Topological phases of matter are characterized by invariants that resist local perturbation. Three apparently different diagnostic tools have been used to detect these phases: the feature spectrum (projecting the Hamiltonian onto subspaces defined by quantum observables like spin), the entanglement spectrum (eigenvalues of the reduced density matrix for a spatial partition), and the Wilson loop spectrum (the holonomy of the Berry connection around the Brillouin zone). Each has its own mathematical framework, physical motivation, and community of practitioners.
Note (2026-11-11): checked against the paper's full text (this essay had been logged as held up on an abstract-level read). The authors, the nested construction, the feature-energy complementarity and the claim that boundary modes can persist in the feature spectrum when the energy spectrum is gapped are as in the paper. Three statements go beyond it. (1) The essay says the authors prove that the three spectra are equivalent. The paper's figure caption says the equivalence between the Wilson loop spectrum, the spatially resolved entanglement spectrum and the spatially resolved feature spectrum has been extensively studied in earlier work; what the paper adds is a proof that the feature and entanglement spectra, with a consistent partitioning, have identical nonzero eigenvalues (through Sylvester's determinant theorem), and the extension of the equivalence to sectors within the feature spectrum (the nested construction). (2) The essay says the feature spectrum's spectral flow is identical to the Wilson loop winding. The paper shows that the spectral flows can be adiabatically connected to the Wilson loop winding. (3) The closing generalisation, that topology lives in the complementarity between any two conjugate descriptions of a system, is the essay's, not the paper's.
Hung, Ong, and Lin (arXiv:2603.13128) prove that for non-interacting fermionic systems, the three are equivalent. Feature spectrum topology, entanglement properties, and Wilson loop spectra are different views of the same underlying structure. The feature spectrum encodes entanglement between quantum observable sectors (spin-up versus spin-down, orbital angular momentum components), and its spectral flow is identical to Wilson loop winding — both manifest what the authors call feature-energy complementarity.
The recursive construction makes this precise: applying projection operators to feature spectra subsectors generates a nested hierarchy, and the topology at each level maps exactly onto the entanglement spectrum at the corresponding partition and the Wilson loop at the corresponding cycle. Topological boundary modes can appear in the feature spectrum even when the energy spectrum is fully gapped — the feature spectrum sees topology that the energy spectrum misses.
This refines the bulk-boundary correspondence. The conventional statement — gapless boundary modes signal nontrivial bulk topology — is a special case of a broader principle where the “boundary” can be in feature space rather than real space. The topology isn't bound to edges; it lives in the complementarity between any two conjugate descriptions of the same system.