arXiv:2610.03108
Determining the Number of Symmetry Sectors in Composite Quantum Spectra
Nisarg Vyas, S. Harshini Tekur, M. S. Santhanam
Abstract
The spectral fluctuations in complex quantum systems are modeled in terms of spectra from random matrices. Quantitative agreement with random matrices holds only when the quantum spectrum is a pure sequence, and not a composite spectrum that mixes levels from different symmetry sectors. It was shown earlier that discrete symmetries in composite spectra can be detected using higher-order spacing ratio statistics provided the symmetry sectors are of equal dimensions. The general case is when the symmetry sectors have unequal dimensions, and the symmetries of the Hamiltonian system are unknown or only partially known. In this work, we address how the number of symmetry sectors can be determined from a composite spectrum arising in Hermitian and non-Hermitian settings. A Best Subset Technique is proposed for determining the number of the symmetry sectors by comparing the spectral fluctuations to that of an appropriate random matrix ensemble without requiring de-symmetrization. It is demonstrated using spectra from Gaussian, Wishart and Ginibre ensembles, and also for spectra drawn from many-body quantum systems.
AI Summary
Composite spectra mixing independent symmetry sectors can obscure random-matrix correlations. This paper asks whether the number of sectors can be inferred without symmetry resolution or spectral unfolding when sector dimensions differ, extending earlier higher-order spacing-ratio diagnostics for equal-size GOE blocks. The proposed Best Subset Technique scans the spacing-ratio order $k$ and a spectral fraction $f$, minimizing the Kullback–Leibler divergence between the empirical ratio distribution and a random-matrix spacing-ratio surmise. Windows are taken from the spectral center for Gaussian ensembles and the lower end for Wishart eigenvalues and Ginibre singular values. The inferred sector count is $m=k^*$, using an empirical effective-index relation $\beta'=k+\beta-1$ for Dyson classes $\beta=1,2,4$. Numerical demonstrations cover Gaussian, square Wishart–Laguerre, and Ginibre ensembles. For four GUE blocks of dimensions $5000,10000,15000,20000$, the full-spectrum diagnostic returns three blocks, whereas optimization returns four using approximately $32\%$ of the spectrum. Composite spectra from three transverse-field Ising systems similarly give the correct count using approximately $21\%$. A non-Hermitian hard-core-boson example uses singular values and returns three blocks with or without subset selection. These many-body tests combine spectra from different system sizes. For two random-matrix blocks, the optimal fraction approximately collapses by Dyson index and relative block size across ensembles. The method is empirical rather than analytically established and requires sufficiently overlapping spectra with enough levels for reliable statistics. Analytical justification and extension to rectangular Wishart ensembles with unequal matrix dimensions are left open.
Central claim may not hold: BST is supported as an empirical estimator under overlap and sample-size conditions, not as a generally established recovery method for arbitrary unequal-dimensional sectors. The stated TFIM Hamiltonians each retain global spin-flip parity, giving six parity-resolved blocks in the three-system composite unless an unreported projection was applied. The non-Hermitian demonstrations require singular values and therefore do not establish inference from complex eigenvalue spectra alone. The written KL objective omits the ensemble-dependent shifts of its reference parameter required by the paper's own unitary and symplectic prescriptions. The hard-core-boson example specifies impossible exact half-filling for its odd system size L=17 without identifying the actual particle-number sector.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 10 pages, 13 figures