Pith. sign in

REVIEW 1 major objections 4 minor 3 cited by

Near-optimal performance of square-root measurement for general score functions and quantum ensembles

T0 review · 1 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper proves that for any quantum ensemble—continuous or discrete, finite- or infinite-dimensional—a generalized pretty good measurement is near-optimal, with mean square error at most twice the optimum.

desk verdict A rigorous and useful extension of Barnum-Knill to continuous ensembles and infinite-dimensional systems; the positivity restriction on score functions is explicit and the proofs hold up. read the letter →

arxiv 2505.20447 v1 pith:KT44YMXQ submitted 2025-05-26 quant-ph math-phmath.FAmath.MP

Hemant K. Mishra , Ludovico Lami , Mark M. Wilde This is my paper · ORCID
classification quant-phmath-phmath.FAmath.MP MSC 81P50
keywords quantumstatediscriminationprettygoodmeasurementsquare-rootBarnum–KnilltheoremBayesianestimationmeansquareerrorpositivescorefunctionscontinuousensembles
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

For any quantum ensemble—finite or infinite-dimensional, discrete or continuous in its parameter space—the paper claims that a canonical measurement built from the ensemble's average state, the generalized pretty good measurement (GPGM), is near-optimal. The main theorem states that for every score function that factors as a convolution of a bounded function with itself, the optimal expected gain is at most the square root of the GPGM's expected gain. As a direct consequence, in Bayesian estimation with mean square error, the GPGM's expected error is at most twice the minimum achievable. This matters because it extends the Barnum–Knill guarantee, previously confined to finite ensembles, to the continuous and infinite-dimensional settings where success probability is meaningless and practical estimation problems live.

What carries the argument

The load-bearing object is the pair (generalized pretty good measurement, expected gain) together with the factorization identity for positive score functions. The GPGM, $P^{\mathrm{PG}}(A)=\Lambda_A^\dagger\Lambda_A+\mu(A)\Pi_{\ker\rho}$, extends the square-root measurement to arbitrary alphabets by replacing the Moore–Penrose inverse and the discrete states with a contraction $\Lambda_A$ selected by the range condition $\Lambda_A\rho^{1/2}=\rho_A^{1/2}$. The expected gain $G_{\mathcal{E},s}(\Pi)=\int\int s(x,\hat x)\,\operatorname{Tr}[\rho_x\,\Pi(d\hat x)]\,\mu(dx)$ is a figure of merit that reduces to the success probability when the parameter set is finite and $s$ is the Kronecker delta. The factorization $s = f\star_\nu f$, i.e. $s(x,y)=\int f(x,z)f(y,z)\,\nu(dz)$, is the identity that lets the proof apply Cauchy–Schwarz after inserting a square root of the average state, and the two new integration theorems (Theorems A.1 and A.4) guarantee that the resulting integrals of operator-valued measures remain trace-class or Hilbert–Schmidt as needed.

What would settle it

For a finite parameter set $\mathcal{X}=\{1,2\}$ with two slightly non-orthogonal states, choose a $2\times 2$ score matrix $s$ that is not positive semi-definite (one negative eigenvalue) and compute both the optimal expected gain and the GPGM expected gain; a violation of $G^{\star} \le \sqrt{G^{\rm PG}}$ would show that the positivity condition in Definition 3.2 is necessary rather than merely sufficient.

Watch

Extended reading notes

Core claim

The paper's central discovery is a generalization of the Barnum–Knill theorem: for any quantum ensemble $\mathcal{E} = ((\mu(dx), \rho_x))$ on a complex separable Hilbert space and any positive score function $s(x,y) = \int f(x,z)f(y,z)\,\nu(dz)$ with bounded measurable $f$, the optimal expected gain satisfies $G_{\mathcal{E},s}^{\star} \le \sqrt{G_{\mathcal{E},s}^{\mathrm{PG}}}$, where $G^{\mathrm{PG}}$ is the expected gain of the generalized pretty good measurement. The GPGM is constructed from the ensemble's average state $\rho$: for each Borel set $A$ it is $P^{\mathrm{PG}}(A) = \Lambda_A^\dagger \Lambda_A + \mu(A)\Pi_{\ker \rho}$, with $\Lambda_A$ the unique contraction obeying $\Lambda_A \rho^{1/2} = \rho_A^{1/2}$ and $\ker\rho \subseteq \ker\Lambda_A$. A corollary (Theorem 3.6) is that the expected mean square error of the GPGM is no more than twice the optimal mean square error in Bayesian estimation, for any ensemble parameterized by $\mathbb{R}^d$ with a prior of finite second moment. This is the first near-optimality guarantee of this type that covers continuous parameter spaces and infinite-dimensional Hilbert spaces.

Load-bearing premise

The load-bearing premise is that the score function factors as $s(x,y)=\int f(x,z)f(y,z)\,\nu(dz)$ for some bounded measurable $f$; only for such positive score functions is the square-root inequality proved, and non-positive kernels fall outside the claim.

Editorial extensions

If this is right

  • The original Barnum–Knill theorem is recovered exactly: for finite parameter sets, discrete priors, and the identity score function, Theorem 3.2 reduces to the inequality $\sup p_{\rm succ} \le \sqrt{p_{\rm succ}^{\rm PGM}}$.
  • For Bayesian estimation with mean square error, the GPGM is a universal near-optimal estimator: ${\rm MSE}(\mathcal{E}, P^{\rm PG}) \le 2\,{\rm MSE}(\mathcal{E}, \Pi)$ for every POVM $\Pi$.
  • For bosonic Gaussian ensembles, where the GPGM is already known to be a Gaussian measurement with an explicit mean square error, the new bound gives a provable near-optimality guarantee without any numerical search.
  • The two integration theorems (A.1 and A.4) are standalone tools: any estimation proof that needs to integrate bounded measurable functions against trace-class or Hilbert–Schmidt operator-valued measures can now do so with the trace inside the integral.
  • The positivity condition defines a class of admissible figures of merit, so the result extends beyond success probability and mean square error to any estimation loss expressible as a positive kernel.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: the factor of two in Theorem 3.6 is a worst-case constant; for translation-invariant ensembles the GPGM may be much closer to optimal, and numerical checks on bosonic Gaussian ensembles could reveal a smaller universal constant.
  • Editorial inference: the restriction to positive score functions marks the boundary of the proof technique; testing non-positive-definite kernels on small finite ensembles would show whether the square-root bound actually fails outside that class.
  • Editorial inference: because the proof only uses the factorization and operator-valued integration machinery, analogous square-root bounds may hold for other loss functions (e.g., polynomial or exponential costs) that admit a positive factorization.
  • Editorial inference: the result suggests a practical prescription for continuous-variable metrology: use the GPGM (often a standard Gaussian measurement) and one is guaranteed to be within a factor of two of the optimal estimator, without solving the generally hard optimization over POVMs.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

1 major / 4 minor

Summary. The paper extends the pretty good measurement (PGM) and the Barnum–Knill theorem from finite discrete ensembles on finite-dimensional Hilbert spaces to arbitrary quantum ensembles, including continuous parameter spaces and infinite-dimensional Hilbert spaces. It constructs a generalized pretty good measurement (GPGM) canonically from the ensemble, proves it is a POVM (Theorem 3.1), and proves a generalized Barnum–Knill inequality for positive score functions: the optimal expected gain is at most the square root of the expected gain of the GPGM (Theorem 3.2). As an application, it shows that in Bayesian parameter estimation with mean square error, the GPGM's expected MSE is at most twice the optimal MSE (Theorem 3.6), via a Gaussian-score limiting identity (Lemma 3.4). The appendices supply functional-analytic integration tools for trace-class and Hilbert-Schmidt-valued operator measures.

Significance. This is a substantial generalization of one of the standard tools in quantum state discrimination and quantum estimation. The construction is canonical, the main theorem is proved with detailed functional-analytic arguments rather than formal manipulations, and the restriction to positive (factorizable) score functions is stated explicitly and verified for Gaussian kernels in Proposition C.1. The corollary for Bayesian estimation, with a clean factor-of-two bound, is likely to be useful for bosonic Gaussian ensembles and continuous-variable settings. If the technical point about σ-finiteness in the theorem's statement is settled, I regard the contribution as a strong and welcome one.

major comments (1)
  1. [Definition 3.2 and Eqs. (3.30), (3.46)] The proof of Theorem 3.2 invokes Fubini's theorem for the measure ν on the auxiliary space Z, but Definition 3.2 does not assume ν to be σ-finite. The standard Fubini theorem used in the proof (Cohn, Theorem 5.2.2) requires σ-finiteness of the measure spaces. Since all examples in the paper (finite alphabets, probability measures, Lebesgue measure for Gaussian scores) are σ-finite, this is likely a missing hypothesis rather than a conceptual flaw. Please add σ-finiteness of ν to Definition 3.2, or supply an argument that the relevant integrand is supported on a σ-finite subset of Z so that Fubini applies on that subset. Without this, Theorem 3.2 is not proved in the full generality in which it is stated.
minor comments (4)
  1. [Proposition 3.5, Eqs. (3.76)-(3.79)] The identities in these equations are used for the unbounded kernel ||\hat x||^2, whereas Theorem A.4 and the proof of Theorem 3.2 are stated for bounded score functions. Please add a truncation and monotone-convergence argument, or state the underlying bimeasure identity Tr[ℓ(dx)ℓ(dx')]=Tr[ρ_x' A_PG(dx)] μ(dx') separately, so that the unbounded case is justified.
  2. [Theorem A.4, Eq. (A.34)] In Eq. (A.34), the displayed formula introduces the measure as 'Tr[A(·)A(·)]', but the theorem is about an operator-valued measure ℓ; the notation should be made consistent.
  3. [Section 3.3.1, Eq. (3.57)] The phrase 'probability density Tr[ρ_x A(d\hat x)]' is imprecise, since A is a measure and in general is not absolutely continuous; it should read 'probability measure' or 'probability distribution'.
  4. [Definition 3.2] The paper calls the score functions 'positive' although the factor f is only real-valued; a remark that positivity refers to the resulting kernel s(x,y)=∫ f(x,z)f(y,z)ν(dz) being positive semidefinite would help the reader.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity identified: the generalized Barnum–Knill inequality is proved from the stated factorization assumption and external integration theory, with no fitted parameters or load-bearing self-citations.

full rationale

The paper's central claim, Theorem 3.2, is not circular. The GPGM in Definition 3.1 is constructed solely from the ensemble via the operators Lambda_A determined by Lambda_A rho^{1/2} = rho_A^{1/2}; it does not depend on the score function. The inequality G_{E,s} <= sqrt(G_{E,s}^{PG}) is therefore not an identity forced by the construction of the measurement. The proof uses the factorization s = f *_nu f from Definition 3.2 exactly at Eq. (3.29), then Fubini and Cauchy-Schwarz steps (3.30)-(3.43) together with the trace/Bochner integration theorems in Appendices A and B. This is a genuine derivation, and the restriction to positive score functions is an explicit hypothesis, not a hidden restatement of the conclusion. The MSE application is likewise self-contained: Lemma 3.4 is an exact pointwise limiting identity (Eqs. (3.60)-(3.67)), Proposition C.1 verifies that Gaussian score functions satisfy Definition 3.2, and Theorem 3.6 passes the inequality of Theorem 3.2 through that limit. No parameter is fitted to data or to the target inequality. The self-citations [7] and [21] appear only as literature context and as illustrations of previously known Gaussian-ensemble facts; they are not used to justify Theorem 3.2 or Theorem 3.6. The proof relies on external references [26], [28], and [29] for standard facts, and the stated limitation (the positivity/factorizability condition on score functions) is an assumption that narrows the scope of the theorem rather than a circular step.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central result rests on standard functional analysis and a stated positivity condition on score functions, with no fitted parameters.

assumptions (5)
  • domain assumption Borel measurability and Bochner integrability of the ensemble map x -> ρ_x (Eq. (2.1)).
    Standard framework for continuous quantum ensembles; needed for the average state ρ to be a well-defined trace-class operator.
  • domain assumption Score function is positive, i.e., factorizable as s = f ★ f with f bounded (Definition 3.2).
    Load-bearing: the proof of Theorem 3.2 substitutes this factorization at Eq. (3.29).
  • standard math Douglas lemma (Ex. 10.8.8 in [26]) ensuring existence of contraction Λ_A with Λ_A ρ^{1/2} = ρ_A^{1/2} whenever ρ_A ≤ ρ.
    Used to define GPGM in Definition 3.1.
  • standard math Operator-valued integration theory (Theorems A.1, A.4) for trace-class and Hilbert-Schmidt valued measures.
    Proved in appendices; relies on [29] and semivariation arguments.
  • domain assumption Finite second moment of the prior μ for the MSE finiteness result (Proposition 3.5).
    Needed for MSE(GPGM) to be finite.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Near-optimal performance of square-root measurement for general score functions and quantum ensembles." pith.science (2026). https://pith.science/paper/KT44YMXQ

@misc{pith2026250520447,
  author       = {Pith},
  title        = {Pith review of: Near-optimal performance of square-root measurement for general score functions and quantum ensembles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KT44YMXQ}},
  note         = {Machine review of arXiv:2505.20447}
}
read the original abstract

The Barnum-Knill theorem states that the optimal success probability in the multiple state discrimination task is not more than the square root of the success probability when the pretty good or square-root measurement is used for this task. An assumption of the theorem is that the underlying ensemble consists of finitely many quantum states over a finite-dimensional quantum system. Motivated in part by the fact that the success probability is not a relevant metric for continuous ensembles, in this paper we provide a generalization of the notion of pretty good measurement and the Barnum-Knill theorem for general quantum ensembles, including those described by a continuous parameter space and an infinite-dimensional Hilbert space. To achieve this, we also design a general metric of performance for quantum measurements that generalizes the success probability, namely, the expected gain of the measurement with respect to a positive score function. A notable consequence of the main result is that, in a Bayesian estimation task, the mean square error of the generalized pretty good measurement does not exceed twice the optimal mean square error.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Dimension-Free Polylogarithmic Quantum Shadow Tomography from Sequential Pretty-Good Measurements

    quant-ph 2026-08 conditional novelty 8.0 of 10

    New sequential pretty-good measurement protocol achieves dimension-free shadow tomography with sample complexity O(1/eps^2 * (log(M/delta))^4 / (log log(M/delta))^3).

  2. Measurement incompatibility in Bayesian multiparameter quantum estimation

    quant-ph 2025-11 accept novelty 7.0 of 10

    Measurement incompatibility at most doubles the minimum mean-square loss in Bayesian multiparameter quantum estimation, relative to the symmetric-posterior-mean bound.

  3. Unifying quantum measurement constructions via a relative-entropy minimum change principle

    quant-ph 2026-08 conditional novelty 6.0 of 10

    A relative-entropy minimum change principle yields a unified closed-form family of optimal measurements, including pretty good, Fermi-Dirac thermal, and new softmin thermal measurements.

Reference graph

Works this paper leans on

31 extracted references · 14 canonical work pages · cited by 3 Pith papers

  1. [1]

    Helstrom

    Carl W. Helstrom. Quantum detection and estimation theo ry. Journal of Statistical Physics , 1:231–252, June 1969. http://dx.doi.org/10.1007/BF01007479

    doi:10.1007/bf01007479 1969

  2. [2]

    Alexander S. Holevo. Investigations in the general theo ry of statistical decisions. Trudy Matematicheskogo Instituta imeni V.A. Steklova , 124:3–140, 1976. (English translation: Proceedings of the Steklov Institute of Mathematics 124:1–140, 1978)

    work page 1976

  3. [3]

    Barnett and Sarah Croke

    Stephen M. Barnett and Sarah Croke. Quantum state discri mination. Advances in Optics and Photonics , 1(2):238–278, April 2009. http://dx.doi.org/10.1364/AOP.1.000238. arXiv:0810.1970

    arXiv 2009

  4. [4]

    Conclusive exclusion of quantu m states

    Somshubhro Bandyopadhyay, Rahul Jain, Jonathan Oppenh eim, and Christopher Perry. Conclusive exclusion of quantu m states. Physical Review A, 89:022336, February 2014. http://dx.doi.org/10.1103/PhysRevA.89.022336. arXiv:1306.4683

    arXiv 2014

  5. [5]

    Antidistinguishab ility of pure quantum states

    Teiko Heinosaari and Oskari Kerppo. Antidistinguishab ility of pure quantum states. Journal of Physics A: Mathematical and Theoretical , 51(36):365303, July 2018. http://dx.doi.org/10.1088/1751-8121/aad1fc. arXiv:1804.10457. Near-optimal performance of square-root measurement for general score functions and quantum ensembles 20

    arXiv 2018

  6. [6]

    Inner products of pure st ates and their antidistinguishability

    Vincent Russo and Jamie Sikora. Inner products of pure st ates and their antidistinguishability. Physical Review A, 107:L030202, March

    work page

  7. [7]

    Mishra, Michael Nussbaum, and Mark M

    Hemant K. Mishra, Michael Nussbaum, and Mark M. Wilde. On the optimal error exponents for classical and quantum antidistinguisha- bility. Letters in Mathematical Physics, 114(3):76, June 2024. http://dx.doi.org/10.1007/s11005-024-01821-z. arXiv:2309.03723

    arXiv 2024

  8. [8]

    Optimal distinction of non-ortho gonal quantum signals

    Viacheslav Belavkin. Optimal distinction of non-ortho gonal quantum signals. Radio Engineering and Electronic Physics, 20:39–47, 1975

    work page 1975

Show all 31 references
  1. [9]

    Optimal multiple quantum statist ical hypothesis testing

    Viacheslav Belavkin. Optimal multiple quantum statist ical hypothesis testing. Stochastics, 1:315–345, 1975

    1975

  2. [10]

    Alexander S. Holevo. On asymptotically optimal hypoth esis testing in quantum statistics. Theory of Probability & Its Applications, 23(2): 411–415, 1979. http://dx.doi.org/10.1137/1123048

    1979 doi

  3. [11]

    pretty good

    Paul Hausladen and William K. Wootters. A “pretty good”measurement for distinguishing quantum states. Journal of Modern Optics, 41 (12):2385–2390, December 1994. http://dx.doi.org/10.1080/09500349414552221

    1994 doi

  4. [12]

    Reversing quantum dy namics with near-optimal quantum and classical fidelity

    Howard Barnum and Emmanuel Knill. Reversing quantum dy namics with near-optimal quantum and classical fidelity. Journal of Mathematical Physics, 43(5):2097–2106, May 2002. http://dx.doi.org/10.1063/1.1459754. arXiv:quant-ph/0004088

    2002

  5. [13]

    Eldar and G.D

    Y.C. Eldar and G.D. Forney. On quantum detection and the square-root measurement. IEEE Transactions on Information Theory, 47(3): 858–872, March 2001. http://dx.doi.org/10.1109/18.915636. arXiv:quant-ph/0005132

    2001 arXiv

  6. [14]

    Eldar, A

    Y.C. Eldar, A. Megretski, and G.C. Verghese. Designing optimal quantum detectors via semidefinite programming. IEEE Transactions on Information Theory, 49(4):1007–1012, April 2003. http://dx.doi.org/10.1109/TIT.2003.809510. arXiv:quant-ph/0205178

    2003

  7. [15]

    Optimal mea surements for the dihedral hidden subgroup problem

    Dave Bacon, Andrew M Childs, and Wim van Dam. Optimal mea surements for the dihedral hidden subgroup problem. Chicago Journal of Theoretical Computer Science, 2006(2), October 2006. http://dx.doi.org/10.4086/cjtcs.2006.002. arXiv:quant-ph/0501044

    2006 arXiv

  8. [16]

    pretty good

    Carlos Mochon. Family of generalized “pretty good” mea surements and the minimal-error pure-state discriminatio n prob- lems for which they are optimal. Physical Review A , 73:032328, March 2006. http://dx.doi.org/10.1103/PhysRevA.73.032328. arXiv:quant-ph/0506061

    2006 arXiv

  9. [17]

    Optimality of the pretty good measurem ent for port-based teleportation

    Felix Leditzky. Optimality of the pretty good measurem ent for port-based teleportation. Letters in Mathematical Physics , 112(5):98, October 2022. http://dx.doi.org/10.1007/s11005-022-01592-5 . arXiv:2008.11194

    2022 arXiv

  10. [18]

    Alexander S. Holevo. Gaussian maximizers for quantum G aussian observables and ensembles. IEEE Transactions on Information Theory, 66(9):5634–5641, April 2020. http://dx.doi.org/10.1109/TIT.2020.2987789. arXiv:1908.03038

    2020

  11. [19]

    Alexander S. Holevo. On the classical capacity of gener al quantum Gaussian measurement. Entropy, 23(3):377, March 2021. http://dx.doi.org/10.3390/e23030377. arXiv:2101.00423

    2021 arXiv

  12. [20]

    Alexander S. Holevo. Accessible information of a gener al quantum Gaussian ensemble. Journal of Mathematical Physics, 62(9):092201, September 2021. http://dx.doi.org/10.1063/5.0048112. arXiv:2102.01981

    2021 arXiv

  13. [21]

    Mishra, Ludovico Lami, Prabha Mandayam, and M ark M

    Hemant K. Mishra, Ludovico Lami, Prabha Mandayam, and M ark M. Wilde. Pretty good measurement for bosonic Gaussian e nsem- bles. International Journal of Quantum Information , 22(5):2440010, August 2024. http://dx.doi.org/10.1142/S0219749924400100. arXiv:2303.04949

    2024 arXiv

  14. [22]

    Frigyes Riesz and Béla Sz. Nagy. Functional Analysis , volume 41. Blackie & Son Limited, London and Glasgow, 1956. http://dx.doi.org/10.2307/3611560

    1956 doi

  15. [23]

    Holevo and Maksim E

    Alexander S. Holevo and Maksim E. Shirokov. Continuous ensembles and the capacity of infinite-dimensional quantum channels. The- ory of Probability & Its Applications, 50(1):86–98, 2006. http://dx.doi.org/10.1137/S0040585X97981470. arXiv:quant-ph/0408176

    2006 arXiv

  16. [24]

    Shirokov

    Maksim E. Shirokov. On average output entropy of a quant um channel. 2024. arXiv:2404.07978

    2024 arXiv

  17. [25]

    Alexander S. Holevo. Statistical decision theory for q uantum systems. Journal of Multivariate Analysis , 3(4):337–394, 1973. ISSN 0047-259X. http://dx.doi.org/10.1016/0047-259X(73)90028-6

    1973 doi

  18. [26]

    Unbounded Self-adjoint Operators on Hilbert Space

    Konrad Schmüdgen. Unbounded Self-adjoint Operators on Hilbert Space . Springer, 2012. http://dx.doi.org/10.1007/978-94-007-4753-1

    2012 doi

  19. [27]

    Analysis in Banach Spaces

    Tuomas Hytönen, Jan Van Neerven, Mark Veraar, and Lutz W eis. Analysis in Banach Spaces . Springer, 2016. http://dx.doi.org/10.1007/978-3-319-48520-1

    2016 doi

  20. [28]

    Donald L. Cohn. Measure Theory, volume 2. Birkhäuser Basel, 2013. http://dx.doi.org/10.1007/978-1-4614-6956-8

    2013 doi

  21. [29]

    S. K. Mitter and S. K. Young. Integration with respect to operator-valued measures with applications to quantum est imation theory. Annali di Matematica pura ed applicata, 137:1–39, December 1984. http://dx.doi.org/10.1007/BF01789387. Near-optimal performance of square-root me...

    1984 doi

  22. [30]

    Positive Definite Matrices

    Rajendra Bhatia. Positive Definite Matrices. Princeton University Press, 2009. Appendix A Integration with respect to operator-valued mea sures Let /u1D4B3be a locally compact Hausdorff space, let /u1D707be a Borel probability measure on /u1D4B3, and let ℋ be a complex separab...

    2009

  23. [2023]

    arXiv:2206.08313

    http://dx.doi.org/10.1103/PhysRevA.107.L030202. arXiv:2206.08313

    arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.