Scatterer arrays with latent symmetries enable intruder localization and identification by quantifying symmetry breaking in a capacitance matrix model, with Bayesian inference and neural networks outperforming dictionary methods under noise.
Latent symmetry in a minimal non-Hermitian trimer
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
We study a minimal non-Hermitian trimer with latent symmetry formed by a cospectral pair of sites embedded in a three-site network with nonreciprocal couplings. We show that cospectrality provides a structural latent-symmetry constraint, whereas exact dark-state decoupling requires an additional algebraic matching condition among the couplings. For a dark-state-compatible representative of this cospectral class, the model admits an exact decomposition into dark and bright sectors: the dark mode is spectrally isolated and retains a complex eigenvalue, while the bright sector reduces to an effective non-Hermitian dimer. For a suitable choice of parameters, this reduced subsystem becomes $\mathcal{PT}$-symmetric and exhibits partial spectral reality, with two real eigenvalues coexisting with the complex dark eigenvalue. At the critical point, the bright sector hosts an embedded second-order exceptional point, which renders the full trimer defective and gives rise to the characteristic Jordan-block dynamics. These results establish the non-Hermitian trimer as a minimal analytically solvable setting in which latent symmetry, sector-resolved $\mathcal{PT}$ symmetry, and exceptional-point physics naturally coexist.
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Target localization, identification and sensing using latent symmetries
Scatterer arrays with latent symmetries enable intruder localization and identification by quantifying symmetry breaking in a capacitance matrix model, with Bayesian inference and neural networks outperforming dictionary methods under noise.