Unitary Schur Sampling of Qudits via Random SWAP Tests: Hunt for Antisymmetry
π arXiv:2610.02103 Β· π₯ PDF Β· 2026-10-01 Β· hep-th
Authors: Shrigyan Brahmachari [arXiv Β· scholar] , David Jakab [arXiv Β· scholar] , Henry D. Pfister [arXiv Β· scholar] , Iman Marvian [arXiv Β· scholar]
π Abstract
Schur sampling is a fundamental primitive for extracting permutation-invariant information from many-body quantum systems and has broad applications in quantum information science. Circuit-based implementations typically rely on coherent representation-theoretic operations such as Clebsch-Gordan transforms, generalized phase estimation, or quantum Fourier transforms over the symmetric group. We show that unitary Schur sampling of arbitrary permutation-invariant mixed states on $n$ $d$-dimensional qudits can instead be implemented using only pairwise SWAP tests, namely, two-qudit projective measurements onto the symmetric and antisymmetric subspaces. Our protocol achieves error $Ξ΅$ in diamond distance using $O(n\min\{n,d\}^{3}\log^2(n/Ξ΅))$ random SWAP tests. The central idea is to repeatedly identify and extract the largest antisymmetric subsystem. This turns unitary Schur sampling into a search for antisymmetry and gives the algorithm a natural interpretation as a stochastic traversal of a Young diagram. The protocol preserves the $\mathrm{SU}(d)$-irrep state while preparing a canonical pure state in the multiplicity subsystem. As an application, we consider quantum purity amplification (QPA), in which multiple noisy copies of a pure quantum state are combined to produce a state of higher purity. We show that the optimal purified state can be obtained by retaining a single designated output qudit and discarding the rest.