Weak Schur Sampling in Quantum Systems

• Weak Schur Sampling is a quantum measurement protocol that projects multipartite states onto irreducible representations defined by Schur–Weyl duality.
• WSS enables efficient quantum tomography, spectrum estimation, and metrology via streaming algorithms that reduce quantum memory and gate complexity.
• The protocol’s design, including classical simulatability in sparse regimes, serves as a benchmark for evaluating quantum advantage in state estimation tasks.

Weak Schur Sampling (WSS) is a quantum measurement protocol grounded in Schur–Weyl duality, enabling the extraction of global symmetry information from multipartite quantum systems. WSS plays a fundamental role in quantum tomography, spectrum estimation, and quantum information processing, leveraging the structure of irreducible representations of the symmetric and unitary groups acting on composite quantum systems. WSS protocols are defined both by their representation-theoretic properties and by efficient algorithmic realizations with implications for resource complexity and classical simulation (Cervero-Martín et al., 2024, Cervero et al., 2023, Havlicek et al., 2018).

1. Representation-Theoretic Foundations

In WSS, one considers a state $\rho$ on $(\mathbb{C}^d)^{\otimes n}$—that is, $n$ $d$-level quantum systems. By Schur–Weyl duality, the total Hilbert space decomposes as

$$(\mathbb{C}^d)^{\otimes n} \cong \bigoplus_{\lambda \vdash n, \ell(\lambda) \leq d}\; \mathcal{P}_\lambda \otimes \mathcal{Q}_\lambda^d$$

where $\lambda$ runs over Young diagrams (partitions of $n$ into at most $d$ parts), $\mathcal{P}_\lambda$ is the irreducible representation (irrep) of $S_n$ (the "permutation register"), and $\mathcal{Q}^d_\lambda$ the irrep of $U(d)$ (the "unitary register").

Weak Schur Sampling is defined as the projective measurement onto the isotypic subspaces labeled by $\lambda$. The POVM $\{\Pi_\lambda\}$, where

$$\Pi_\lambda = I_{\mathcal{P}_\lambda} \otimes I_{\mathcal{Q}^d_\lambda},$$

yields outcome $\lambda$ with probability $p(\lambda) = \operatorname{Tr}[\Pi_\lambda \rho]$. This measurement extracts only the irreducible symmetry type (the Young label) of the state, discarding multiplicity and finer register information (Cervero-Martín et al., 2024, Cervero et al., 2023, Havlicek et al., 2018).

2. Algorithmic Realizations and Streaming Protocols

Traditional WSS can be implemented by executing the Schur transform (a quantum circuit that changes basis to the joint irrep basis), followed by a measurement of the Young label $\lambda$. Standard approaches (e.g., Bacon–Chuang–Harrow, Generalized Phase Estimation) require $O(n)$ quantum memory and $\mathrm{poly}(n, d)$ gates.

Cervero–Mančinska introduced a logarithmic-space streaming WSS algorithm. At each step $k=1,\dots,n-1$, one consumes a fresh qudit, updates a small workspace (tracking the current irrep label $\lambda^{(k)} \vdash k$), and applies a Clebsch–Gordan isometry. The algorithm maintains only $O(\log_d n)$ qudits of memory and achieves total gate complexity $O(d\,n^{2d}\log_2^p(n^{2d}/\epsilon))$ for qudits, $O(n^3\log(n/\epsilon))$ for qubits (Cervero et al., 2023, Cervero-Martín et al., 2024). This yields an exponential quantum-memory reduction compared to previous circuits.

A further advance is presented in (Cervero-Martín et al., 2024), where a streaming unitary mixed Schur sampling protocol generalizes the task to $m$ "inputs" and $n$ "outputs", leveraging the mixed Schur–Weyl duality. The algorithm processes a stream of $m+n$ qudits, utilizing super–Clebsch–Gordan transforms and projective measurements on isotypic sectors, yielding exponential memory ($O(d^2\log^p(d, m, n, 1/\epsilon))$ qubits) and polynomial gate ($O((m+n)d^4\log^p(d,m,n,1/\epsilon))$) savings. For rank-constrained inputs, further reductions are achieved both in working space and gate count.

A notable physical realization of WSS for qubit systems is the random-SWAP-test protocol (Brahmachari et al., 7 Aug 2025), where $O(n\log(n/\epsilon))$ random two-qubit SWAP tests suffice to extract permutationally invariant information, matching the outcome statistics of the Schur transform.

3. Complexity and Resource Bounds

The table below contrasts the main algorithms for WSS:

Method	Quantum Memory	Gate Complexity	Streaming/Online Capability
Full Schur Transform	$O(n)$ (qudits)	$\mathrm{poly}(n,d,1/\epsilon)$	No
Generalized Phase Estimation	$O(n)$	$\mathrm{poly}(n,d)$	No
(Cervero et al., 2023) Streaming	$O(\log_d n)$	$O(d n^{2d}\log^p(n^{2d}/\epsilon))$	Yes
SWAP-testing (Brahmachari et al., 7 Aug 2025)	$O(n)$ (registers)	$O(n \log(n/\epsilon))$ (SWAPs)	Yes (randomized)

Memory requirement for streaming WSS: $O(\log_d n)$ qudits; for qubits: $O(\log_2 n)$ qubits. Gate complexity for qubits is $O(n^3\log(n/\epsilon))$ (Cervero et al., 2023); for general $d$, $O(d\,n^{2d}\log_2^p(n^{2d}/\epsilon))$ (Cervero-Martín et al., 2024). For reduced-rank inputs, both memory and gates scale as $O((r+r') d \log^p(d, m, n, 1/\epsilon))$ and $O((m+n)(r+r')^3 d\log^p(d, m, n, 1/\epsilon))$, where $r,r'$ are input/output subspace dimensions (Cervero-Martín et al., 2024).

4. Applications in Quantum Information and Metrology

WSS is instrumental in quantum state tomography, spectrum estimation, quantum metrology, and quantum hypothesis testing. In such applications, only the Young label $\lambda$ and the corresponding U(d) irrep register are informationally relevant, rendering permutation-invariant measurements optimal.

In quantum metrology settings, WSS enables rank, purity, and spectral gap hypothesis tests for unknown density matrices (Gardner et al., 19 Feb 2026). Sample complexity for WSS-based tests:

• Rank test: $M = \Theta(r_n^2 \log(1/\beta)/\theta)$.
• Purity test: $M = \Theta(\log(1/\beta)/\theta)$.
• Spectral gap test: $M = O(r^2 \log(1/\beta)/\min(\theta^2,\Lambda^2))$.

These exhibit no scaling with Hilbert space dimension, circumventing the Rayleigh curse, and vastly outperform full-state tomography ($M = O(rd \log(1/\beta)/\theta^2)$) for low-rank states (Gardner et al., 19 Feb 2026). WSS subroutines underpin quantum-enhanced superresolution, exoplanet detection, gravitational wave and dark matter spectroscopy, and Lindblad operator detection.

5. Classical Simulatability and Sparsity Regimes

Quantum Schur Sampling circuits, including WSS, can be efficiently simulated classically for qubits and, under a mild sparsity condition, for higher $d$. For $n$-qubit systems, both amplitude and weak sampling (i.e., drawing outputs distributed as $p(\lambda)$) can be achieved in $\mathrm{poly}(n,1/\epsilon)$ time (Havlicek et al., 2018). The algorithm computes inner products of Schur basis states using products of Clebsch–Gordan coefficients, enabling construction of conditional marginals and chain-rule samplers.

Further, if the Schur output distribution is $\epsilon$-approximately $t$-sparse ($t=\mathrm{poly}(n)$), there exists a classical algorithm that produces samples within total variation distance $6\epsilon$ of the true WSS distribution, with overall complexity $\mathrm{poly}(n,1/\epsilon,t)$ (Havlíček et al., 2018). This result refutes prior conjectures regarding the classical hardness of Permutational Quantum Computing for sparse-output regimes. Only genuinely high-entropy output regimes remain potentially classically hard.

6. Extensions: Unitary and Mixed Schur Sampling

Unitary Schur Sampling generalizes WSS by retaining, along with the Young label, the reduced density matrix on the $U(d)$ irrep register, discarding the permutation register. Formally, for input $\rho$ the output is $$\{ \lambda, \sigma_\lambda = \operatorname{Tr}_{\mathcal{P}_\lambda} [\Pi_\lambda \rho \Pi_\lambda] \}$$, with probability $p(\lambda) = \operatorname{Tr}[\Pi_\lambda \rho]$ (Cervero-Martín et al., 2024). In many quantum protocols, e.g., spectrum estimation and tomography, only the $Q^d_\lambda$ register is required.

Mixed Schur Sampling (and unitary mixed Schur sampling) further generalizes the protocol to inputs on $$\mathcal{H} = (\mathbb{C}^d)^{\otimes m} \otimes (\mathbb{C}^{d*})^{\otimes n}$$, using the mixed Schur–Weyl transform and walled-Brauer algebra. Output states correspond to staircases $\gamma$ (partitioning $(m, n)$), and the protocol retains only the $U(d)$-irrep sector, with streaming algorithms generalizing the memory and gate efficiency benefits to the mixed setting (Cervero-Martín et al., 2024).

7. Implications, Limitations, and Outlook

WSS protocols are foundational both theoretically and as practical subroutines in quantum algorithms. Streaming and low-memory architectures bring these tasks within reach for near-term devices and large-scale systems. Classical simulatability results delimit the boundaries of possible quantum advantage: in regimes where Schur-output distributions are approximately sparse, no exponential quantum-over-classical sampling separation is possible (Havlicek et al., 2018, Havlíček et al., 2018). A plausible implication is that further quantum advantage from WSS-derived circuits may require leveraging non-sparse regimes or incorporating additional non-simulable gate structures.

Recent advances generalizing WSS to the mixed Schur–Weyl and reduced-rank input settings suggest that improvements in quantum resource requirements will continue, further enhancing the applicability of WSS in scalable quantum state estimation, metrology, and beyond (Cervero-Martín et al., 2024, Cervero et al., 2023).