Generalized Grassmann graphs associated to conjugacy classes of finite-rank self-adjoint operators
Mark Pankov, Krzysztof Petelczyc, Mariusz Zynel

TL;DR
This paper generalizes Grassmann graphs to conjugacy classes of finite-rank self-adjoint operators, characterizing their automorphisms as compositions of unitary or anti-unitary transformations and eigenspace permutations.
Contribution
It extends the adjacency concept to broader conjugacy classes and classifies automorphisms, linking them to unitary, anti-unitary, and semilinear transformations.
Findings
Automorphisms are composed of unitary/anti-unitary operators and eigenspace permutations.
Classical Chow's theorem applies to two-eigenvalue cases, involving semilinear automorphisms.
Generalized graphs reveal structure of conjugacy classes beyond classical Grassmann graphs.
Abstract
Two distinct projections of finite rank are adjacent if their difference is an operator of rank two or, equivalently, the intersection of their images is -dimensional. We extend this adjacency relation on other conjugacy classes of finite-rank self-adjoint operators which leads to a natural generalization of Grassmann graphs. Let be a conjugacy class formed by finite-rank self-adjoint operators with eigenspaces of dimension greater than . Under the assumption that operators from have at least three eigenvalues we prove that every automorphism of the corresponding generalized Grassmann graph is the composition of an automorphism induced by a unitary or anti-unitary operator and the automorphism obtained from a permutation of eigenspaces with the same dimensions. The case when the operators from have two eigenvalues only is…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
