Search arXiv⌕ Search

arXiv · 2610.14540

Multiplicity One and van Dijk's Conjecture for Connected Complex Symmetric Spaces

Abstract

We give new invariant-theoretic proofs of regularity for the three classical residual families DIII, balanced CII, and BDI/Spin. In DIII and balanced CII, trace contractions and a supertrace identity show that the relevant character spaces in the quadratic moment quotient vanish. These quotient-level vanishings lift to the corresponding equivariant tempered-distribution vanishings. In BDI/Spin, Gram matrices and a volume tensor give the analogous vanishing for the two single-determinant characters. We also give a direct proof of the required equivariant-distribution vanishing for EVII. These results, together with corrected distinguished-nilpotent calculations for the nice factors, establish the required componentwise regularity for the finite-component data attached to all irreducible factors and repair gaps in the previously proposed proofs. Combining these factorwise results across products and finite central quotients yields regularity for every symmetric pair of a connected complex reductive group. Generalized Harish-Chandra descent and a compatible Chevalley involution give multiplicity one for irreducible Casselman-Wallach representations in the connected-group case. A finite-component restriction argument extends this multiplicity-one statement to every possibly disconnected complex reductive $G$ whose symmetric space $G/G^θ$ is connected. Consequently, we establish van Dijk's conjecture on the unitary Gelfand property for connected complex symmetric spaces.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Yuxuan Ge. 2026-10-10. Multiplicity One and van Dijk's Conjecture for Connected Complex Symmetric Spaces. https://arxiv.org/abs/2610.14540

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

Pointwise character bounds for $\mathrm{SU}(3)$

We present a basic pointwise bound for the irreducible characters of $\mathrm{SU}(3)$ and, as an application, derive new $L^p$ bounds for these characters. Our approach is based on the descent of characters to singular sets and the cancellation in this formula.

math.RT↗

Stranding $\mathfrak{sl}_n$ webs

Webs are planar directed graphs that encode invariant vectors for tensor products of fundamental $U_q(\mathfrak{sl}_n)$-representations. For $\mathfrak{sl}_2$ and $\mathfrak{sl}_3$, web calculus is governed by effective reduction rules and well-understood reduced-web bases, but these features break down in higher rank. In this paper we develop a global combinatorial framework for untagged $\mathfrak{sl}_n$ webs, called \emph{strandings}, which organizes local labeling data into systems of colored directed paths on the web graph. Our main result is an explicit state-sum formula for the invariant vector of an untagged web as a weighted sum over its valid strandings. The usual labeling-based construction requires a choice of decomposition into elementary web pieces and a vertex-by-vertex coefficient calculation. Our global reformulation avoids both. We prove that the vectors produced by strandings are $U_q(\mathfrak{sl}_n)$-invariant, compare the untagged theory with the tagged-web framework of Cautis, Kamnitzer, and Morrison, and use this comparison to obtain a complete set of relations for untagged web graphs. We also give applications of strandings to nonvanishing results, basis constructions from tableaux, and connections with Springer-theoretic combinatorics.

math.RT↗