A Random Matrix Theory of Pauli Tomography

Nathan Keenan, John Goold, Alex Nico-Katz · 2025

Read A Random Matrix Theory of Pauli Tomography by Nathan Keenan, John Goold, Alex Nico-Katz (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-12010.txt.

Introduction

Nathan Keenan, John Goold, and Alex Nico-Katz bring random matrix theory to quantum state checks. Their paper, from twenty twenty-five, studies tomography, the rebuild of an unknown quantum state from many measures. Knowing the size and shape of rebuild errors is key for quantum machines. The authors treat the error itself as a random matrix.

They find a deep link between the error and a classic random matrix family. That link lets them swap real errors for samples from the family in many formulas. They derive mean error, spread, and a bound on how many samples a check needs. Numeric tests on common methods back the theory. A standard fix-up step leaves the bound unchanged.

Catalog record

CREATORNathan Keenan, John Goold, Alex Nico-Katz
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE63 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-12010

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A Random Matrix Theory of Pauli Tomography Nathan Keenan, John Goold, Alex Nico-Katz Abstract Quantum state tomography (QST), the process of reconstructing some unknown quantum state ρ ^ \hat{\rho} from repeated measurements on copies of said state, is a foundationally important task in the context of quantum computation and simulation. For this reason, a detailed characterization of the error Δ ​ ρ ^ = ρ ^ − ρ ^ ′ \Delta\hat{\rho}=\hat{\rho}-\hat{\rho}^{\prime} in a QST reconstruction ρ ^ ′ \hat{\rho}^{\prime} is of clear importance to quantum theory and experiment. In this work, we develop a fully random matrix theory (RMT) treatment of state tomography in informationally-complete bases; and in doing so we reveal deep connections between QST errors Δ ​ ρ ^ \Delta\hat{\rho} and the gaussian unitary ensemble (GUE). By exploiting this connection we prove that wide classes of functions of the spectrum of Δ ​ ρ ^ \Delta\hat{\rho} can be evaluated by substituting samples of an appropriate GUE for realizations of Δ ​ ρ ^ \Delta\hat{\rho} . This powerful and flexible result enables simple analytic treatments of the mean value and variance of the error as quantified by the trace distance ‖ Δ ​ ρ ^ ‖ Tr \|\Delta\hat{\rho}\|_{\mathrm{Tr}} (which we validate numerically for common tomographic protocols), allows us to derive a bound on the QST sample complexity, and subsequently demonstrate that said bound doesn’t change under the most widely-used rephysicalization procedure. These results collectively demonstrate the flexibility, strength, and broad applicability of our approach; and lays the foundation for broader studies of RMT treatments of QST in the future. I Introduction Quantum state tomography (QST) Bagan et al. 2004 ; Cai et al. 2016 ; Keyl 2006 ; Guţă et al. 2008 is the process of experimentally reconstructing an unknown quantum state \hat{\rho} from the outcomes of repeated measurements on copies of said state. Efficient and precise implementation of QST is a foundational task in quantum computation and simulation: state verification, experimental determination of non-trivial functions of \hat{\rho} Nico-Katz et al. 2024 , and channel tomography Yu 2023 ; Huang et al. 2024 (via e.g. the Choi-Jamiłkowski isomorphism), all hinge on high-fidelity QST. Moreover, QST bears importance in the context of quantum foundations: the ontological state \hat{\rho} is fundamentally inaccessible to any ‘classical’ process, and hence there exist limitations on how accurately (implicit) phenomenal tomographic reconstructions can reflect reality. Thus, the behaviour (and bounding) of errors in QST, here quantified by the trace distance \|\hat{\rho}-\hat{\rho}^{\prime}\|_{\mathrm{Tr}}=\|\hat{\Delta}\hat{\rho}\|_{\mathrm{Tr}} , are of basic importance across theoretical, experimental, and computational quantum physics. In this work, we address this problem using a complete random matrix theory treatment of tomographic processes. In doing so we reveal deep connections between states produced by tomographic protocols and the gaussian unitary ensemble (GUE), and develop and deploy a powerful method for evaluating functions over empirical spectral measures. In this work, we are solely concerned with complete, single-copy, fixed shot count QST: in which we seek to reconstruct \hat{\rho} as completely as possible given access to exactly one copy of \rho for any single measurement Flammia et al. 2012 ; Lowe and Nayak 2022 . This represents the (widely used) prototypical form of QST deployed in e.g. native QST on IBM devices. Examples of other QST protocols include those that allow simultaneous measurements on multiple entangled copies of \rho Yuen 2023 ; Bubeck et al. 2020 ; O’Donnell and Wright 2016 ; O’Donnell and Wright 2017 ; Haah et al. 2016 ; Chen et al. 2024 ; Flammia and O’Donnell 2024 , protocols that partially reconstruct the state such as shadow tomography Akhtar et al. 2022 ; Zhang et al. 2024 ; Sengupta et al. 2025 ; Rath et al. 2021 ; Aaronson 2018 ; Elben et al. 2023 ; Chen et al. 2022a ; Chen et al. 2024 ; Huang et al. 2020 and quantum overlapping tomography García-Pérez et al. 2020 ; Cotler and Wilczek 2020 , and state verification of \hat{\rho}^{\prime} against known Bădescu et al. 2019 ; Chen et al. 2022b ; Liu and Acharya 2024 ; Guţă et al. 2020 or unknown Yu 2023 ontological quantum states \hat{\rho} . For a more comprehensive overview of possible tomographic protocols (and related topics), we refer the reader to Ref. Acharya et al. 2025 . Figure 1: (a) Eigenvalue distributions of naive tomographic excess \Delta\hat{\rho} for different shot counts S , overlayed with the GUE predicted Wigner semi-circles of radius 2/\sqrt{S} . (b) Schematic for tomographic procedure. An a priori state is measured in an IC-basis, which leads to coefficients \vec{a} . The basis expansion map \mathcal{A}(\vec{a}) then gives us an estimate \hat{\rho}^{\prime} of \hat{\rho} . If we then want to ensure that our estimator for \hat{\rho} is physical, we project back to \mathrm{phys}_{D} via \Pi which depends on the rephysicalization scheme. We first introduce and motivate the basic elements of the full single-copy QST protocols we consider in Section II . In Section III , we introduce the GUE measure and develop a series of theorems which connect it to full QST in a fixed (symmetric) informationally-complete basis; this forms the basis of our full RMT treatment of tomographic errors. In Section IV we develop two more theorems which bound the difference between Lipschitz functions over the one-point and two-point empirical spectral measures of the tomographic excess \Delta\hat{\rho} , and identical functions evaluated over a (rescaled) GUE spectral measure; we then demonstrate that these distances vanish for both a naïve QST protocol (in Section IV.1 ) and a widely used more sophisticated QST protocol (in Section IV.2 ). This is the major result of our work: that for common tomographic protocols, one can substitute random samples from the GUE for the random error matrix \Delta\hat{\rho} in calculations of functions of the QST error. Equipped with this powerful and generic result, we proceed to: calculate analytic values for the expected value \mathbb{E}[\|\Delta\rho\|_{\mathrm{Tr}}] and variance \mathrm{var}[\|\Delta\rho\|_{\mathrm{Tr}}] of QST errors in Section V.1 , validate these derived values numerically in Section V.2 , validate a recently calculated bound (c.f. Ref. Acharya et al. 2025 ) on the QST sample complexity in Section V.3 , and finally demonstrate that the most widely-used process of state rephysicalization (i.e. artificial enforcement of positivity after determination of \hat{\rho} ) does not improve the existing complexity bound in Section V.4 . In doing so, we demonstrate the strength and general flexibility of our RMT approach to QST, and conclude by speculating on further possible avenues of research in Section V.5 . II Motivation Given an N -qubit system, an arbitrary state of the system \rho\in\mathbb{C}^{D\times D} (where D=2^{N} is the total dimension) can be written in the orthogonal Pauli basis as \rho=\frac{1}{D}\sum_{j}c_{j}\hat{P}_{j}, (1) where c_{j}=\mathrm{Tr}(\rho P_{j}) , and P_{j} runs over all the N -length Pauli strings \mathcal{P}_{N}=\{\mathbb{I},X,Y,Z\}^{\otimes N} . Full state Pauli tomography consists of performing measurements in the basis defined by elements of \mathcal{P}_{N} on copies of the state \hat{\rho} Cai et al. 2016 ; Gross et al. 2010 ; Acharya et al. 2025 . Since the basis \mathcal{P}_{N} is informationally-complete (IC) (i.e. spans the physical state space), an unbiased estimator \hat{\rho}^{\prime} for \hat{\rho} can then be constructed: \hat{\rho}^{\prime}=\frac{1}{D}\sum_{j}a_{j}\hat{P}_{j} (2) where the a_{j} are ultimately derived from the measurement data (see Section IV.1 and

Read the complete text