Spectral Localizer Framework in Topological Systems

• The spectral localizer framework is a rigorous operator-based method that computes local topological invariants using real-space data from quantum and classical Hamiltonians.
• It introduces local spectral gap conditions and relative resolvent-damped bounds to overcome the limitations of traditional momentum-space invariants in inhomogeneous systems.
• Recent advances refine tapering constants and spectral-flow measures, enhancing numerical efficiency and robust detection of phase boundaries in disordered and aperiodic materials.

The spectral localizer framework is a rigorous operator-theoretic scheme for analyzing and computing local topological invariants of quantum and classical Hamiltonians, particularly in nonperiodic, disordered, aperiodic, or heterostructured systems. It leverages real-space data to circumvent the limitations of momentum-space invariants, enabling the direct numerical evaluation of topological markers and their robustness. Recent advances have refined the foundational criteria of the framework, introducing truly local gap conditions, relative operator bounds, and improved constants, thereby extending its precision and applicability in highly inhomogeneous systems (Cerjan et al., 17 Jun 2025).

1. Formal Definition of the Spectral Localizer

The spectral localizer is defined using a self-adjoint Hamiltonian $H$ on a Hilbert space $\ell^2(\mathbb{Z}^d;\mathbb{C}^L)$, together with position operators $X = (X_1, ..., X_d)$. For a fixed spatial point $x\in\mathbb{R}^d$, energy shift $E\in\mathbb{R}$, and tapering parameter $\kappa > 0$, the localizer is constructed as

$L_\kappa(H,x) = -H\Gamma + \kappa D(x),$

where, in even $d$, $D(x) = \sum_{j=1}^d \gamma_j (X_j - x_j)$ is a dual Dirac operator using an irreducible Clifford representation $\{\gamma_j\}$, and $\Gamma = \mathrm{diag}(1, -1)$. In block form,

$$L_\kappa(H,x) = \begin{pmatrix} -(H-E) & \kappa D_0(x)^* \ \kappa D_0(x) & +(H-E) \end{pmatrix}.$$

For a finite region $B_\rho(x)\subset\mathbb{Z}^d$, the finite-volume restriction is $L_{\kappa,\rho}(H,x)$. When $0$ is gapped in this matrix, the local topological marker (index) is given by

$$\mathrm{Index}_{\kappa,\rho}(H,x) = \frac{1}{2}\,\mathrm{Sig}(L_{\kappa,\rho}(H,x)),$$

where $\mathrm{Sig}$ denotes the matrix signature. The smallest singular value,

$$\mu_{\kappa,\rho}(H,x) = \min \mathrm{spec}|L_{\kappa,\rho}(H,x)|,$$

is the localizer gap, a quantitative local measure of topological protection (Cerjan et al., 17 Jun 2025).

2. Fundamental Advances in Local Criteria

a) Local (Dirichlet-type) Spectral Gap:

Earlier formulations required a global bulk gap $\inf \mathrm{spec}(H^2) \ge g^2$. The present framework now allows for a strictly local spectral gap: $(H^2)_\rho(x) \ge g_\rho^2 \cdot 1_\rho(x)$ where $(\cdot)_\rho$ denotes restriction to the finite region $B_\rho(x)$. This "local gap" $g_\rho(H,x)$ is sufficient for topological protection of the localizer gap, drastically relaxing global constraints and accommodating internal boundaries, defects, or inhomogeneous structures.

b) Relative Resolvent-Damped Bounds:

The operator norms $\|H\|$ and $\|[D,H]\|$ are replaced by relative, resolvent-damped versions: $$\|H R\|, \qquad \|[D, H] R\|, \qquad R = (i + c\kappa/g_\rho D)^{-1}$$ which localize the control to the vicinity of $x$. These bounds allow perturbations far from $x$ to be absorbed in the resolvent, and are essential for heterostructures and disordered samples with large remote inhomogeneities.

c) Sharpened Tapering Constant:

The commutator estimate for a tapering function $F_\rho$ localizing in $B_\rho(x)$ is improved: $\|[F_\rho(D), H]\| \leq \frac{C_F}{\rho} \|[D,H]\|$ with $C_F$ reduced from $8$ to approximately $4.56$, and numerically down to near $2$ for nearest-neighbor models, improving quantitative efficiency (Cerjan et al., 17 Jun 2025).

3. Principal Technical Results and Invariants

Theorem 3.2 establishes that for "weakly local" Hamiltonians (bounded $[D(x), H]$), a finite $\rho$-local gap $g_\rho > 0$, and for $\kappa$ in a precise range defined by $g_\rho$, $\rho$, resolvent-damped norms, and the sharp $C_F$, the finite-volume localizer $L_{\kappa,\rho'}(H,x)$ (for any $\rho'\geq\rho$) satisfies: $$L_{\kappa,\rho'}(H,x)^2 \geq b^2 g_\rho^2 \cdot 1_{\rho'}(x)$$ yielding a localizer gap $\mu_{\kappa,\rho'}(H,x) \geq b g_\rho > 0$ and index stability across $\rho'$. The conditions relate $\kappa$ to the local gap and locality constants: $$\kappa > \frac{2 g_\rho}{\rho}, \qquad \kappa \leq \frac{g_\rho^3}{(1/(1-a-b^2))(C_F \|H R\| + g_\rho)\|[D(x), H] R\|}$$ for auxiliary parameters $a, b \geq 0$ with $a + b^2 < 1$.

Propositions ensure stability under distant perturbations (local gap $g_\rho$ unchanged if the perturbation $W$ is outside $B_\rho(x)$), and spectral-flow invariance (the local index difference along paths $x_t$ is robust to any $W$ vanishing at the endpoints, regardless of overlap with the interface path) (Cerjan et al., 17 Jun 2025).

4. Applications to Nonperiodic, Aperiodic, and Disordered Systems

The framework is applicable even when global Bloch band theory fails:

• Heterostructures: The localizer gap and index detect robust local topology within spatially varying, disconnected domains.
• Aperiodic and Quasiperiodic Lattices: Valid without Bloch periodicity; local spectral gaps suffice for local topological markers.
• Disordered Systems: In regimes without a global spectral gap, the localizer gap may remain finite for modest regions and low local density of states, ensuring well-defined invariants even in the mobility-gap regime. Numerical studies confirm that values such as $\rho \sim 10\!-\!20$ unit cells and $\kappa \sim 0.2\,t/a$ yield stable local Chern markers into localized phases.

These developments provide tools for the computation of local invariants in systems where translation symmetry, periodicity, or a bulk gap are absent (Cerjan et al., 17 Jun 2025).

5. Spectral-Flow, Robustness, and Phase Boundary Analysis

The framework quantifies topological phase transitions and their robustness:

• Spectral Flow: The line integral of $\partial_t \mu$ along a path $x_t$ in space counts the signed zero-crossings of the spectral localizer, equaling the difference in local indices between endpoints.
• Defect Robustness: Localized perturbations along the interface path do not alter the spectral flow or transversality at crossings, provided they vanish at endpoints. Explicit bounds give a radius-dependent criterion for a defect's distance such that eigenvalue crossing signs remain unchanged.
• Slope-Stability: For a simple zero crossing of the restricted localizer $L^Λ_\kappa(H, x_{t_c})$ with nonzero slope, distant perturbations only shift the eigenvalue and its slope by quantities controlled by the decay from the crossing region, preserving crossing transversality.

These results provide enhanced, quantitative robustness bounds for the detection of topological phase boundaries in inhomogeneous media (Cerjan et al., 17 Jun 2025).

6. Theoretical and Algorithmic Implications

• Truly Local Topological Probe: The local criteria enable computation of invariants even where global approaches fail or are meaningless, making the spectral localizer a direct local diagnostic.
• Numerical Accessibility: The improved tapering constant and relative bounds decrease the required system size and computational resources.
• Interface Physics: The framework naturally predicts the presence and location of robust boundary or interface states, and its explicit spectral-flow invariants underpin bulk-edge correspondence in the absence of global gaps.
• Platform Generality: Applicable to artificial and natural materials, including electronic, photonic, mechanical, or cold atomic systems lacking clean periodicity.
• Future Extensions: The advances suggest further analytic and algorithmic refinement in detecting topology in interacting, time-dependent, and more strongly disordered settings.

7. Connection to Other Real-Space Invariants

A systematic perturbative expansion in the localizer parameter $\kappa$ reveals that the spectral localizer index reduces to widely used local Chern and winding markers at leading order, thus providing an explicit operator-theoretic origin and clarifying their equivalence for non-crystalline, disordered, and amorphous topological systems (Jezequel et al., 31 Jul 2025).

---

References:

• "Detecting local topology with the spectral localizer" (Cerjan et al., 17 Jun 2025)
• "Explicit equivalence between the spectral localizer and local Chern and winding markers" (Jezequel et al., 31 Jul 2025)

These provide the definitive technical account of the local spectral gap framework, relative operator norm improvements, and quantitative criteria underlying the modern spectral localizer approach.