_v are the energies in the conduction and bands, respectively. This gap determines key material properties, such as whether a material behaves as an (E_g > 0), (small E_g), or metal (E_g \leq 0). In , a finite E_g prevents electrical conduction at low temperatures unless thermal excitation or doping promotes electrons across the gap. In topological insulators, the bulk spectral gap is protected by symmetries such as time-reversal invariance, which ensures the robustness of gapless edge or despite the insulating bulk. These protected states arise from nontrivial topological invariants, like the Chern number in two-dimensional systems, enabling dissipationless transport along boundaries. A prototypical theoretical model is the Haldane model on a lattice, where a mass term m from next-nearest-neighbor hopping with complex phases opens a bulk gap \Delta \propto |m| while yielding a nonzero Chern number, realizing a quantum anomalous Hall state without external magnetic fields. In superconductors, the spectral gap manifests as the superconducting gap $2\Delta in the excitation spectrum, arising from the pairing of electrons into Cooper pairs within Bardeen-Cooper-Schrieffer (BCS) theory. This gap equals twice the pairing energy \Delta, suppressing single-particle excitations below $2\Delta and enabling zero-resistance current flow. The value of $2\Delta / k_B T_c \approx 3.5 in weak-coupling BCS superconductors links the gap directly to the critical temperature T_c. Experimental measurements of the spectral gap often employ (ARPES) to map band structures or (STM) to probe local . For instance, ARPES on the \mathrm{Bi_2Se_3} reveals a bulk gap of approximately 0.3 eV, with Dirac-like crossing into the gap. Gap closing in the spectral structure signals phase transitions, such as the metal-insulator transition in Mott insulators, where strong electron correlations open a gap that vanishes under doping or pressure, restoring metallic behavior. In Mott systems like \mathrm{NiS_2}, the gap closure near structural defects or edges can enable localized conduction channels. This transition highlights how interactions in many-body Hamiltonians drive insulating phases, distinct from band insulators.

Significance and Applications

Theoretical Implications

The presence of a spectral gap in the spectrum of operators frequently implies stability and rigidity in the underlying mathematical structures. In , a positive gap above the ensures the uniqueness of the ground state and its robustness against local perturbations, as demonstrated in models like the Affleck-Kennedy-Lieb-Tasaki (AKLT) chain where cluster expansions confirm local indistinguishability of ground states. Similarly, in , a nontrivial spectral gap for the normalized Laplacian characterizes expander graphs, which possess uniform expansion properties that prevent bottlenecks in connectivity and enable pseudorandom behavior despite sparsity. In optimization contexts, such as semidefinite relaxations for graph partitioning problems, the spectral gap provides conditions for exact recovery by ensuring the construction of dual certificates for global optimality of the partition. Universality phenomena arise in the distribution of gaps across random ensembles. In random matrix theory, the fluctuations of the largest eigenvalue and associated gaps in the Gaussian Orthogonal Ensemble (GOE) converge to the , a that governs statistics in diverse physical and statistical models, reflecting non-Gaussian correlations near the spectrum's boundary. Several open s highlight the role of spectral gaps in . Selberg's eigenvalue posits that the smallest positive eigenvalue of the Laplacian on the modular surface is at least 1/4, implying a uniform spectral gap for automorphic forms and influencing distribution via the Selberg . Relatedly, Bourgain's bounds on thin subgroups of semisimple Lie groups establish quantitative spectral gaps, controlling in sparse sets and yielding applications to by restricting representations of integers in these groups. Interconnections between discrete and continuous settings underscore the robustness of spectral gaps under discretization. For graph Laplacians approximating differential operators on manifolds, the spectral gap in the discrete model inherits positivity from the continuous one through finite element or finite difference schemes, preserving essential spectral properties like connectivity bounds as mesh size refines. The absence of a spectral gap signals critical phenomena, including slow relaxation dynamics in gapless systems. In quantum many-body physics, gapless spectra lead to power-law or logarithmic decay in correlation functions, manifesting critical slowing down near phase transitions where relaxation times diverge. This contrasts with gapped systems, where exponential decay ensures rapid equilibration.

Computational and Engineering Uses

The spectral gap of a matrix, particularly the Laplacian of a graph, can be computed using iterative algorithms suited for large sparse matrices arising in practical applications. The Lanczos algorithm is a widely used method for approximating extreme eigenvalues of symmetric sparse matrices, generating a tridiagonal matrix whose eigenvalues provide bounds on the original spectrum. For an n × n sparse matrix with O(n) nonzeros, the algorithm requires k matrix-vector multiplications and orthogonalization steps, leading to a computational complexity of O(n k^2) for k iterations, where k is typically much smaller than n to target the spectral gap near the smallest non-zero eigenvalue. This efficiency makes Lanczos particularly valuable for graphs with millions of nodes, such as those in network analysis. For faster approximations, especially when high precision is not required, the power method can approximate the principal eigenvectors of the normalized , whose eigenvalue gaps relate to the Laplacian spectral properties; convergence occurs at a rate determined by the ratio of consecutive eigenvalue magnitudes. In graph contexts, this facilitates approximations in tasks, where the gap between the k-th and (k+1)-th eigenvalues informs cluster quality, with a small number of iterations often sufficing for relative error guarantees. This method is computationally lightweight, requiring only O(m) time per iteration for m edges, and has been rigorously analyzed for . In engineering applications, the spectral gap plays a key role in optimizing interconnection networks. For VLSI design, expander graphs with large spectral gaps are employed to construct efficient routing topologies, ensuring low congestion and short paths for signal propagation across chips; for instance, Ramanujan expanders provide near-optimal expansion, minimizing wire lengths and improving routability in multi-layer layouts. In , spectral leverages the spectral gap to select the number of clusters and initialize k-means, where a large gap between the k-th and (k+1)-th Laplacian eigenvalues indicates well-separated components, projecting data onto the corresponding eigenspace for robust initialization and reducing sensitivity to random starts. Quantum computing exploits the spectral gap in adiabatic algorithms, where the guarantees that a system remains in its during slow evolution if the minimum gap along the path is at least Ω(1/poly(n)) for n qubits, enabling polynomial-time computation for optimization problems. A prominent example is the adiabatic implementation of Grover's search, where engineered Hamiltonians maintain a constant or polynomially large gap throughout the interpolation from an initial superposition to the marked state, achieving the quadratic speedup over classical search while avoiding small-gap bottlenecks through path optimization. In networking, high-spectral-gap graphs underpin load balancing in data centers by ensuring rapid of processes, such as protocols or iterative averaging, where the inversely bounds the mixing time to O(1/ · log n) steps for balancing workloads across servers. Expander-based topologies, like those in fat-tree alternatives, leverage this property to distribute evenly, reducing in large-scale clusters with thousands of nodes.

References

  1. [1]
    [PDF] arXiv:2412.13391v1 [math.SP] 18 Dec 2024
    Dec 18, 2024 · A bounded component of the complement of the spectrum is called a spectral gap (or just a gap). ... open spectral gap for some operator family ...
  2. [2]
    Spectral Gap - an overview | ScienceDirect Topics
    Spectral gap is defined as a property of a strongly continuous unitary representation of a locally compact second countable (lcsc) group, ...
  3. [3]
    [PDF] an introduction to spectral graph theory
    Abstract. Spectral graph theory is the study of properties of the Laplacian matrix or adjacency matrix associated with a graph. In this paper, we focus.
  4. [4]
    [PDF] Spectral Graph Theory - UCSD Math
    αivik2 ≤ C maxi6=1 |λi| d t. 22. Page 23. Thus we define the following. Definition 8.2. For a graph G, its spectral gap λ(G) is maxi6=1 |λi| = max{|λ2|,|λn|}.
  5. [5]
    [PDF] The Eigenvalue Gap and Mixing Time
    Mar 14, 2011 · The eigenvalue or spectral gap of a Markov chain is the difference between the two largest eigenvalues of the transition matrix of its ...
  6. [6]
    [PDF] Lecture 28 : The Spectral Gap - UC Berkeley Statistics
    The spectral gap (λ) is defined as min E(f,f) / varπ(f), where E is the Dirichlet form. It's the smallest nonzero eigenvalue of I - 1/(K + K*).
  7. [7]
    [PDF] The Importance of the Spectral Gap in Estimating Ground-State ...
    The spectral gap, the difference between the smallest two eigenvalues, is crucial for understanding the ground space of any Hamiltonian in many-body physics.
  8. [8]
    [PDF] Chapter 9: The Spectrum of Bounded Linear Operators
    In Chapter 7, we used Fourier series to solve various constant coefficient, linear par- tial differential equations, such as the heat equation.
  9. [9]
    [PDF] The spectral gap. Existence and Implications
    The Spectral Gap – Definition. Let H = H∗ on a Hilbert space H be such that 0 = inf specH is an eigenvalue. The corresponding eigenspace, ker H is then.
  10. [10]
    Spectral gaps of two- and three-dimensional many-body quantum ...
    The spectral gap is the energy difference between the ground and first excited state, key for quantum phenomena. This paper presents a new method to calculate ...
  11. [11]
    [PDF] Expander graphs and their applications - CS - Huji
    Aug 7, 2006 · EXPANDER GRAPHS AND THEIR APPLICATIONS. 441. We are also grateful for ... SHLOMO HOORY, NATHAN LINIAL, AND AVI WIGDERSON. 3.46. 3.465 n ...
  12. [12]
    [PDF] Markov Chains and Mixing Times, second edition David A. Levin ...
    Page 1. Markov Chains and Mixing Times, second edition. David A. Levin. Yuval Peres. With contributions by Elizabeth L. Wilmer. University of Oregon. E-mail ...
  13. [13]
    Undecidability of the spectral gap - Nature
    Dec 9, 2015 · The spectral gap—the energy difference between the ground state and first excited state of a system—is central to quantum many-body physics.
  14. [14]
    Geometric Bounds for Eigenvalues of Markov Chains - Project Euclid
    We develop bounds for the second largest eigenvalue and spectral gap of a reversible Markov chain. The bounds depend on geometric quantities.
  15. [15]
    [PDF] Spectral Gaps for Self-Adjoint Second Order Operators - arXiv
    Sep 4, 2012 · Abstract. We consider a second order self-adjoint operator in a domain which can be bounded or unbounded. The boundary is partitioned into ...
  16. [16]
    [PDF] On Approximation of the Eigenvalues of Perturbed Periodic ... - arXiv
    Feb 14, 2007 · Abstract. This paper addresses the problem of computing the eigenvalues lying in the gaps of the essential spectrum of a periodic ...
  17. [17]
    [PDF] Spectral Theorem for Symmetric Operators with Compact Resolvent
    A compact operator has a convergent subsequence of bounded sequences. The goal is to give an elementary proof of spectral theorem for compact normal operators.Missing: gap | Show results with:gap
  18. [18]
    [PDF] arXiv:1811.08285v1 [math.AP] 20 Nov 2018
    Nov 20, 2018 · Firstly we obtain estimates of the Poincaré constant in the Poincaré-Sobolev ... spectral gap for Dirichlet-Laplacian satisfies λ2(Ω) − λ1(Ω) ≥ λ2 ...
  19. [19]
    [PDF] A spectral gap theorem in simple Lie groups - arXiv
    May 8, 2014 · The purpose of the paper is to study the spectral gap property for measures on a compact simple Lie group G. If µ is a Borel probability ...
  20. [20]
    [PDF] Algebraic connectivity of graphs - Czechoslovak Mathematical Journal
    It is the purpose of this paper to find its relation to the usual vertex and edge connectivities. We recall that many authors, e.g. A. J. HOFFMAN, M. DOOB, ...Missing: original | Show results with:original
  21. [21]
    [1201.0425] Spectral gaps of random graphs and applications - arXiv
    Jan 2, 2012 · We study the spectral gap of the Erdős--Rényi random graph through the connectivity threshold. In particular, we show that for any fixed \delta > 0 if p \ge \ ...
  22. [22]
    Concentration inequalities for Markov chains by Marton couplings ...
    In the case of non-reversible chains, we introduce a new quantity called the “pseudo spectral gap”, and show that it plays a similar role for non-reversible ...
  23. [23]
    [PDF] Lectures on Spectral Graph Theory Fan R. K. Chung
    The discrete analogue of the Cheeger in- equality has been heavily utilized in the study of random walks and rapidly mixing. Markov chains [224]. New spectral ...Missing: gap | Show results with:gap
  24. [24]
    [PDF] Spectral Graph Theory: Random Walks - csail
    Given an undirected graph G = (V,E) and some starting vertex s, a random walk in G of length t is defined as a randomized process in which, starting from the ...
  25. [25]
    [PDF] Spectral graph theory and random walks on graphs - Kimball Martin
    Feb 3, 2014 · One of the main themes of algebraic graph theory comes from the following question: what do matrices and linear algebra tell us about graphs?
  26. [26]
    Random walks on graphs: new bounds on hitting, meeting ...
    In this paper, we derive new bounds on return probabilities, hitting times, and meeting deterministic trajectories for these processes. We also bound the ...Missing: d_min) | Show results with:d_min)
  27. [27]
    [PDF] Maximum Hitting Time for Random Walks on Graphs
    Abstract. For x and y vertices of a connected graph G, let TG(x, y) denote the expected time before a random walk starting from x reaches y.Missing: spectral gap (γ d_min)
  28. [28]
    [PDF] Lecture 11 - Cornell University
    Sep 22, 2016 · Therefore the mixing time for random walks is O(ln n α. ). For the lazy random walk, we can say more about the mixing time. The spectral gap ...Missing: cycle | Show results with:cycle<|separator|>
  29. [29]
    [PDF] (Lazy) random walks, their stationary distribution and l2 ...
    Sep 22, 2009 · The natural next step at this point would be to claim that the random walk of a graph G always converges to the stationary distribution π. This ...
  30. [30]
    [PDF] Modern Discrete Probability VI - Spectral Techniques Background
    Dec 1, 2014 · The relaxation time for lazy simple random walk on the hypercube is trel = n. Proof: The eigenvalues are n−|J| n for J ⊆ [n]. The spectral gap ...
  31. [31]
    [PDF] Mixing Times of Markov Chains: Techniques and Examples
    Feb 4, 2019 · Thus φj is an eigenfunction with eigenvalue cos(2πj/n). If n is even, the chain is periodic and the absolute spectral gap is 0. If n is odd, the ...
  32. [32]
    [1708.00530] The Spectral Gap of Sparse Random Digraphs - arXiv
    Aug 1, 2017 · This is the first result concerning the asymptotic behavior of the spectral gap for sparse non-reversible Markov chains with an unknown stationary distribution.Missing: connectivity | Show results with:connectivity
  33. [33]
    [PDF] S-T Connectivity on Digraphs with a Known Stationary Distribution∗
    We present a deterministic logspace algorithm for solving S-T CONNECTIVITY on directed graphs if. (i) we are given a stationary distribution for random walk ...
  34. [34]
    Spectral analysis of product formulas for quantum simulation - Nature
    Apr 11, 2022 · By Weyl's inequality, the perturbation in the eigenvalues satisfies ... For instance, for 2-local Hamiltonian with zero spectral gap, it's ...
  35. [35]
    Bandgap engineering of two-dimensional semiconductor materials
    Aug 24, 2020 · In the bulk, the valence (conduction) band edge is observed at the Γ (Q) point and the bandgap is therefore indirect. As the number of layers ...
  36. [36]
    Revisiting the optical bandgap of semiconductors and the proposal ...
    Aug 2, 2019 · Instead, the bandgap of amorphous semiconductors can be defined by taking the photon energy at which the optical absorption coefficient reaches ...
  37. [37]
    [PDF] First Principles Calculation of the Topological Phases of the ... - arXiv
    respectively, M is the so-called “mass term” and ϕ is the phase factor determined by the ... For the electronic Haldane model the gap energy can be taken equal to ...Missing: formula | Show results with:formula
  38. [38]
    [PDF] Determination of Superconducting Gap of SmFeAsFxO1-x ... - arXiv
    The determined gap value of 2∆ = 13.34 ±0.47 meV for. SmFeAs(O0.85F0.15) gives 2∆/kBTC = 3.68, close to the BCS prediction of 3.53. The superconducting gap.
  39. [39]
    [PDF] We studied band gap in ultrathin Bi2Se3 film by electronic transport ...
    ARPES measurements indicate Bi2Se3 has 0.3 eV bulk band gap and a single ... that in few-QL Bi2Se3 the surface states may hybridize and open a bulk energy gap, ...
  40. [40]
    Mott materials: unsuccessful metals with a bright future - Nature
    Oct 1, 2024 · Mott insulators are unsuccessful metals with a large density of carriers that are made inactive by Coulomb repulsion but could become available for electric ...
  41. [41]
    [2407.19636] Closing of the Mott gap near step edges in NiS2 - arXiv
    Jul 29, 2024 · Instead, the Mott gap is closed near step edges, suggesting possible electrical conduction from one-dimensional channels. The edge anomaly was ...
  42. [42]
    Disorder induced power-law gaps in an insulator–metal Mott transition
    Oct 15, 2018 · Theory predicts that disorder might play a role in driving a Mott insulator across an IMT, with the emergent metallic state hosting a power-law suppression of ...
  43. [43]
    Stability of the spectral gap and ground state indistinguishability for ...
    Sep 2, 2022 · We use cluster expansions to establish local indistiguishability of the finite-volume ground states for the AKLT model on decorated hexagonal lattices.Missing: rigidity expanders optimization global minima
  44. [44]
    [PDF] Certifying Global Optimality of Graph Cuts via Semidefinite Relaxation
    Jul 7, 2018 · This is made possible by considering minimal ratio cuts and normalized cuts, which represent a traditional problem in graph theory [14, 5].
  45. [45]
    [PDF] The Distributions of Random Matrix Theory and their Applications∗
    This paper surveys the largest eigenvalue distributions appearing in random matrix theory and their application to multivariate statistical analysis. Contents.
  46. [46]
    254B, Notes 3: Quasirandom groups, expansion, and Selberg's 3/16 ...
    Dec 16, 2011 · Since the numerical method does count the eigenvalues, Selberg's Conjecture follows. However, this does not work all the time. The first ...
  47. [47]
    [PDF] Some Diophantine applications of the theory of group expansion
    The recent years have seen considerable progress in the theory of expansion and spectral gaps for so-called thin groups (as opposed to the classical theory.
  48. [48]
    [PDF] Spectral gap for quantum graphs and their connectivity - arXiv
    Feb 20, 2013 · Abstract. The spectral gap for Laplace operators on metric graphs is investigated in relation to graph's connectivity, in particular what ...
  49. [49]
    Rigorous lower bound on dynamical exponents in gapless ... - arXiv
    Jul 19, 2025 · The vanishing of the spectral gap implies a divergence of the relaxation time, a phenomenon known as critical slowing down. The dynamical ...
  50. [50]
    [PDF] 10 The Lanczos Method - cond-mat.de
    The Lanczos iteration [1] was conceived as a method for tridiagonalizing Hermitian matrices. Like the related Arnoldi method [2] for non-Hermitian matrices, ...
  51. [51]
    [PDF] Spectral Clustering via the Power Method - Provably
    We prove that solving the k-means problem on the approx- imate eigenvectors obtained via the power method gives an additive-error approximation to solving the k ...
  52. [52]
    Scalable expanders | Proceedings of the twenty-sixth annual ACM ...
    Design and analysis of algorithms · Approximation algorithms analysis · Routing and network design problems · Graph algorithms analysis · Network flows.
  53. [53]
    Measuring the stability of spectral clustering - ScienceDirect.com
    Feb 1, 2021 · In this paper we measure the stability of spectral k-clustering, for any number k of clusters, by the structured distance to ambiguity.Missing: initialization | Show results with:initialization<|separator|>
  54. [54]
    [PDF] Adiabatic quantum computation is equivalent to standard ... - arXiv
    Mar 26, 2005 · More precisely, the adiabatic computation is polynomial time if this minimal spectral gap is at least inverse polynomial. The motivation for ...
  55. [55]
    Adiabatic quantum computation | Rev. Mod. Phys.
    Jan 29, 2018 · Since the gap for the adiabatic process is large, QA takes only polynomial time. ... gap and the run time does not hold as a general theorem.
  56. [56]
    [PDF] ASAP: Asynchronous Approximate Data-Parallel Computation - arXiv
    Dec 27, 2016 · The spectral gap of a communication graph determines how rapidly a dis- tributed, iterative sparse reduce converges when perform- ing ...
  57. [57]
    Large-Scale Spectral Graph Neural Networks via Laplacian ... - arXiv
    Jan 8, 2025 · Graph Neural Networks (GNNs) have gathered increasing research attention because of their versatility in handling graph-structured data.