Pith. sign in

REVIEW 3 major objections 4 minor 117 references

Enhancing Noisy Quantum Sensing by GHZ State Partitioning

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read For a fixed number of noisy sensors, the best strategy is to partition them into m* ≈ −n ln k independent GHZ sub-ensembles, which beats one monolithic GHZ state by an exponential quantum Fisher information advantage.

desk verdict Useful closed-form resource-allocation rules for noisy GHZ sensing, built on a model the authors state plainly; the formulas verify against numerics and deserve a referee, with requests for error bounds and a correlated-noise caveat. read the letter →

arxiv 2507.02829 v1 pith:2KQYNYJD submitted 2025-07-03 quant-ph

Allen Zang , Tian-Xing Zheng , Peter C. Maurer , Frederic T. Chong , Martin Suchara , Tian Zhong This is my paper · ORCID
classification quant-ph
keywords GHZstatesquantumFisherinformationentanglement-enhancedsensingphaseestimationdepolarizingnoiseparticlelossdephasingresourceallocation
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

The paper proposes and analyzes a resource-allocation strategy for entanglement-enhanced phase estimation under realistic noise: instead of preparing one GHZ state across all n sensors, split the sensors into m independent, smaller GHZ sub-ensembles. For state-preparation errors modeled by an entangling-gate fidelity k, the optimal number of sub-ensembles is $m^*\approx -n\ln k$, so each sub-ensemble holds about $1/(1-k)$ sensors. This partitioning turns what would otherwise be a noise-limited monolithic sensor into a sensor with an exponentially larger quantum Fisher information, by a factor roughly $-(1/e)\,k^{-n}/(n\ln k)$. The same logic is extended to particle loss, where $k$ is replaced by $kp$, and to dephasing, where $k$ is replaced by $k(2q-1)^2$, and it is applied to time-dependent and sequential sensing protocols. If correct, the results reduce noisy entanglement-enhanced sensing design to simple closed-form rules.

What carries the argument

The argument rests on the depolarized-GHZ preparation model $\rho = V(n)|\mathrm{GHZ}_n\rangle\langle\mathrm{GHZ}_n| + [1-V(n)] I_n/2^n$ with fidelity $F(n)=k^{n-1}F$, together with the closed-form quantum Fisher information for GHZ-diagonal states under z-rotation encoding. From that, the paper derives the QFI of a monolithic n-qubit GHZ sensor and uses additivity of the QFI under tensor products to write the partitioned QFI as $m\,F(F,k,n/m)$. The optimal partition number $m^*$ is then obtained by setting $\partial F/\partial m = 0$ and keeping leading-order terms for $k,F$ close to 1, giving $m^*\approx -n\ln k$. Inequality-based concavity of $F$ as a function of n is used to justify equal partition sizes.

What would settle it

Measure the QFI (or estimation variance) of a fixed n-sensor array as a function of partition number m, using an independently characterized entangling-gate fidelity k; Eq. (7) predicts a peak at $m^*\approx -n\ln k$. If the measured peak is elsewhere, or if the measured GHZ fidelity as a function of n is not $F(n)=k^{n-1}F$, the derived closed-form partition rules do not hold.

Watch

Extended reading notes

Core claim

The central claim is that, with a fixed total number n of sensor qubits and a state-preparation fidelity that decays as $F(n)=k^{n-1}F$, the quantum Fisher information is maximized not by entangling all n qubits but by partitioning them into $m^*\approx -n\ln k$ equal GHZ sub-ensembles. At this partition, the ratio of the optimally partitioned QFI to the monolithic QFI is approximately $-(1/e)\,k^{-n}/(n\ln k)$, an exponential advantage in n. Equivalently, the optimal sub-ensemble size $n/m^*\approx -1/\ln k$ is set only by the error rate, so additional sensors should be used to create more parallel sub-ensembles rather than larger GHZ states. With sensor loss probability p per qubit, the rule becomes $m^*\approx -n\ln(kp)$; with dephasing probability q of no phase flip, $m^*\approx -n[\ln k + 2\ln(2q-1)]$. The same partitioning viewpoint also yields closed-form QFI dynamics, peak times, optimal sensing bandwidth in the sequential scheme, and a parameter regime where noisy partitioned GHZ states outperform noiseless one-axis-twisting spin-squeezed states.

Load-bearing premise

The load-bearing premise is that preparing an n-qubit GHZ state yields fidelity $F(n)=k^{n-1}F$ with a single per-gate parameter k that also applies identically to every sub-ensemble, and that preparation errors are independent across sub-ensembles.

Editorial extensions

If this is right

  • For any sensor array with entangling-gate fidelity k, the optimal GHZ sub-ensemble contains about $-1/\ln k$ sensors; adding more sensors should create more sub-ensembles, not larger ones.
  • The QFI of an optimally partitioned sensor grows like $k^{-n}/(n\ln k)$, exponentially in the total sensor count, whereas a monolithic GHZ sensor's useful size is capped near $-2/\ln k$.
  • Particle loss and dephasing enter the same rule through effective fidelities $kp$ and $k(2q-1)^2$, so the optimal partition can be read off from independently measured error parameters.
  • In the sequential scheme, the optimal number of sub-ensembles also grows with the allowed evolution time, and partitioning raises the short-time QFI accumulation rate up to $m\approx -n\ln k$.
  • Noisy partitioned GHZ Ramsey spectroscopy can beat noiseless one-axis-twisting spin-squeezed Ramsey spectroscopy once preparation fidelity is high enough.

Reading between the lines

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

  • A broader reading is that the exponential-in-size fidelity penalty forces a crossover: Heisenberg scaling is only available up to a size set by the per-gate error, after which parallelism replaces entanglement as the optimal use of resources. This crossover could be tested by sweeping m on any platform with calibrated gate fidelity.
  • The same optimization logic should apply to other entangled probe families, such as W states or graph states, whenever their preparation fidelity decays exponentially with system size; verifying that would require a closed-form QFI analogous to Eq. (4) for each family.
  • Because the optimal sub-ensemble size is independent of n, the result suggests a calibration recipe: measure k once, set each sub-ensemble size to about $1/(1-k)$, then add as many sub-ensembles as sensors allow. This also provides a direct experimental test of Eq. (7) by measuring estimation variance versus m.
  • If preparation errors are correlated across sub-ensembles, the additivity assumption $mF(F,k,n/m)$ breaks down and the optimal partition would shift; quantifying that shift is a natural next step beyond the paper's independent-error model.
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

3 major / 4 minor

Summary. The paper studies phase estimation with GHZ states under state-preparation noise, particle loss, and dephasing. It models an n-qubit GHZ preparation as a depolarized state with fidelity F(n)=k^{n-1}F (Eq. 2) and proposes partitioning the n sensors into m independent sub-ensembles of size n/m. The main results are closed-form approximations for the optimal partition number, m* ≈ -n ln k (Eq. 7), and for the QFI ratio between optimally partitioned and monolithic strategies, F(m*)/F(monolithic) ≈ -(1/e) k^{-n}/(n ln k) (Eq. 10). The paper also derives analogous formulas with particle loss and dephasing, studies QFI dynamics, analyzes the sequential scheme, and compares noisy GHZ Ramsey spectroscopy with spin-squeezed states.

Significance. If correct, the static partitioning rule m* ≈ n(1-k) provides a simple, practical guide for resource allocation in entanglement-enhanced sensing under realistic preparation errors. The paper's analytic derivations are explicit and the approximations are checked against exact integer programming in Figs. 4, 6, and 7, which is a notable strength. The closed forms for QFI with loss and dephasing extend the authors' earlier formula. However, the headline results are conditional on a specific noise model and on unquantified asymptotic approximations, and the sequential-scheme optimization in Sec. VI contains an internal inconsistency; these issues require revision before the claims can be considered fully established.

major comments (3)
  1. [Sec. II.A, Eqs. (1)-(2); Sec. III.C, Eq. (7)] The central optimal-partition rule m*≈-n ln k and the exponential ratio Eq. (10) rest on the assumption, stated after Eq. (5), that sub-ensembles can be independently prepared with uniform errors and the same per-gate fidelity k for all m. The text acknowledges in Sec. II.A that the depolarized/exponential-fidelity model is 'not unique,' but it does not quantify the error incurred when the assumption fails. If parallel GHZ preparation shares control hardware, crosstalk or global noise can make the effective k depend on m or correlate errors across sub-ensembles; then the derivative condition dF/dm≈-F n^2 k^{n/m-1}(m+n ln k)/m^3 in Appendix A.2 is not the correct optimality condition and m* can differ from Eq. (7). The paper should either justify the independence microscopically for a concrete platform or provide a robustness bound, for example through a numerical study with correlated noise or an m-dependent k.
  2. [Appendix A; Eqs. (6)-(7), (19)-(21), (37)-(39)] The closed-form expressions for n* and m* are obtained by dropping 'exponentially suppressed terms' (e.g., Eqs. (A1)-(A3) and (A5)-(A8)) without a bound on the remainder. The stated validity conditions, 'k, F close to 1' and 'n, n/m not small,' are not made quantitative, and the figures validate only isolated parameter points (k=0.99 in Fig. 4; k=0.995 in Fig. 7). Since the optimal partition numbers are the main quantitative output, a parameter-region validity statement or a numerical sweep over (k, p, q, n) is needed to establish the range in which Eqs. (7), (21), and (39) can be used.
  3. [Sec. VI.C, Eqs. (54)-(58)] The sequential-scheme optimization is not consistent when t can be chosen freely. For fixed m, the best evolution time is t̃*=m/[n(2γ+η)] from Eq. (54); substituting this into Eq. (57) gives m̃*(t̃*)=m-n ln k > m, so the stationary-point equations for t and m cannot be satisfied simultaneously. In fact, the maximal QFI per unit time at the t-optimal point, Eq. (55), is monotone increasing in m, so the unconstrained joint optimum is m=n (separable sensors), not m≈-n ln k. The recommendation that m and n should be matched by m≈-n ln k is therefore valid only when the evolution time per cycle is bounded by tth; the text should state this restriction explicitly and the claimed advantage of partitioning in the sequential scheme should be framed accordingly.
minor comments (4)
  1. [Throughout] The symbol F is used both for the initialization fidelity (Eq. (2)) and for the quantum Fisher information (e.g., Eq. (4)); this is confusing, especially in Eqs. (8)-(11) and (24)-(26). Please use a different symbol, such as a script F, for one of the two quantities.
  2. [Secs. III.E, IV.A, IV.C, V.A, VI.C] There are several typographical errors that should be corrected: 'beacuse' (end of Sec. III.E), 'wiothout' and 'yileds' (Sec. IV.C), 'ocurred' (Sec. IV.A), 'costist' (Sec. V.A), and 'highest highest maximum' (Sec. VI.C). Also, in Sec. VI.C, 'm < k−1' should be 'm < l−1'.
  3. [Sec. VI.C, Eq. (58)] The asymptotic performance in Eq. (58), I → F nT/[e(2γ+η)], appears to omit a factor 1/k relative to Eq. (55), which gives F̃_peak ≈ F n k^{n/m}/[e k (2γ+η)]; in the limit k^{n/m}→1 the result should be F nT/[e k (2γ+η)] unless an additional approximation is intended. Please check and clarify.
  4. [Sec. II.D, Eq. (3)] The QFI formula in Eq. (3) and its evaluation in Eq. (4) are taken from the authors' prior work [59]. A brief self-contained derivation in an appendix would make the paper more readable and would reduce the dependence on an external reference for a central quantity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the closed-form partition rule is derived from the explicitly stated exponential-fidelity noise model, and the model is presented as an assumption, not as a conclusion of the paper.

full rationale

The derivation chain is self-contained conditional on the noise model stated in Sec. II A, Eqs. (1) and (2): a depolarized GHZ state with fidelity F(n) = k^(n-1)F. The central QFI expression Eq. (4) is imported from the authors' prior work [59], but it is a formula for the QFI of a depolarized GHZ state under that same stated model; it is not equivalent to the partitioning result and is re-derivable from Eqs. (1)-(3). The optimal partition number Eq. (7), m* ≈ -n ln k, is obtained by differentiating the QFI with respect to m in Appendix A2 and setting the derivative to zero, not by assuming the answer. The exponential advantage ratio Eq. (10) is then algebra from substituting Eq. (7) into Eq. (4); its exponential form tracks the assumed exponential fidelity decay, which the paper explicitly labels as an assumption. The loss and dephasing results are derived from the same model with no fitted parameters and with no target result used as an input. The self-citations to [59] and [87] are present, and [59] supplies the QFI formula used, but the cited formula is an auxiliary mathematical result rather than the paper's conclusion, and no uniqueness theorem or ansatz is smuggled in via citation. The known scaling of optimal sub-ensemble size with inverse error rate is acknowledged as corroboration from [56-58], not presented as a derivation from them. Concerns about correlated preparation errors or crosstalk invalidating the model are model-risk issues, not circularity, because the model is transparently an assumption. No step in the derivation reduces by construction to its own inputs, so the circularity score is 0.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

No data fitting occurs; k, F, p, q are model parameters the authors treat as known from characterization. The paper's claims depend on these values through the domain assumptions above, so the contribution is conditional on the noise model being accurate.

assumptions (7)
  • domain assumption Depolarized GHZ preparation with fidelity F(n) = k^{n-1} F
    Sec. II A, Eqs. (1)-(2). The authors state this model is not unique but operationally meaningful; the exponential fidelity decay is central to all optimal partition formulas.
  • domain assumption Independent particle loss as a binary process with survival probability p; lost qubits replaced by maximally mixed state
    Sec. II B and Appendix B. This allows loss to be folded into state preparation errors, a key step for Eqs. 13 and 18.
  • domain assumption Independent dephasing described by the single-qubit phase-flip channel with no-flip probability q in (1/2, 1]
    Sec. II C and Sec. V A, Eqs. (29)-(30). The commutativity of encoding and dephasing is used to derive Eq. 35.
  • domain assumption Uniform, independent sub-ensembles, so QFI is additive and each sub-ensemble has the same k
    Sec. III A, Eq. (5) and Sec. III E. Additivity under tensor product is standard, but independence and uniform k across sub-ensembles is an assumption about experimental control.
  • ad hoc to paper Asymptotic approximations valid for k, F, p, q close to 1 and n, n/m large
    Derivations in Appendices A-C repeatedly drop exponentially suppressed terms and take coth x approximately 1; these approximations are not rigorously bounded.
  • domain assumption The dynamics are parameterized by p = e^{-eta t} and q = (1 + e^{-gamma t})/2, with state preparation error k independent of t
    Sec. VI A-B, used to turn the static QFI formulas into time-dependent ones.
  • domain assumption For the OAT SSS comparison, noiseless preparation of the spin-squeezed state and use of the Kitagawa-Ueda lower bound xi_S^2 as a proxy for the Ramsey variance
    Sec. VI D. The authors correctly note this is a lower bound, so their comparison is conservative toward spin squeezing.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhancing Noisy Quantum Sensing by GHZ State Partitioning." pith.science (2026). https://pith.science/paper/2KQYNYJD

@misc{pith2026250702829,
  author       = {Pith},
  title        = {Pith review of: Enhancing Noisy Quantum Sensing by GHZ State Partitioning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2KQYNYJD}},
  note         = {Machine review of arXiv:2507.02829}
}
read the original abstract

Presence of harmful noise is inevitable in entanglement-enhanced sensing systems, requiring careful allocation of resources to optimize sensing performance in practical scenarios. We advocate a simple but effective strategy to improve sensing performance in the presence of noise. Given a fixed number of quantum sensors, we partition the preparation of GHZ states by preparing smaller, independent sub-ensembles of GHZ states instead of a GHZ state across all sensors. We perform extensive analytical studies of the phase estimation performance when using partitioned GHZ states under realistic noise -- including state preparation error, particle loss during parameter encoding, and sensor dephasing during parameter encoding. We derive simple, closed-form expressions that quantify the optimal number of sub-ensembles for partitioned GHZ states. We also examine the explicit noisy quantum sensing dynamics under dephasing and loss, where we demonstrate the advantage from partitioning for maximal QFI, short-time QFI increase, and the sensing performance in the sequential scheme. The results offer quantitative insights into the sensing performance impact of different noise sources and reinforce the importance of resource allocation optimization in realistic quantum applications.

Figures

Figures reproduced from arXiv: 2507.02829 by the authors.

Figure 1
Figure 1. FIG. 1. Schematic of using the same number of quantum sen [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Visualization of the parameter estimation variance [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Visualization of the highest achievable QFI under op [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Visualization of the optimal partitioning number [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Visualization of [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Visualization of [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 8
Figure 8. Figure 8: FIG. 8. Ratio between the QFI for the case with loss de [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: FIG. 9. Time dynamics of the (rescaled) QFI, i.e. QFI mul [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 10
Figure 10. Figure 10: FIG. 10. Time dynamics of the (rescaled) QFI (QFI multi [PITH_FULL_IMAGE:figures/full_fig_p013_10.png]
Figure 11
Figure 11. Figure 11: FIG. 11. Time dynamics of the (rescaled) QFI (QFI multi [PITH_FULL_IMAGE:figures/full_fig_p014_11.png]
Figure 12
Figure 12. Figure 12: FIG. 12. (Rescaled) QFI per unit evolution time as a function [PITH_FULL_IMAGE:figures/full_fig_p015_12.png]
Figure 13
Figure 13. Figure 13: FIG. 13. Comparison between scaled cross time [PITH_FULL_IMAGE:figures/full_fig_p017_13.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

117 extracted references · 61 canonical work pages

  1. [59]

    A. Cao, W. J. Eckner, T. L. Yelin, A. W. Young, S. Jan- dura, L. Yan, K. Kim, G. Pupillo, J. Ye, N. D. Oppong, et al.,Multi-qubitgatesand’Schrödingercat’statesinan optical clock, arXiv preprint arXiv:2402.16289 (2024)

    work page Pith review arXiv 2024

  2. [1]

    To make the time dependence explicit and intuitive, we visualize the QFI dynamics in Fig

    Without loss detection When we do not have the capability to detect losses that happen during the parameter encoding evolution, the QFI as a function of the evolution timet is Floss,1(t) =Floss,1(F, k, e−ηt, n, m)t2 = n2 k − F (2k) n m 2 e− 2n m ηtt2/ mk2 2 n m − 1 1+e−ηt 2 n m + e− n m ηt h k+F k n m 2 n m −2 i k 2 n m −1 − e− n m ηt ≈ n2F 2 (ke−ηt) 2n m...

  3. [2]

    With loss detection In the case where we are able to detect losses, the QFI dynamics becomes Floss,2(t) =Floss,2(F, k, e−ηt, n, m)t2 = n2 F (2k) n m − k 2 e− n m ηtt2 mk 2 n m − 1 F 2 n m − 2 k n m + k . (44) Qualitatively, the behavior of the QFI dynamics is the same as the first case in that the QFI will first increase and then decrease, since the compe...

  4. [3]

    Here we compare the two cases considered above

    Advantage of loss detection When there are particle losses during the probing evo- lution, whether having the capability of detecting the losses will make a difference in the performance. Here we compare the two cases considered above. A basic exam- ple is illustrated in Fig. 10, where we visualizeFloss,1(t) and Floss,2(t) for n = 200, 300, 400, while fix...

  5. [4]

    The evolution time t is restricted by t ∈ [0, tth] with tth < ˜t∗

  6. [5]

    Therefore, we only need to perform optimization arg maxm≥1,m∈Z[Floss,dp(t)/t] to find the optimized m

    We have thatFloss,dp(t)/t monoton- ically increases throughout this interval for allm. Therefore, we only need to perform optimization arg maxm≥1,m∈Z[Floss,dp(t)/t] to find the optimized m

  7. [6]

    Then we know thatFloss,dp(t)/t can achieve its maximum by choosing an evolution time within the interval for m = 1, 2,

    The evolution time t is restricted by t ∈ [0, tth] with tth ∈ [˜t∗ l−1, ˜t∗ l ), for l > 1, l∈ Z. Then we know thatFloss,dp(t)/t can achieve its maximum by choosing an evolution time within the interval for m = 1, 2, . . . , l− 1, while takingm = l − 1 achieves the highest highest maximum in comparison with other m < k− 1. Then we should perform the optim...

  8. [7]

    We first consider the scenario with state prepara- tion errors only as the minimal example. We derive a simple, closed-form approximation for the opti- mal total number of sensorn∗ under a fixed parti- tion number m and the optimal partition number m∗ for a fixed total number of sensorsn. We also demonstrate the impact of entangling gate fideli- ties on t...

Show all 117 references
  1. [8]

    Therein we consider two cases, one with the ability to detect losses in a nondemolition manner, and the other without

    We combine state preparation errors and qubit losses during the encoding process. Therein we consider two cases, one with the ability to detect losses in a nondemolition manner, and the other without. We derive approximations for the opti- mal total number of sensors n∗ and th...

  2. [9]

    Again, we derive approximations for the optimal total number of sensorsn∗ and the optimal partition number m∗

    We then study a scenario that includes both state preparation errors and qubit dephasing during the encoding process. Again, we derive approximations for the optimal total number of sensorsn∗ and the optimal partition number m∗. We derive that n∗ and m∗ are 4 times more sensit...

  3. [10]

    Last but not least, we perform comprehensive in- vestigation into the explicit quantum sensing dy- namics under losses and dephasing. We derive closed-form expressions for QFI dynamics, and demonstrate that partitioning continues to offer benefits to the maximal achievable QFI...

  4. [11]

    C. L. Degen, F. Reinhard, and P. Cappellaro, Quantum sensing, Reviews of Modern Physics89, 035002 (2017)

    2017

  5. [12]

    Safronova, D

    M. Safronova, D. Budker, D. DeMille, D. F. J. Kimball, A. Derevianko, and C. W. Clark, Search for new physics with atoms and molecules, Reviews of Modern Physics 90, 025008 (2018)

    2018

  6. [13]

    J.YeandP.Zoller,Essay: Quantumsensingwithatomic, molecular, andopticalplatformsforfundamentalphysics, Physical Review Letters132, 190001 (2024)

    2024

  7. [14]

    A. D. Ludlow, M. M. Boyd, J. Ye, E. Peik, and P. O. Schmidt, Optical atomic clocks, Reviews of Modern Physics 87, 637 (2015)

    2015

  8. [15]

    Müller, A

    H. Müller, A. Peters, and S. Chu, A precision measure- ment of the gravitational redshift by the interference of matter waves, Nature463, 926 (2010)

    2010

  9. [16]

    B.P.Abbott, R.Abbott, T.D.Abbott, M.R.Abernathy, F.Acernese, K.Ackley, C.Adams, T.Adams, P.Addesso, R. X. Adhikari,et al., Observation of gravitational waves from a binary black hole merger, Physical Review Letters 116, 061102 (2016)

    2016

  10. [17]

    S. D. Bass and M. Doser, Quantum sensing for particle physics, Nature Reviews Physics6, 329 (2024)

    2024

  11. [18]

    J. M. Taylor, P. Cappellaro, L. Childress, L. Jiang, D. Budker, P. R. Hemmer, A. Yacoby, R. Walsworth, and M. D. Lukin, High-sensitivity diamond magnetometer with nanoscale resolution, Nature Physics4, 810 (2008)

    2008

  12. [19]

    J. R. Maze, P. L. Stanwix, J. S. Hodges, S. Hong, J. M. Taylor, P. Cappellaro, L. Jiang, M. G. Dutt, E. Togan, A. Zibrov, et al., Nanoscale magnetic sensing with an individual electronic spin in diamond, Nature455, 644 (2008)

    2008

  13. [20]

    J. F. Barry, J. M. Schloss, E. Bauch, M. J. Turner, C. A. Hart, L. M. Pham, and R. L. Walsworth, Sensitivity optimization for nv-diamond magnetometry, Reviews of Modern Physics92, 015004 (2020)

    2020

  14. [21]

    Dolde, H

    F. Dolde, H. Fedder, M. W. Doherty, T. Nöbauer, F. Rempp, G. Balasubramanian, T. Wolf, F. Reinhard, L. C. Hollenberg, F. Jelezko, and J. Wrachtrup, Electric- field sensing using single diamond spins, Nature Physics 7, 459 (2011)

    2011

  15. [22]

    K. Bian, W. Zheng, X. Zeng, X. Chen, R. Stöhr, A. Denisenko, S. Yang, J. Wrachtrup, and Y. Jiang, Nanoscale electric-field imaging based on a quantum sen- sor and its charge-state control under ambient condition, Nature Communications12, 2457 (2021)

    2021

  16. [23]

    Casola, T

    F. Casola, T. Van Der Sar, and A. Yacoby, Probing condensed matter physics with magnetometry based on nitrogen-vacancy centres in diamond, Nature Reviews Materials 3, 1 (2018)

    2018

  17. [24]

    Rovny, S

    J. Rovny, S. Gopalakrishnan, A. C. B. Jayich, P. Maletinsky, E. Demler, and N. P. de Leon, Nanoscale diamond quantum sensors for many-body physics, Na- ture Reviews Physics , 1 (2024)

    2024

  18. [25]

    Schirhagl, K

    R. Schirhagl, K. Chang, M. Loretz, and C. L. Degen, Nitrogen-vacancy centers in diamond: nanoscale sen- sors for physics and biology, Annual Review of Physical Chemistry 65, 83 (2014)

    2014

  19. [26]

    Aslam, H

    N. Aslam, H. Zhou, E. K. Urbach, M. J. Turner, R. L. Walsworth, M. D. Lukin, and H. Park, Quantum sensors for biomedical applications, Nature Reviews Physics5, 157 (2023)

    2023

  20. [27]

    L. M. Pham, D. Le Sage, P. L. Stanwix, T. K. Yeung, D. Glenn, A. Trifonov, P. Cappellaro, P. R. Hemmer, M. D. Lukin, H. Park,et al., Magnetic field imaging with nitrogen-vacancy ensembles, New Journal of Physics13, 045021 (2011)

    2011

  21. [28]

    D. R. Glenn, D. B. Bucher, J. Lee, M. D. Lukin, H. Park, and R. L. Walsworth, High-resolution magnetic reso- nance spectroscopy using a solid-state spin sensor, Na- ture 555, 351 (2018)

    2018

  22. [29]

    Aeppli, K

    A. Aeppli, K. Kim, W. Warfield, M. S. Safronova, and J. Ye, Clock with 8× 10-19 systematic uncertainty, Phys- ical Review Letters133, 023401 (2024)

    2024

  23. [30]

    Giovannetti, S

    V. Giovannetti, S. Lloyd, and L. Maccone, Quantum- enhanced measurements: beating the standard quantum limit, Science306, 1330 (2004)

    2004

  24. [31]

    Pezzé and A

    L. Pezzé and A. Smerzi, Entanglement, nonlinear dynam- ics, and the heisenberg limit, Physical Review Letters 102, 100401 (2009)

    2009

  25. [32]

    D. M. Greenberger, M. A. Horne, and A. Zeilinger, Going beyond bell’s theorem, inBell’s theorem, quantum theory and conceptions of the universe(Springer, 1989) pp. 69– 72

    1989

  26. [33]

    D. J. Wineland, J. J. Bollinger, W. M. Itano, F. Moore, and D. J. Heinzen, Spin squeezing and reduced quan- tum noise in spectroscopy, Physical Review A46, R6797 (1992)

    1992

  27. [34]

    M.KitagawaandM.Ueda,Squeezedspinstates,Physical Review A47, 5138 (1993)

    1993

  28. [35]

    Cappellaro, J

    P. Cappellaro, J. Emerson, N. Boulant, C. Ramanathan, S. Lloyd, and D. G. Cory, Entanglement assisted metrol- ogy, Physical Review Letters94, 020502 (2005)

    2005

  29. [36]

    Giovannetti, S

    V. Giovannetti, S. Lloyd, and L. Maccone, Quantum metrology, Physical Review Letters96, 010401 (2006)

    2006

  30. [37]

    Cappellaro and M

    P. Cappellaro and M. D. Lukin, Quantum correlation in disordered spin systems: Applications to magnetic sens- ing, Physical Review A—Atomic, Molecular, and Optical Physics 80, 032311 (2009)

    2009

  31. [38]

    Goldstein, P

    G. Goldstein, P. Cappellaro, J. R. Maze, J. Hodges, L. Jiang, A. S. Sørensen, and M. Lukin, Environment- assisted precision measurement, Physical Review Letters 106, 140502 (2011)

    2011

  32. [39]

    Pezze, A

    L. Pezze, A. Smerzi, M. K. Oberthaler, R. Schmied, and P. Treutlein, Quantum metrology with nonclassical 19 states of atomic ensembles, Reviews of Modern Physics 90, 035005 (2018)

    2018

  33. [40]

    Pedrozo-Peñafiel, S

    E. Pedrozo-Peñafiel, S. Colombo, C. Shu, A. F. Adiy- atullin, Z. Li, E. Mendez, B. Braverman, A. Kawasaki, D. Akamatsu, Y. Xiao,et al., Entanglement on an optical atomic-clock transition, Nature588, 414 (2020)

    2020

  34. [41]

    H. Yu, L. McCuller, M. Tse, N. Kijbunchoo, L. Barsotti, and N. Mavalvala, Quantum correlations between light and the kilogram-mass mirrors of ligo, Nature583, 43 (2020)

    2020

  35. [42]

    T.Xie, Z.Zhao, X.Kong, W.Ma, M.Wang, X.Ye, P.Yu, Z. Yang, S. Xu, P. Wang,et al., Beating the standard quantum limit under ambient conditions with solid-state spins, Science advances7, eabg9204 (2021)

    2021

  36. [43]

    W. Wu, E. J. Davis, L. B. Hughes, B. Ye, Z. Wang, D. Kufel, T. Ono, S. A. Meynell, M. Block, C. Liu,et al., Spin squeezing in an ensemble of nitrogen-vacancy cen- tersindiamond,arXivpreprintarXiv:2503.14585 (2025)

    2025

  37. [44]

    Cooper, W

    A. Cooper, W. K. C. Sun, J.-C. Jaskula, and P. Cap- pellaro, Environment-assisted quantum-enhanced sens- ing with electronic spins in diamond, Physical Review Applied 12, 044047 (2019)

    2019

  38. [45]

    S. F. Huelga, C. Macchiavello, T. Pellizzari, A. K. Ekert, M. B. Plenio, and J. I. Cirac, Improvement of frequency standards with quantum entanglement, Physical Review Letters 79, 3865 (1997)

    1997

  39. [46]

    Escher, R

    B. Escher, R. L. de Matos Filho, and L. Davidovich, Gen- eral framework for estimating the ultimate precision limit in noisy quantum-enhanced metrology, Nature Physics7, 406 (2011)

    2011

  40. [47]

    Demkowicz-Dobrzański, J

    R. Demkowicz-Dobrzański, J. Kołodyński, and M. Guţă, The elusive heisenberg limit in quantum-enhanced metrology, Nature communications3, 1063 (2012)

    2012

  41. [48]

    Borregaard and A

    J. Borregaard and A. S. Sørensen, Efficient atomic clocks operated with several atomic ensembles, Physical review letters 111, 090802 (2013)

    2013

  42. [49]

    Rosenband and D

    T. Rosenband and D. R. Leibrandt, Exponential scal- ing of clock stability with atom number, arXiv preprint arXiv:1303.6357 (2013)

    2013 arXiv

  43. [50]

    E. M. Kessler, P. Komar, M. Bishof, L. Jiang, A. S. Sørensen, J. Ye, and M. D. Lukin, Heisenberg-limited atom clocks based on entangled qubits, Physical review letters 112, 190403 (2014)

    2014

  44. [51]

    Schulte, C

    M. Schulte, C. Lisdat, P. O. Schmidt, U. Sterr, and K. Hammerer, Prospects and challenges for squeezing- enhanced optical atomic clocks, Nature Communications 11, 5955 (2020)

    2020

  45. [52]

    W.Li, S.Wu, A.Smerzi,andL.Pezzè,Improvedabsolute clock stability by the joint interrogation of two atomic ensembles, Physical Review A105, 053116 (2022)

    2022

  46. [53]

    J. M. Martinis, Qubit metrology for building a fault- tolerant quantum computer, npj Quantum Information 1, 1 (2015)

    2015

  47. [54]

    Harper, S

    R. Harper, S. T. Flammia, and J. J. Wallman, Effi- cient learning of quantum noise, Nature Physics16, 1184 (2020)

    2020

  48. [55]

    Eisert, D

    J. Eisert, D. Hangleiter, N. Walk, I. Roth, D. Markham, R. Parekh, U. Chabaud, and E. Kashefi, Quantum cer- tification and benchmarking, Nature Reviews Physics2, 382 (2020)

    2020

  49. [56]

    Kliesch and I

    M. Kliesch and I. Roth, Theory of quantum system cer- tification, PRX Quantum2, 010201 (2021)

    2021

  50. [57]

    Schioppo, R

    M. Schioppo, R. C. Brown, W. F. McGrew, N. Hinkley, R. J. Fasano, K. Beloy, T. Yoon, G. Milani, D. Nicolodi, J. Sherman, et al., Ultrastable optical clock with two cold-atom ensembles, Nature Photonics11, 48 (2017)

    2017

  51. [58]

    J. A. Hines, S. V. Rajagopal, G. L. Moreau, M. D. Wahrman, N. A. Lewis, O. Marković, and M. Schleier- Smith, Spin squeezing by rydberg dressing in an array of atomic ensembles, Physical Review Letters131, 063401 (2023)

    2023

  52. [60]

    Gefen, J

    R.Finkelstein, R.B.-S.Tsai, X.Sun, P.Scholl, S.Direkci, T. Gefen, J. Choi, A. L. Shaw, and M. Endres, Univer- sal quantum operations and ancilla-based read-out for tweezer clocks, Nature634, 321 (2024)

    2024

  53. [61]

    A. L. Shaw, R. Finkelstein, R. B.-S. Tsai, P. Scholl, T. H. Yoon, J. Choi, and M. Endres, Multi-ensemble metrology by programming local rotations with atom movements, Nature Physics20, 195 (2024)

    2024

  54. [62]

    J. M. Robinson, M. Miklos, Y. M. Tso, C. J. Kennedy, T.Bothwell, D.Kedar, J.K.Thompson,andJ.Ye,Direct comparison of two spin-squeezed optical clock ensembles at the 10- 17 level, Nature Physics20, 208 (2024)

    2024

  55. [63]

    E. S. Cooper, P. Kunkel, A. Periwal, and M. Schleier- Smith, Graph states of atomic ensembles engineered by photon-mediated entanglement, Nature Physics20, 770 (2024)

    2024

  56. [64]

    Zheng, J

    X. Zheng, J. Dolde, and S. Kolkowitz, Reducing the in- stability of an optical lattice clock using multiple atomic ensembles, Physical Review X14, 011006 (2024)

    2024

  57. [65]

    Y. S. Hassan, T. Kobayashi, T. Bothwell, J. L. Seigel, B. D. Hunt, K. Beloy, K. Gibble, T. Grogan, and A. D. Ludlow,Ratchetloadingandmulti-ensembleoperationin anopticallatticeclock,QuantumScienceandTechnology 9, 045023 (2024)

    2024

  58. [66]

    C. Yin, V. V. Albert, and S. Zhou, Small correlation is sufficient for optimal noisy quantum metrology, arXiv preprint arXiv:2408.00079 (2024)

    2024

  59. [67]

    Jarzyna and R

    M. Jarzyna and R. Demkowicz-Dobrzański, Matrix prod- uct states for quantum metrology, Physical Review Let- ters 110, 240405 (2013)

    2013

  60. [68]

    Chabuda, J

    K. Chabuda, J. Dziarmaga, T. J. Osborne, and R. Demkowicz-Dobrzański, Tensor-network approach for quantum metrology in many-body quantum systems, Na- ture communications11, 250 (2020)

    2020

  61. [69]

    A. Zang, A. Kolar, A. Gonzales, J. Chung, S. K. Gray, R. Kettimuthu, T. Zhong, and Z. H. Saleem, Quantum advantage in distributed sensing with noisy quantum net- works, arXiv preprint arXiv:2409.17089 (2024)

    2024 arXiv

  62. [70]

    T. Monz, P. Schindler, J. T. Barreiro, M. Chwalla, D. Nigg, W. A. Coish, M. Harlander, W. Hänsel, M. Hen- nrich, and R. Blatt, 14-qubit entanglement: Creation and coherence, Physical Review Letters106, 130506 (2011)

    2011

  63. [71]

    C. Song, K. Xu, H. Li, Y.-R. Zhang, X. Zhang, W. Liu, Q. Guo, Z. Wang, W. Ren, J. Hao,et al., Generation of multicomponent atomic Schrödinger cat states of up to 20 qubits, Science365, 574 (2019)

    2019

  64. [72]

    Omran, H

    A. Omran, H. Levine, A. Keesling, G. Semeghini, T. T. Wang, S. Ebadi, H. Bernien, A. S. Zibrov, H. Pich- ler, S. Choi, et al., Generation and manipulation of Schrödinger cat states in rydberg atom arrays, Science 365, 570 (2019)

    2019

  65. [73]

    W. W. Ho, C. Jonay, and T. H. Hsieh, Ultrafast varia- tional simulation of nontrivial quantum states with long- 20 range interactions, Physical Review A99, 052332 (2019)

    2019

  66. [74]

    Pogorelov, T

    I. Pogorelov, T. Feldker, C. D. Marciniak, L. Postler, G. Jacob, O. Krieglsteiner, V. Podlesnic, M. Meth, V. Negnevitsky, M. Stadler, et al., Compact ion-trap quantum computing demonstrator, PRX Quantum 2, 020343 (2021)

    2021

  67. [75]

    G. J. Mooney, G. A. White, C. D. Hill, and L. C. Hollenberg, Generation and verification of 27-qubit Greenberger-Horne-Zeilinger states in a superconducting quantum computer, Journal of Physics Communications 5, 095004 (2021)

    2021

  68. [76]

    Y. Zhao, R. Zhang, W. Chen, X.-B. Wang, and J. Hu, Creation of Greenberger-Horne-Zeilinger states with thousands of atoms by entanglement amplification, npj Quantum Information7, 24 (2021)

    2021

  69. [77]

    Comparin, F

    T. Comparin, F. Mezzacapo, and T. Roscilde, Multi- partite entangled states in dipolar quantum simulators, Physical Review Letters129, 150503 (2022)

    2022

  70. [78]

    Zhang, Z

    X. Zhang, Z. Hu, and Y.-C. Liu, Fast generation of GHZ- like states using collective-spin XYZ model, Physical Re- view Letters132, 113402 (2024)

    2024

  71. [79]

    Yin, Fast and accurate GHZ encoding using all-to-all interactions, arXiv preprint arXiv:2406.10336 (2024)

    C. Yin, Fast and accurate GHZ encoding using all-to-all interactions, arXiv preprint arXiv:2406.10336 (2024)

    2024 arXiv

  72. [80]

    Sekatski, M

    P. Sekatski, M. Skotiniotis, J. Kołodyński, and W. Dür, Quantum metrology with full and fast quantum control, Quantum 1, 27 (2017)

    2017

  73. [81]

    Demkowicz-Dobrzański, J

    R. Demkowicz-Dobrzański, J. Czajkowski, and P. Sekatski, Adaptive quantum metrology under general markovian noise, Physical Review X7, 041009 (2017)

    2017

  74. [82]

    S. Zhou, M. Zhang, J. Preskill, and L. Jiang, Achiev- ing the heisenberg limit in quantum metrology using quantum error correction, Nature communications9, 78 (2018)

    2018

  75. [83]

    C. W. Helstrom, Quantum detection and estimation the- ory, Journal of Statistical Physics1, 231 (1969)

    1969

  76. [84]

    M. G. Paris, Quantum estimation for quantum technol- ogy, International Journal of Quantum Information 7, 125 (2009)

    2009

  77. [85]

    Petz and C

    D. Petz and C. Ghinea, Introduction to quantum Fisher information, inQuantum Probability and Related Topics (World Scientific, 2011) pp. 261–281

    2011

  78. [86]

    Tóth and I

    G. Tóth and I. Apellaniz, Quantum metrology from a quantum information science perspective, Journal of Physics A: Mathematical and Theoretical 47, 424006 (2014)

    2014

  79. [87]

    J. Liu, H. Yuan, X.-M. Lu, and X. Wang, Quantum fisher information matrix and multiparameter estima- tion,JournalofPhysicsA:MathematicalandTheoretical 53, 023001 (2020)

    2020

  80. [88]

    These two differ by at most one, so they will not make a big difference in practical regimes with largen

    The number of qubits must be an integer, thus we actu- ally refer to⌊n/m⌋ or ⌈n/m⌉. These two differ by at most one, so they will not make a big difference in practical regimes with largen

  81. [89]

    J. J. Meyer, Fisher information in noisy intermediate- scale quantum applications, Quantum5, 539 (2021)

    2021

  82. [90]

    Y. Wang, S. Crain, C. Fang, B. Zhang, S. Huang, Q. Liang, P. H. Leung, K. R. Brown, and J. Kim, High- fidelity two-qubit gates using a microelectromechanical- system-based beam steering system for individual qubit addressing, Physical Review Letters125, 150505 (2020)

    2020

  83. [91]

    Google Quantum AI, Suppressing quantum errors by scaling a surface code logical qubit, Nature 614, 676 (2023)

    2023

  84. [92]

    Bluvstein, S

    D. Bluvstein, S. J. Evered, A. A. Geim, S. H. Li, H. Zhou, T. Manovitz, S. Ebadi, M. Cain, M. Kalinowski, D. Hangleiter,et al., Logical quantum processor based on reconfigurable atom arrays, Nature626, 58 (2024)

    2024

  85. [93]

    Bartling, J

    H. Bartling, J. Yun, K. Schymik, M. van Riggelen, L. Enthoven, H. van Ommen, M. Babaie, F. Sebastiano, M. Markham, D. Twitchen,et al., Universal high-fidelity quantumgatesforspin-qubitsindiamond,arXivpreprint arXiv:2403.10633 (2024)

    2024 arXiv

  86. [94]

    Z. H. Saleem, A. Shaji, and S. K. Gray, Optimal time for sensing in open quantum systems, Physical Review A 108, 022413 (2023)

    2023

  87. [95]

    J. Ma, X. Wang, C.-P. Sun, and F. Nori, Quantum spin squeezing, Physics Reports509, 89 (2011)

    2011

  88. [96]

    G. P. Egorychev,Integral representation and the compu- tation of combinatorial sums, Vol. 59 (American Mathe- matical Soc., 1984)

    1984

  89. [97]

    A. Zang, X. Chen, E. Chitambar, M. Suchara, and T. Zhong, No-go theorems for universal entanglement pu- rification, Physical Review Letters134, 190803 (2025). Appendix A: Derivations for the scenario with only state preparation errors

    2025

  90. [98]

    Note that k and F should both be close to unity in practically meaningful regime, and the optimal number of sensors is generally large

    Optimal number of sensors To determine the optimal number of sensors, we would like to evaluate the partial derivative of the QFI with respect to n. Note that k and F should both be close to unity in practically meaningful regime, and the optimal number of sensors is generally...

  91. [99]

    Optimal number of sub-ensembles Now we take the partial derivative ofF (F, k, n, m) with respect tom and perform analytical approximations based on similar assumptions to the above. 2 n m −1 2 k k+F 2 n m −2 k n m 2 m3 k − F (2k) n m n2 ∂ ∂m F (F, k, n, m) = F k1+ n m 2 n m (2...

  92. [100]

    Advantage of partitioning We present the explicit expression of the ratio between optimally partitioned QFI and the monolithic QFI, using the closed-form approximation ofm∗. F (F, k, n, m∗) F (F, k, n) ≈ (2n −1) F −2 1 ln k ek 2 [F kn(2n −2) +k] e 2 1 ln k − 1 [ k − F (2k)n]2 ...

  93. [101]

    Equal partition To justify the equal partition from concavity ofF (F, k, n), we evaluate its second-order partial derivative with respect to n. After a tedious derivation, one is able to show − k(1 − 2n)3[(2n − 2)F kn + k]3 ∂2 ∂n2 F (F, k, n) = (2n + 1)(2nknF − k)2 ln2 2 + 2(2...

  94. [102]

    Effect of particle losses We first demonstrate that the result of any particle loss pattern is a GHZ-diagonal state withλa = λb, ∀(a, b) ∈ S. Consider a GHZ basis state (|i⟩ + |i⟩)/ √ 2, where |i⟩ is an arbitrary computational basis state, and |i⟩ is the 23 computational basis...

  95. [103]

    One can see that the resulting state is indeed a GHZ-diagonal state which satisfiesλa = λb, ∀(a, b) ∈ S

    Suppose the first qubit is lost, then the resulting state is 1√ 2 (|000⟩ + |111⟩) 1st qubit lost − − − − − − − − →1 2 (|0⟩⟨0| + |1⟩⟨1|) ⊗ 1 2 (|00⟩⟨00| + |11⟩⟨11|) = 1 4 (|000⟩⟨000| + |100⟩⟨100| + |011⟩⟨011| + |111⟩⟨111|) = 1 4 1 2 (|000⟩ + |111⟩)(⟨000| + ⟨111|) +1 2 (|000⟩ − ...

  96. [104]

    We first examine the second case which assumes loss detection capability

    Optimal number of sensors We take the partial derivative with respect ton of the QFI. We first examine the second case which assumes loss detection capability. km2 2 n m −1 2 k+F 2 n m −2 k n m n F (2k) n m −k p n m ∂ ∂n Floss,2(F, k, p, n, m) = k2 2 n m n ln 2 p − 2m ((((((+n...

  97. [105]

    Optimal number of sub-ensembles We first examine the second case which assumes loss detection capability. km3 2 n m −1 2 k + F k n m 2 n m −2 2 kn2p n m 1 − (2k) n m ∂ ∂m Floss,2(F, k, p, n, m) = k2 m + n ln p + 2 n m n ln 2 p − m + f k m+n m 4 n m n ln(8k3) + 2 n m (2m + 2n l...

  98. [106]

    2 n m − 4p 1 +p n m −1 # ≈1 + k F 1 +p 4kp n m

    Advantage of loss detection In this section we compare the sensing performance in the two considered cases, that is,Feff,1(F, k, p, n/m) and pn/mF (F, k, n/m). According to previous derivation details, we have Feff,1(F, k, p, n) =n2 p2n h F (n) − 1−F (n) 2n−1 i2 pn h F (n) +1−...

  99. [107]

    We have used the fact that in practicek, p≲ 1 so that ln k, ln p ≲ 0, which makes x 1 ln k+ln p ≪ 1 for x >1

    Advantage of partitioning The explicit expression of the ratio between optimally partitioned QFI and the monolithic QFI, and corresponding approximation are as follows Floss,2(F, k, p, n, m∗ loss,2) Floss,2(F, k, p, n) ≈ (2n −1) [k + F (2n −2) kn] h F −k(2k) 1 ln k+ln p i2 en(...

  100. [108]

    (C1) The derivation of the series is straightforward

    Spin-off: elementary representation of a hypergeometric function Recall that when deriving the effect of dephasing on sensing performance we encountered a series ˜q = X even # Z n #Z p#Z = ⌊n/2⌋X i=0 n 2i qn−2i(1 − q)2i = 1 + (2q − 1)n 2 . (C1) The derivation of the series is ...

  101. [109]

    Optimal number of sensors Now we would like to evaluate the partial derivative of the QFI with respect ton to determine the optimal total number of sensors. km2 2 n m −1 2 k+F k n m 2 n m −2 2 n F (2k) n m −k (2q − 1) 2n m ∂ ∂n Fdp(F, k, q, n, m) 30 = F 2(2k2) n m h 4m+m2 2n m...

  102. [110]

    Optimal number of sub-ensembles Here we evaluate the partial derivative of the QFI with respect tom to determine the optimal partition number. km3 2 n m −1 2 k+F k n m 2 n m −2 2 n2 k − F (2k) n m (2q − 1) 2n m ∂ ∂m Fdp(F, k, q, n, m) = k2 ((((((((m + 2n ln(2q − 1) − 2 n m (2n...

  103. [111]

    We have used the fact that in practicek, q≲ 1 so that ln k, ln(2q − 1) ≲ 0, which makes x −1 ln k+2 ln(2q−1) ≫ 1 for x >1

    Advantage of partitioning The explicit expression of the ratio between optimally partitioned QFI and the monolithic QFI, and corresponding approximation are as follows Fdp(F, k, q, n, m∗ dp) Fdp(F, k, q, n) ≈ − (2n −1) [ k+F (2n −2) kn] h k − F (2k) −1 ln k+2 ln(2q−1) i2 (2q −...

  104. [112]

    1 2eηt + F k n m − k k 1 +eηt 2eηt 2e−ηt 1 +e−ηt n m # (2m − nηt) + (m − nηt) ! . (D4) Therefore, the equation∂Floss,1(t)/∂k = 0is approximated by

    Sensor loss - without loss detection a. Monotonicity of QFI with initial state preparation quality We prove that higherk leads to higherFloss,1(t) by taking the partial derivative with respect tok as ∂ ∂k Floss,1(t) = F k2 (n − m) 4ke−2ηt n m F (2k) n m − k n2t2 2 n m − 1 h mk...

  105. [113]

    (D9) Obviously the fraction on the right hand side of the above equation is non-negative

    Sensor loss - with loss detection We prove that higherk leads to higherFloss,2(t) by taking the partial derivative with respect tok as ∂ ∂k Floss,2(t) = F n2(n − m) F (2k) n m − k e− n m ηtt2 m2k2 2 n m − 1 F 2 n m − 2 k n m + k 2 F (2k) n m 2 n m − 2 + k 3 × 2 n m − 2 . (D9) ...

  106. [114]

    Highest achievable QFI Here we present the exact expression of the highest achievable QFI when combining sensor loss and dephasing, and also assuming loss detectability. Floss,dp(t∗ loss,dp) = 4m F (2k) n m −k 2 e2k 2 n m −1 F 2 n m −2 k n m +k (2γ + η)2 ≈ 4F ek(2γ + η)2 k n m...

  107. [115]

    ∂ ∂t Floss,dp(t) = 2n2 F (2k) n m − k 2 mk 2 n m − 1 F 2 n m − 2 k n m + k t + O(t2)

    Short time QFI Here we examine thet derivative of the QFI at early evolution timest → 0. ∂ ∂t Floss,dp(t) = 2n2 F (2k) n m − k 2 mk 2 n m − 1 F 2 n m − 2 k n m + k t + O(t2). (D11) We are interested in how the factor of the linear term change whenm changes, so we take itsm der...

  108. [116]

    Highest QFI per unit time Here we present the exact expression of the highest achievable QFI when combining sensor loss and dephasing, and also assuming loss detectability

    Sequential scheme a. Highest QFI per unit time Here we present the exact expression of the highest achievable QFI when combining sensor loss and dephasing, and also assuming loss detectability. Floss,dp(˜t∗ loss,dp) ˜t∗ loss,dp = n F (2k) n m −k 2 ek 2 n m −1 F 2 n m −2 k n m ...

  109. [117]

    Comparison with spin squeezed states Here we provide the derivations for GHZ Ramsey spectroscopy performance. Recall that the probability to measure 0 for n-qubit noisy GHZ Ramsey is P0 = 1 2 + F kn−1V (n)[2˜q(t) − 1] 2 cos(nωt), (D27) 35 with ˜q(t) = (1 +e−nγt)/2 and V (n) = ...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.