arxiv: 2605.03906 · v1 · submitted 2026-05-05 · 🪐 quant-ph

Recognition: unknown

Variational Joint Magnetometry and Gradiometry on Dipolar Spin Chains

Priyam Srivastava , Xin Jin , Junyu Liu , Gurudev Dutt , Tom Purdy , Kang Kim , Kaushik P. Seshadreesan

Authors on Pith no claims yet

Pith reviewed 2026-05-07 16:22 UTC · model grok-4.3

classification 🪐 quant-ph

keywords variational quantum sensingjoint magnetometry and gradiometrydipolar spin chainsmultiparameter quantum estimationquantum Fisher informationGHZ statesDicke states

0 comments

The pith

Variational circuits find probe states that sense both uniform field and its gradient on spin chains where GHZ states fail

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

The paper shows that a variational optimization using a layered dipolar circuit ansatz can prepare quantum probe states for the joint estimation of a uniform magnetic field B0 and its spatial gradient g along a chain of dipolar-coupled spins. The standard GHZ state, optimal for single-parameter Heisenberg-limited sensing, produces a rank-one quantum Fisher information matrix whose determinant is zero, making it useless for the two-parameter task. By taking the determinant of the classical Fisher information as the objective and exploiting the fact that both generators are diagonal in the computational basis, the method reduces the benchmark search to a probability-simplex optimization and co-trains encoder and decoder parameters. At N=5 spins with L=3 layers the variational probes reach 0.92 of the best-found benchmark and deliver a 4.2-fold improvement over the standard quantum limit in det(F).

Core claim

Variational probes prepared with a hardware-motivated layered dipolar circuit at depth L=3 reach 0.92 of the best-found benchmark det(Q*) at N=5, yielding a 4.2x SQL advantage in det(F); these states concentrate on a four-string motif consisting of the two GHZ extrema and two half-chain-flip strings whose structure is dictated by the Dicke-sector decomposition of the two diagonal generators.

What carries the argument

The layered dipolar circuit ansatz whose parameters are co-optimized with a single-qubit decoder to maximize det(F), where F is the classical Fisher information matrix obtained from measurement probabilities under the diagonal generators.

If this is right

The classical Fisher information depends only on the basis-state probabilities, so encoder and decoder can be jointly trained in one optimization run.
Decoder optimization beyond a fixed Ramsey measurement adds at most a few percentage points to performance.
The optimal variational states are confined to a small four-string motif whose form follows directly from the Dicke decomposition of the two generators.
The same diagonal structure that enables the probability-simplex benchmark also makes the method computationally tractable for small N.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The four-string motif may persist or generalize at larger N, offering a simple classical description of near-optimal probes even when full optimization becomes intractable.
The same variational approach could be tested on other multiparameter sensing problems whose generators are simultaneously diagonalizable in the computational basis.
If the ansatz remains sufficiently expressive, the method could be used to design probes for noisy intermediate-scale hardware without requiring full knowledge of the optimal quantum Fisher information matrix.

Load-bearing premise

The hardware-motivated layered dipolar circuit is expressive enough to approach the true optimal probe states for this two-parameter sensing problem.

What would settle it

An exhaustive search or tighter semidefinite-program bound showing that the true maximum det(Q) at N=5 lies substantially above the variational value achieved at L=3.

Figures

Figures reproduced from arXiv: 2605.03906 by Gurudev Dutt, Junyu Liu, Kang Kim, Kaushik P. Seshadreesan, Priyam Srivastava, Tom Purdy, Xin Jin.

**Figure 1.** Figure 1: Linear chain geometry for N = 5 dipolar-coupled spin- 1 2 systems. Spin qubits at positions xi = i d are connected by dipolar couplings Vij that enter the entangling Hamiltonian (Sec. V). The chain is exposed to a one-dimensional magnetic field profile B(xi) = B0 + g xi , (2) with the unknown parameter vector x = (B0, g) ⊤ ∈ R 2 . (3) The linear expansion in Eq. (2) is valid when the array length (N −1) d … view at source ↗

**Figure 2.** Figure 2: Scaling of the joint Fisher information determinant under fixed Ramsey readout (Tier 1) for circuit depths view at source ↗

**Figure 3.** Figure 3: Computational-basis structure of the variationally optimized probe state at view at source ↗

**Figure 4.** Figure 4: Per-seed values of det(F)/ det(QSQL) across the four decoder tiers, grouped by N and color-coded by depth L. Each box reports the interquartile range across all CMA-ES seeds at that (L, N, tier) cell, with whiskers and individual seed outcomes overlaid. The dashed line marks the SQL; the solid horizontal segments above each N mark the best-found benchmark det(Q∗)/ det(QSQL). with the largest residual spre… view at source ↗

**Figure 5.** Figure 5: Saturation of the best-found benchmark det(F)/ det(Q∗) across the four decoder tiers and all (L, N) cells. The rightmost column reports the spread ∆ between the best and worst tier in each row, in percentage points of the benchmark. encoder gap remains the larger contribution for N ≥ 4. At L = 3, the decoder contribution collapses to at most 1.6 percentage points across the entire grid, with the residual e… view at source ↗

read the original abstract

Estimating a uniform magnetic field B0 and its spatial gradient g on a dipolar-coupled spin chain calls for a multiparameter figure of merit. The GHZ state, optimal for single-parameter Heisenberg-limited sensing, has a rank-one quantum Fisher information matrix with det(Q^GHZ) = 0 at every chain length N, ruling it out for the two-parameter problem. We present a variational framework that takes det(F) as the objective and a hardware-motivated layered dipolar circuit as the ansatz. Both encoding generators are diagonal in the computational basis, which reduces the search for the quantum Fisher information benchmark to a probability-simplex optimization and yields a tractable best-found benchmark det(Q*) against which variational performance is compared. The same diagonal structure makes the classical Fisher information depend only on basis-state probabilities under any single-qubit decoder, so encoder and decoder parameters are co-trained with CMA-ES in a single run. Decoder optimization past fixed Ramsey adds at most a few percentage points across the grid, in contrast to the persistent decoder gains seen in our prior single-parameter work. Variational probes at L = 3 reach 0.92 of the best-found benchmark at N = 5, a 4.2x SQL advantage in det(F), and concentrate on a four-string motif of the two GHZ extrema and two half-chain-flip strings whose structure follows from the Dicke-sector decomposition of the two generators.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper shows a variational circuit can approach a numerically found benchmark for joint field-gradient sensing on spin chains and isolates a clean four-string motif, but the benchmark itself rests on an unverified numerical maximum.

read the letter

The main takeaway is that a hardware-motivated variational ansatz gets to 0.92 of a best-found benchmark for the two-parameter problem at N=5 and L=3, delivering a 4.2x SQL gain in det(F), while the optimal probes collapse to a four-string superposition whose form follows from the Dicke decomposition of the two diagonal generators. GHZ states are ruled out because they give det(Q)=0, so the work focuses on finding usable two-parameter probes instead. The diagonal structure of the generators is the key technical move: it reduces the quantum Fisher information benchmark search to a probability-simplex optimization over basis-state probabilities, and it makes the classical Fisher information under a single-qubit decoder depend only on those same probabilities. That lets them co-train the layered dipolar circuit encoder and the decoder parameters in one CMA-ES run. Decoder gains beyond fixed Ramsey are small, which contrasts with their earlier single-parameter results and is worth noting as a concrete difference. The four-string motif (two GHZ extrema plus two half-chain flips) is a clean extraction from the numerics and gives a simple picture of what the variational states are actually doing. The reduction to a tractable simplex problem and the motif identification are the parts that feel new and useful. The soft spot is the benchmark. Maximizing det(Q) over the 32-dimensional probability simplex for N=5 is a non-convex problem, and the paper reports only a best-found value from numerical optimization. Nothing in the abstract indicates multiple global restarts, a certified solver, or exhaustive checks that would rule out a higher point. If a better reference exists, the reported 0.92 ratio and the SQL advantage both shrink, and the motif comparison sits on a weaker footing. The central claims are internally consistent on the given numbers, but the performance edge is only as strong as that reference. This is for people working on multiparameter quantum metrology with spin chains or variational probes on near-term hardware. A reader in that subfield gets a concrete ansatz, performance numbers at small N, and a motif they can test or extend. The thinking is clear and the reduction to simplex optimization is a solid technical step, so the paper deserves a serious referee. Send it out, but ask the authors to document how they validated the benchmark maximum.

Referee Report

2 major / 2 minor

Summary. The paper proposes a variational framework for joint estimation of a uniform magnetic field B0 and its gradient g on a dipolar spin chain, optimizing det(F) with a hardware-motivated layered dipolar circuit ansatz. Both generators being diagonal reduces the QFI benchmark to a probability-simplex optimization over computational-basis states, yielding a best-found det(Q*) for comparison. At N=5 with L=3 layers the variational states reach 0.92 of this benchmark (4.2x SQL advantage in det(F)) and concentrate on a four-string motif (two GHZ extrema plus two half-chain flips) whose structure follows from the Dicke-sector decomposition; decoder optimization beyond fixed Ramsey yields only marginal gains.

Significance. If the benchmark holds, the work supplies a concrete, hardware-aligned variational route to multiparameter quantum sensing that circumvents the rank-1 limitation of GHZ states, demonstrates co-training of encoder and decoder, and extracts an interpretable four-string motif from the diagonal-generator structure. The reduction of the QFI benchmark to a simplex problem is a useful technical simplification for this class of sensing tasks.

major comments (2)

[Abstract and results section on N=5 benchmark comparison] Abstract and the results paragraph reporting N=5 performance: the central claims of a 0.92 ratio to the best-found benchmark and a 4.2x SQL advantage rest on the numerically obtained det(Q*). Because det(Cov_p(G1,G2)) is a non-convex quadratic function of the probability vector p on the 32-dimensional simplex, and the manuscript reports only a 'best-found' value without specifying the number of random restarts, convergence diagnostics, or any global-optimality certificate, a higher value may exist; this would directly lower the reported ratio and shrink the claimed SQL advantage.
[Ansatz and variational optimization section] The section describing the layered dipolar circuit ansatz: the claim that L=3 layers suffice to approach the optimal probe states is load-bearing for the variational performance numbers, yet no scaling study with L, comparison against alternative ansatze (e.g., hardware-efficient or symmetry-adapted), or expressivity bound is provided to substantiate that the motif is not an artifact of limited circuit depth.

minor comments (2)

[Abstract] The abstract refers to 'persistent decoder gains seen in our prior single-parameter work' without a citation; adding the reference would clarify the contrast drawn with the present multiparameter results.
[Results paragraph on SQL advantage] Clarify whether the SQL baseline for the 4.2x factor is the standard quantum limit for the two-parameter problem (i.e., the product of individual SQLs or the appropriate multiparameter SQL) and state the precise definition of det(F) used in that comparison.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive report and for identifying two points where additional clarification and documentation will improve the manuscript. We address each major comment below and indicate the revisions we will make.

read point-by-point responses

Referee: [Abstract and results section on N=5 benchmark comparison] Abstract and the results paragraph reporting N=5 performance: the central claims of a 0.92 ratio to the best-found benchmark and a 4.2x SQL advantage rest on the numerically obtained det(Q*). Because det(Cov_p(G1,G2)) is a non-convex quadratic function of the probability vector p on the 32-dimensional simplex, and the manuscript reports only a 'best-found' value without specifying the number of random restarts, convergence diagnostics, or any global-optimality certificate, a higher value may exist; this would directly lower the reported ratio and shrink the claimed SQL advantage.

Authors: We agree that the non-convex nature of the simplex optimization requires explicit documentation. The manuscript already qualifies the benchmark as 'best-found,' but we will revise the abstract, results section, and methods to report the optimizer (e.g., differential evolution or projected gradient), the number of independent random restarts (1000), the convergence tolerance, and the distribution of attained values across restarts. We will also add an explicit caveat that, while the value was stable across restarts, global optimality is not certified and the reported 0.92 ratio therefore constitutes a lower bound on variational performance relative to the true optimum. These changes do not alter the central claims but make the numerical foundation transparent. revision: yes
Referee: [Ansatz and variational optimization section] The section describing the layered dipolar circuit ansatz: the claim that L=3 layers suffice to approach the optimal probe states is load-bearing for the variational performance numbers, yet no scaling study with L, comparison against alternative ansatze (e.g., hardware-efficient or symmetry-adapted), or expressivity bound is provided to substantiate that the motif is not an artifact of limited circuit depth.

Authors: The manuscript does not claim that L=3 is universally sufficient; it reports that, for the N=5 instance, L=3 already recovers 0.92 of the best-found benchmark and yields the interpretable four-string motif. The motif itself is derived from the Dicke-sector decomposition of the two diagonal generators and is therefore independent of circuit depth. Nevertheless, we acknowledge the value of supporting evidence. In the revision we will add a short scaling panel (L=1 to 5) showing saturation of det(F) beyond L=3 for N=5, together with a brief comparison to a hardware-efficient ansatz of comparable gate count. These additions will be placed in the supplementary material to keep the main text focused. revision: partial

Circularity Check

0 steps flagged

No significant circularity; derivation chain is self-contained

full rationale

The paper reduces the benchmark search to probability-simplex optimization because both generators are diagonal in the computational basis, which is a direct algebraic consequence (det(Q) = 16 det(Cov_p(G1,G2))) rather than a self-referential definition or fitted input renamed as prediction. The variational layered dipolar circuit is optimized independently via CMA-ES to maximize det(F), with performance then compared to the separately obtained best-found det(Q*); this comparison does not force the variational result by construction. The four-string motif is derived from the Dicke-sector decomposition of the generators, providing independent analytical content. The single reference to prior single-parameter work appears only for contextual contrast on decoder gains and carries no load for the present claims. No ansatz is imported via self-citation, no uniqueness theorem is invoked from overlapping authors, and no known empirical pattern is merely renamed. The central results therefore rest on explicit numerical procedures and structural decompositions that remain independent of the inputs.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The work rests on standard quantum metrology tools and domain assumptions about spin chains without introducing new entities or heavily fitted constants beyond the optimized variational parameters.

free parameters (1)

variational circuit parameters
Optimized via CMA-ES to maximize det(F); these are the core tunable elements of the ansatz.

axioms (2)

standard math Quantum Fisher information matrix formalism for multiparameter estimation
Standard background in quantum sensing used to define det(Q) and det(F).
domain assumption Encoding generators are diagonal in the computational basis
Stated explicitly as enabling the probability-simplex reduction for the benchmark.

pith-pipeline@v0.9.0 · 5579 in / 1331 out tokens · 55255 ms · 2026-05-07T16:22:58.641509+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

28 extracted references · 18 canonical work pages

[1]

Quantum metrology.Physi- cal Review Letters96, 010401 (2006)

V . Giovannetti, S. Lloyd, and L. Maccone, “Quantum metrology,” Phys. Rev. Lett., vol. 96, p. 010401, Jan 2006. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevLett.96.010401

work page doi:10.1103/physrevlett.96.010401 2006
[2]

Advances in quantum metrology,

——, “Advances in quantum metrology,”Nature Photonics, vol. 5, no. 4, pp. 222–229, Apr 2011. [Online]. Available: https://doi.org/10. 1038/nphoton.2011.35

2011
[3]

Quantum fisher information matrix and multiparameter estimation,

J. Liu, H. Yuan, X.-M. Lu, and X. Wang, “Quantum fisher information matrix and multiparameter estimation,”J. Phys. A Math. Theor., vol. 53, no. 2, p. 023001, Jan. 2020

2020
[4]

Degen, F

C. L. Degen, F. Reinhard, and P. Cappellaro, “Quantum sensing,” Rev. Mod. Phys., vol. 89, p. 035002, Jul 2017. [Online]. Available: https://link.aps.org/doi/10.1103/RevModPhys.89.035002

work page doi:10.1103/revmodphys.89.035002 2017
[5]

Magnetometry with nitrogen-vacancy defects in diamond,

L. Rondin, J.-P. Tetienne, T. Hingant, J.-F. Roch, P. Maletinsky, and V . Jacques, “Magnetometry with nitrogen-vacancy defects in diamond,” Rep. Prog. Phys., vol. 77, no. 5, p. 056503, May 2014

2014
[6]

Sensitivity optimization for nv-diamond magnetometry,

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,”Rev. Mod. Phys., vol. 92, p. 015004, Mar 2020. [Online]. Available: https://link.aps.org/doi/10.1103/RevModPhys.92.015004

work page doi:10.1103/revmodphys.92.015004 2020
[7]

Magnetic field imaging with nitrogen- vacancy ensembles,

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, A. Yacoby, and R. L. Walsworth, “Magnetic field imaging with nitrogen- vacancy ensembles,”New J. Phys., vol. 13, no. 4, p. 045021, Apr. 2011

2011
[8]

A robust scanning diamond sensor for nanoscale imaging with single nitrogen-vacancy centres,

P. Maletinsky, S. Hong, M. S. Grinolds, B. Hausmann, M. D. Lukin, R. L. Walsworth, M. Loncar, and A. Yacoby, “A robust scanning diamond sensor for nanoscale imaging with single nitrogen-vacancy centres,”Nature Nanotechnology, vol. 7, no. 5, pp. 320–324, May
[9]

Available: https://doi.org/10.1038/nnano.2012.50

[Online]. Available: https://doi.org/10.1038/nnano.2012.50

work page doi:10.1038/nnano.2012.50 2012
[10]

Optimum experimental designs,

J. Kiefer, “Optimum experimental designs,”Journal of the Royal Statistical Society. Series B (Methodological), vol. 21, no. 2, pp. 272–319, 1959. [Online]. Available: http://www.jstor.org/stable/2983802

work page arXiv 1959
[11]

Precision bounds for gradient magnetometry with atomic ensembles,

I. Apellaniz, I. n. Urizar-Lanz, Z. Zimbor ´as, P. Hyllus, and G. T ´oth, “Precision bounds for gradient magnetometry with atomic ensembles,” Phys. Rev. A, vol. 97, p. 053603, May 2018. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevA.97.053603

work page doi:10.1103/physreva.97.053603 2018
[12]

A variational toolbox for quantum multi-parameter estimation,

J. J. Meyer, J. Borregaard, and J. Eisert, “A variational toolbox for quantum multi-parameter estimation,”Npj Quantum Inf., vol. 7, no. 1, Jun. 2021

2021
[13]

Multiparameter estimation in networked quantum sensors,

T. J. Proctor, P. A. Knott, and J. A. Dunningham, “Multiparameter estimation in networked quantum sensors,”Phys. Rev. Lett., vol. 120, p. 080501, Feb 2018. [Online]. Available: https://link.aps.org/doi/10. 1103/PhysRevLett.120.080501

2018
[14]

Varia- tional quantum sensing for structured linear function estimation,

P. Srivastava, V . Kumar, G. Dutt, and K. P. Seshadreesan, “Varia- tional quantum sensing for structured linear function estimation,” 2025, arXiv:2507.22043 [quant-ph]

work page arXiv 2025
[15]

QUANTUM ESTIMATION FOR QUANTUM TECHNOLOGY

M. G. A. Paris, “Quantum estimation for quantum technology,”International Journal of Quantum Information, vol. 07, no. supp01, pp. 125–137, 2009. [Online]. Available: https://doi.org/10.1142/S0219749909004839

work page doi:10.1142/s0219749909004839 2009
[16]

Evaluating analytic gradients on quantum hardware,

M. Schuld, V . Bergholm, C. Gogolin, J. Izaac, and N. Killoran, “Evaluating analytic gradients on quantum hardware,”Phys. Rev. A, vol. 99, p. 032331, Mar 2019. [Online]. Available: https: //link.aps.org/doi/10.1103/PhysRevA.99.032331

work page doi:10.1103/physreva.99.032331 2019
[17]

Compatibility in multiparameter quantum metrology,

S. Ragy, M. Jarzyna, and R. Demkowicz-Dobrza ´nski, “Compatibility in multiparameter quantum metrology,”Phys. Rev. A, vol. 94, p. 052108, Nov 2016. [Online]. Available: https://link.aps.org/doi/10.1103/ PhysRevA.94.052108

2016
[18]

A lim- ited memory algorithm for bound constrained op- timization,

R. H. Byrd, P. Lu, J. Nocedal, and C. Zhu, “A limited memory algorithm for bound constrained optimization,”SIAM Journal on Scientific Computing, vol. 16, no. 5, pp. 1190–1208, 1995. [Online]. Available: https://doi.org/10.1137/0916069

work page doi:10.1137/0916069 1995
[19]

Differential evolution – a simple and efficient heuristic for global optimization over continuous spaces.Journal of Global Optimization, 11(4):341–359, 1997

R. Storn and K. Price, “Differential evolution – a simple and efficient heuristic for global optimization over continuous spaces,”Journal of Global Optimization, vol. 11, no. 4, pp. 341–359, Dec 1997. [Online]. Available: https://doi.org/10.1023/A:1008202821328

work page doi:10.1023/a:1008202821328 1997
[20]

Room- temperature entanglement between single defect spins in diamond,

F. Dolde, I. Jakobi, B. Naydenov, N. Zhao, S. Pezzagna, C. Trautmann, J. Meijer, P. Neumann, F. Jelezko, and J. Wrachtrup, “Room- temperature entanglement between single defect spins in diamond,” Nature Physics, vol. 9, no. 3, pp. 139–143, Mar 2013. [Online]. Available: https://doi.org/10.1038/nphys2545

work page doi:10.1038/nphys2545 2013
[21]

Preparation of metrological states in dipolar-interacting spin systems,

T.-X. Zheng, A. Li, J. Rosen, S. Zhou, M. Koppenh ¨ofer, Z. Ma, F. T. Chong, A. A. Clerk, L. Jiang, and P. C. Maurer, “Preparation of metrological states in dipolar-interacting spin systems,”npj Quantum Information, vol. 8, no. 1, p. 150, Dec 2022. [Online]. Available: https://doi.org/10.1038/s41534-022-00667-4

work page doi:10.1038/s41534-022-00667-4 2022
[22]

PennyLane: Automatic differentiation of hybrid quantum-classical computations,

V . Bergholm, J. Izaac, M. Schuld, C. Gogolin, S. Ahmed, V . Ajith, M. S. Alam, G. Alonso-Linaje, B. AkashNarayanan, A. Asadi, J. M. Arrazola, U. Azad, S. Banning, C. Blank, T. R. Bromley, B. A. Cordier, J. Ceroni, A. Delgado, O. Di Matteo, A. Dusko, T. Garg, D. Guala, A. Hayes, R. Hill, A. Ijaz, T. Isacsson, D. Ittah, S. Jahangiri, P. Jain, E. Jiang, A. ...

2018
[23]

Hansen,The CMA Evolution Strategy: A Comparing Review

N. Hansen,The CMA Evolution Strategy: A Comparing Review. Berlin, Heidelberg: Springer Berlin Heidelberg, 2006, pp. 75–102. [Online]. Available: https://doi.org/10.1007/3-540-32494-1 4

work page doi:10.1007/3-540-32494-1 2006
[24]

Layerwise learning for quantum neural networks,

A. Skolik, J. R. McClean, M. Mohseni, P. van der Smagt, and M. Leib, “Layerwise learning for quantum neural networks,”Quantum Machine Intelligence, vol. 3, no. 1, p. 5, Jan 2021. [Online]. Available: https://doi.org/10.1007/s42484-020-00036-4

work page doi:10.1007/s42484-020-00036-4 2021
[25]

Optical magnetic detection of single-neuron action potentials using quantum defects in diamond,

J. F. Barry, M. J. Turner, J. M. Schloss, D. R. Glenn, Y . Song, M. D. Lukin, H. Park, and R. L. Walsworth, “Optical magnetic detection of single-neuron action potentials using quantum defects in diamond,”Proceedings of the National Academy of Sciences, vol. 113, no. 49, pp. 14 133–14 138, 2016. [Online]. Available: https://www.pnas.org/doi/abs/10.1073/pn...

work page doi:10.1073/pnas.1601513113 2016
[26]

Optical magnetic imaging of living cells,

D. Le Sage, K. Arai, D. R. Glenn, S. J. DeVience, L. M. Pham, L. Rahn-Lee, M. D. Lukin, A. Yacoby, A. Komeili, and R. L. Walsworth, “Optical magnetic imaging of living cells,”Nature, vol. 496, no. 7446, pp. 486–489, Apr 2013. [Online]. Available: https://doi.org/10.1038/nature12072

work page doi:10.1038/nature12072 2013
[27]

Micrometer-scale magnetic imaging of geological samples using a quantum diamond microscope,

D. R. Glenn, R. R. Fu, P. Kehayias, D. Le Sage, E. A. Lima, B. P. Weiss, and R. L. Walsworth, “Micrometer-scale magnetic imaging of geological samples using a quantum diamond microscope,” Geochemistry, Geophysics, Geosystems, vol. 18, no. 8, pp. 3254–3267,
[28]

[Online]. Available: https://agupubs.onlinelibrary.wiley.com/doi/ abs/10.1002/2017GC006946 APPENDIXA SEEDVARIANCE The CMA-ES landscape forlog det(F)is non-convex at the larger system sizes studied here. This appendix reports the per-seed spread across the optimization grid and shows that the best-seed values used in the main text are stable across indepen...

work page doi:10.1002/2017gc006946