Time-averaged continuous quantum measurement

Antoine Tilloy; Pierre Guilmin; Pierre Rouchon

REVIEW 2 major objections 5 minor 2 cited by

Time-averaged continuous quantum measurement

T0 review · 2 major / 5 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 This paper derives an exact map $K_I(\rho)$ that produces the best Bayesian estimate of a quantum state from a binned, time-averaged measurement record for any bin duration $\Delta t$, and shows it can be evaluated exactly or expanded to…

desk verdict A clean, self-contained derivation of an exact Bayesian filter for digitized continuous measurement records; worth engaging despite its narrow ideal-signal scope. read the letter →

arxiv 2505.20382 v1 pith:MQYE4ZLR submitted 2025-05-26 quant-ph

Pierre Guilmin , Pierre Rouchon , Antoine Tilloy This is my paper · ORCID

classification quant-ph PACS 03.65.Ta03.65.Yz

keywords continuousquantummeasurementstochasticmasterequationBayesianstateestimationreconstructionparametertiltedLiouvilliantrajectory

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

Experiments record continuous quantum measurements only as digitized bins: averages of the instantaneous signal over intervals of duration $\Delta t$. The paper's central claim is that the best Bayesian estimate of the quantum state given only these bins—not the unobservable continuous trajectory—can be obtained by an exact, completely positive map that iterates in time and is valid for any bin duration, however large. This matters because standard practice treats the binned signal as though it were the continuous one, which is only approximately correct for very small $\Delta t$ and can produce unphysical states or biased parameter estimates. The paper shows the exact map can be evaluated numerically (Gauss-Hermite quadrature) or expanded in powers of $\sqrt{\Delta t}$ to arbitrary order, recovering and extending existing low-order discretization schemes. If correct, it removes the numerical-stability constraint on digitization, allowing coarser data storage and unbiased inference at the physically relevant timescale.

What carries the argument

The load-bearing object is the tilted Liouvillian $\mathcal{L}_j = \mathcal{L} + j\mathcal{C} + \frac{j^2}{2}\mathcal{I}$, a deformation of the measurement generator that, when exponentiated over a bin, propagates the state weighted by the Fourier phase of that bin's integrated signal. Its time-ordered exponential over $n$ bins factors into a product of single-bin superoperators $\Phi_p$, and because each $\Phi_p$ is linear, the Fourier integrals from the Dirac deltas collapse into a recursion on the previous linear robinet state. The completely positive map $K_I$ is the quantum instrument for a bin: it produces both the unnormalized conditional state and, by its trace, the exact probability density of the binned signal. The same tilted-generator trick supplies the jump-measurement version and the multi-channel version, and the Gaussian integration against the phase $e^{ipI}$ gives the Hermite-polynomial perturbative expansion in $\sqrt{\Delta t}$.

What would settle it

Run a homodyne or heterodyne experiment on a system with known Hamiltonian and monitoring strength, digitize at a coarse $\Delta t$ where Euler schemes fail, histogram many runs' bin integrals, and compare with $\operatorname{Tr}[K_{I}(\rho_0)]$ computed by Gauss-Hermite quadrature; any discrepancy beyond sampling error refutes exactness. A sharper version inserts a low-pass filter before digitization: the paper's premise says $K_I$ must change, so the unmodified map should visibly mispredict the histogram.

Watch

Extended reading notes

Core claim

The construction is built on the linear robinet state: an unnormalized state carrying the conditional expectation of $\rho_{n\Delta t}$ weighted by a product of Dirac deltas that enforce the measured bin integrals. By writing each delta as a Fourier integral, the expectation becomes a Gaussian-weighted Fourier integral of a state propagated by the tilted Liouvillian $\mathcal{L}_j = \mathcal{L} + j\mathcal{C} + (j^2/2)\mathcal{I}$ with piecewise-constant imaginary $j$ on each bin. Linearity of the resulting superoperator $\Phi_p = e^{\Delta t \mathcal{L}_{-ip}}$ makes the $n$-bin expression factorize, yielding the one-bin recurrence $\rho_k = K_{I_k}(\rho_{k-1})/\operatorname{Tr}[K_{I_k}(\rho_{k-1})]$ with $$K_I(\rho) = \frac{1}{2\pi}\int_{\mathbb{R}} dp\, e^{ipI - \$\Delta$ t $p^{2}$/2} e^{\$\Delta$ t(\mathcal{L} - ip\mathcal{C})}(\rho).$$ The normalization is the exact probability density of the binned signal $I_k$. The paper claims this is the exact Bayesian conditional average $E[\rho_{n\Delta t}\mid I_1,\dots,I_n]$ for any $\Delta t$, that it extends to jump measurements (with a bounded $p$-integral) and multiple monitored operators, and that its perturbative expansion reproduces the Milstein correction and known higher-order completely positive discretizations while extending them to arbitrary order; for unit efficiency, up to fifth order in $\sqrt{\Delta t}$ the map is a single Kraus operator.

Load-bearing premise

The construction assumes the recorded bin value is exactly the time integral of the ideal continuous measurement output obeying the stochastic master equation, with no detector bandwidth limit, filtering, dead time, or non-Markovian noise; if the physical record deviates from this perfect average, the map $K_I$ is not the true Bayesian update.

Editorial extensions

If this is right

State reconstruction from a digitized record at coarse $\Delta t$ yields the true best Bayesian estimate, where Euler and CPTP-1 schemes can produce unphysical states and fidelities lower than even the unobserved Lindblad average.
Maximum-likelihood parameter estimation built from the exact map is unbiased at fixed $\Delta t$, so binning can be chosen at the information-relevant timescale instead of the numerical-stability timescale.
Digitized quantum trajectories can be sampled directly at finite $\Delta t$ by drawing $I_k$ from $\operatorname{Tr}[K_{I_k}(\rho_{k-1})]$ and applying the update, with the perturbative expansion giving schemes of arbitrarily high order.
The perturbative expansion places existing discretization schemes, including the Milstein correction, on a common Bayesian footing and extends them to all orders in $\sqrt{\Delta t}$.
For perfect detection efficiency, purity loss from time averaging appears only at third order in $\Delta t$; up to fifth order in $\sqrt{\Delta t}$ the map remains a single Kraus operator.

Reading between the lines

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

Editorial inference: the exact map gives a natural generative model of coarse-grained trajectories; sampling bins from the exact signal density and updating produces statistics identical to the full stochastic master equation, which could be used for fast, memory-light simulation of long experiments.
Editorial inference: existing experimental datasets recorded at coarse $\Delta t$ could be re-analyzed with the exact map without re-running the experiment, potentially correcting state estimates and parameter biases in previously published results.
Editorial inference: the derivation's reliance on the ideal time-average suggests a concrete stress test—insert a calibrated filter before digitization and check whether a modified, filter-dependent map is needed; where the ideal premise fails, the method would need to be re-derived for realistic detector responses.
Editorial inference: the same Fourier/tilted-generator construction should apply to time-dependent Liouvillians and to nonlinear or colored-noise measurement records, though the paper explicitly treats only constant Liouvillians and ideal averaging.

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.

Desk Editor's Note

read the letter

The paper gives the exact recursive map for reconstructing a quantum state from time-averaged measurement records, which is the right object for experiments with finite digitization. The derivation is straightforward once you see it, but it's new and useful: the tilted-Liouvillian trick is standard, but the recursive construction and the explicit quadrature/perturbative evaluation are solid. The Itô computation is clean, the factorization is valid, and the numerical post-selection verification really checks the conditional average independently of the derivation. The applications show clear advantages over Euler and CPTP-1 schemes, and the perturbative expansion correctly reproduces known low-order terms while extending them to arbitrary order.

The soft spots are real but minor. The exactness is conditional on the digitized signal being the perfect time integral of the ideal diffusive SME output. Finite detector bandwidth, filtering, dead time, or non-Markovian noise break the map. That is a scope condition, not an error, but it should be stated more prominently than it is. Complete positivity of K_I is asserted rather than proved; it looks plausible from the physical construction, but a proof or at least a more careful statement would tighten the paper. The time-dependent generalization is mentioned in the appendix but not worked out; that is fine for a letter. No code or data are shipped, which is a minor reproducibility gap given the numerical claims.

Overall the central argument holds up. The paper is clearly written, the math is internally consistent, and the authors are honest about what the robinet state is and is not: it is the best estimate given the information actually available, not a reconstruction of the continuous trajectory. I found no load-bearing flaw.

Who is it for? Anyone working with continuous measurement in superconducting circuits or quantum optics where digitization time matters. It deserves a serious referee. I would recommend acceptance with minor revisions: add a short discussion of detector non-idealities, and either prove or explicitly defer the complete-positivity claim. I would cite this in future work on quantum filtering.

Referee Report

2 major / 5 minor

Summary. This paper addresses the problem of reconstructing the state of a continuously measured quantum system when only a time-averaged (digitized) measurement record is available, which is the situation in most experiments. The authors define the 'robinet state' as the Bayesian conditional expectation of the true state given the digitized record and derive an exact recursive update map K_I (Eqs. (13)-(14)) expressed as a Fourier integral of tilted-Liouvillian exponentials. They also derive a perturbative expansion in powers of sqrt(Δt), recovering and extending existing discretization schemes, and provide numerical verification via Monte Carlo post-selection. Applications to state reconstruction, parameter estimation, and trajectory sampling are presented.

Significance. The central result is a clean, exact formula for the quantum instrument corresponding to time-averaged continuous measurement, valid for arbitrary bin duration. This is an important contribution because it removes the need for the extremely fine binning currently required for unbiased state reconstruction and parameter estimation in experiments with superconducting circuits and other platforms. The derivation via Itô calculus is rigorous and the recursion is simple to implement. The perturbative expansion provides a systematic way to generate higher-order discretization schemes, and the authors demonstrate clear advantages over existing methods. The paper is well written and the numerical checks, while limited, support the claims.

major comments (2)

[Exact map, Eq. (14)] The map K_I is asserted to be completely positive (CP) without proof. The integral representation involves non-Hermitian generators e^{Δt(L - ipC)}, so CP is not manifest. Since K_I is called an 'exact quantum instrument' and is used to generate normalized states, please provide a proof of CP (e.g., via a Kraus representation or by showing the integral is a superposition of CP maps) or cite a reference where this is established.
[Verifying the formula, Figure 1] The Monte Carlo post-selection verification for a single point uses only 10 trajectories, giving a statistical uncertainty on the order of 30%. This is not a strong quantitative test of the formula. Please provide error bars or use a more efficient estimator (e.g., importance sampling) to demonstrate convergence as epsilon tends to zero.

minor comments (5)

[Verifying the formula, Figure 1] The figure caption states that 10 trajectories out of 10^6 are post-selected, and the text says that averaging these 10 states approximates the robinet state; the statistical significance of this comparison should be quantified.
[Evaluating the formula, Eq. (21)] The truncated polynomial-times-Gaussian density used for sampling may take negative values for some values of \bar{I}; please discuss how the authors ensure a valid proposal density in the rejection sampling procedure.
[Throughout] The symbol I is used both for the measurement signal and for the identity superoperator; please use a distinct symbol (e.g., \mathbb{I}) for the identity to avoid confusion.
[Derivation, Eq. (9)] The statement 'for brevity, we assume the Liouvillian is time-independent' is terse; it would be helpful to indicate explicitly how the formula generalizes to time-dependent L (via time-ordered exponentials) in the main text or appendix.
[Appendix, Perturbative expansion] The claim that the recursion cost is quadratic in the chosen order is not fully explained; please add a sentence on how the terms are recursively computed.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the exact robinet map K_I is derived from the SME model by explicit Itô calculus and linearity; the handful of self-citations are non-load-bearing.

full rationale

The central claim does not reduce to its inputs. Eq. (2) defines the robinet state as the Bayesian conditional expectation given the digitized record; eqs. (5)-(7) re-express this definition via delta constraints and Fourier integrals; eqs. (8)-(9) introduce the tilted Liouvillian L_j = L + jC + (j^2/2)I. That ODE is the only imported step, cited to the authors' prior work [4], but it is a one-line Itô computation (the exponential of the integrated signal picks up the (1/2)j^2 dt variance correction and the martingale term has zero mean), which I verified directly, so the citation is a pointer rather than load-bearing. Eq. (10) factorizes because the p^2/2 scalar superoperator commutes with L - ipC, and eq. (12) closes the recursion by linearity of Phi_p. No parameter is fitted anywhere: K_I in eq. (14) has no free parameters, and the perturbative expansion (17) is a Taylor expansion of this same map integrated with Hermite-polynomial identities in the appendix; it reproduces external results [6]-[8] and extends them, so it is not a renaming. The Monte Carlo verification post-selects fine-grained trajectories to estimate the same conditional average by an independent route, and the joint-law test p_{I1,I2} = Tr[K_{I2} K_{I1}(rho0)] is checked against raw simulation histograms, an externally grounded check of the derived likelihood. Self-citations ([4], [5], [6], [15], [21]) appear only as technique pointers, benchmark baselines (CPTP-1), and software credits; none carries the central argument. The paper honestly states its scope (digitization irreversibly loses information; the digitized signal is the exact integral of the ideal SME output), which is a model-condition, not a circularity, and the unproven complete-positivity assertion for K_I is a correctness presentation gap, not a circular step.

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

The central claim rests on the standard continuous-measurement SME model, Ito calculus, and the Markov property; no physical free parameters are fitted and no new physical entities are introduced. The robinet state is a mathematical definition, not a new physical object with independent evidence.

assumptions (5)

domain assumption The system and measurement are described by the Markovian stochastic master equations (3)-(4) with Wiener noise and Lindblad dissipation.
The derivation of the tilted Liouvillian, eqs (8)-(9), starts from this model; if the true detector noise or backaction differs, the robinet state is not the true conditional expectation.
standard math Ito calculus and the martingale properties of the Wiener process are used to derive the linear ODE for the tilted expectation in eq (9).
Standard mathematical background; not proved in the paper.
domain assumption The Liouvillian is time-independent within each bin, so that Phi_p = e^{Delta t (L - ipC)} in eq (10); the appendix indicates the time-dependent extension via time-ordered exponentials.
The main-text closed form assumes constant L; if L varies on timescales shorter than Delta t, the simple exponential is replaced by a time-ordered product, weakening the closed form as written.
domain assumption Markov property of the joint (state, record) process, so that the future increment depends only on the current state and future noise.
The recursive factorization rho_tilde_n = K_{I_n}(rho_tilde_{n-1}) uses the fact that evolution on the new bin is independent of the past given rho_{(n-1)Delta t}.
standard math Fourier representation of the Dirac delta and Gaussian/Hermite integral identities, eqs (7) and (28), are used without proof.
Standard analytical tools used to convert delta-function conditioning and Gaussian integrals into the map K_I.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Time-averaged continuous quantum measurement." pith.science (2026). https://pith.science/paper/MQYE4ZLR

@misc{pith2026250520382,
  author       = {Pith},
  title        = {Pith review of: Time-averaged continuous quantum measurement},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MQYE4ZLR}},
  note         = {Machine review of arXiv:2505.20382}
}

abstract

The theory of continuous quantum measurement allows to reconstruct the state $\rho_t$ of a system from a continuous stochastic measurement record $I_t$. However, this truly continuous-time signal $I_t$ is never available in practice. In experiments, one generally has access to its digitization, i.e., to a series of time averages $I_k$ over finite intervals of duration $\Delta t$. In this letter, we take this digitization seriously and define $\bar{\rho}_n$ as the best Bayesian estimate of the quantum state given (only) a digitized record $(I_1,\dots,I_n)$. We show that $\bar{\rho}_{n+1}$ can be computed recursively from $I_{n+1}$ and $\bar{\rho}_n$ using an exact formula. The latter can be evaluated numerically exactly, or used as the basis for a perturbative expansion into successive powers of $\sqrt{\Delta t}$. This allows reconstructing quantum trajectories in regimes of coarse $\Delta t$ where existing methods fail, estimating parameters at fixed $\Delta t$ without bias, and directly sampling digitized quantum trajectories with schemes of arbitrarily high order.

Figures

Figures reproduced from arXiv: 2505.20382 by the authors.

**Figure 3.** FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

Calculus of Robinet: completely positive reconstruction of time-averaged diffusive quantum trajectories
quant-ph 2026-07 accept novelty 7.0 of 10

A completely-positive numerical method computes the optimal time-averaged (Robinet) quantum trajectory from a system–transmission-line dilation to arbitrarily high order, plus a proof that finitely many extra binned r...
Optimal schedule of multi-channel quantum Zeno dragging with application to solving the k-SAT problem
quant-ph 2025-07 conditional novelty 6.0 of 10

Multi-channel Zeno dragging converges fastest in the weak continuous measurement limit, and optimal control finds schedules that beat linear interpolation.

Reference graph

Works this paper leans on

22 extracted references · 13 canonical work pages · cited by 2 Pith papers

[1]

H. M. Wiseman and G. J. Milburn,Quantum Measure- ment and Control(Cambridge University Press, 2009)

2009
[2]

binnedρ” in improper French, which gives “rho binn´ e

In French, robinet means faucet / tap. The name comes 6 from pronouncing “binnedρ” in improper French, which gives “rho binn´ e”, which sounds like robinet

work page
[3]

Jacobs and D

K. Jacobs and D. A. Steck, A straightforward introduc- tion to continuous quantum measurement, Contempo- rary Physics47, 279 (2006)

2006
[4]

Guilmin, P

P. Guilmin, P. Rouchon, and A. Tilloy, Param- eters estimation by fitting correlation functions of continuous quantum measurement, arXiv preprint 10.48550/ARXIV.2410.11955 (2024)

arXiv doi:10.48550/arxiv.2410.11955 2024
[5]

Rouchon, A tutorial introduction to quantum stochas- tic master equations based on the qubit/photon system, Annual Reviews in Control54, 252 (2022)

P. Rouchon, A tutorial introduction to quantum stochas- tic master equations based on the qubit/photon system, Annual Reviews in Control54, 252 (2022)

work page 2022
[6]

Rouchon and J

P. Rouchon and J. F. Ralph, Efficient quantum filtering for quantum feedback control, Physical Review A91, 012118 (2015)

work page 2015
[8]

Wonglakhon, H

N. Wonglakhon, H. M. Wiseman, and A. Chantasri, Completely positive trace-preserving maps for higher- order unraveling of lindblad master equations, Physical Review A110, 10.1103/physreva.110.062207 (2024)

work page doi:10.1103/physreva.110.062207 2024
[9]

K. W. Murch, S. J. Weber, C. Macklin, and I. Siddiqi, Observing single quantum trajectories of a superconduct- ing quantum bit, Nature502, 211–214 (2013)

work page 2013

Show all 22 references

[10]

S. J. Weber, A. Chantasri, J. Dressel, A. N. Jordan, K. W. Murch, and I. Siddiqi, Mapping the optimal route between two quantum states, Nature511, 570–573 (2014)

2014
[11]

S. J. Weber, K. W. Murch, M. E. Kimchi-Schwartz, N. Roch, and I. Siddiqi, Quantum trajectories of su- perconducting qubits, Comptes Rendus. Physique17, 766–777 (2016)

2016
[12]

Campagne-Ibarcq, P

P. Campagne-Ibarcq, P. Six, L. Bretheau, A. Sarlette, M. Mirrahimi, P. Rouchon, and B. Huard, Observ- ing quantum state diffusion by heterodyne detection of fluorescence, Physical Review X6, 10.1103/phys- revx.6.011002 (2016)

2016 doi
[13]

Ficheux, S

Q. Ficheux, S. Jezouin, Z. Leghtas, and B. Huard, Dy- namics of a qubit while simultaneously monitoring its relaxation and dephasing, Nature Communications9, 10.1038/s41467-018-04372-9 (2018)

2018 doi
[14]

Marquet, A

A. Marquet, A. Essig, J. Cohen, N. Cottet, A. Murani, E. Albertinale, S. Dupouy, A. Bienfait, T. Peronnin, S. Jezouin, R. Lescanne, and B. Huard, Autoparametric resonance extending the bit-flip time of a cat qubit up to 0.3 s, Physical Review X14, 10.1103/physrevx.14.021019 (2024)

2024 doi
[15]

A. N. Jordan, A. Chantasri, P. Rouchon, and B. Huard, Anatomy of fluorescence: quantum trajectory statis- tics from continuously measuring spontaneous emission, Quantum Studies: Mathematics and Foundations3, 237 (2016)

2016
[16]

Mabuchi, Dynamical identification of open quantum systems, Quantum and Semiclassical Optics: Journal of the European Optical Society Part B8, 1103–1108 (1996)

H. Mabuchi, Dynamical identification of open quantum systems, Quantum and Semiclassical Optics: Journal of the European Optical Society Part B8, 1103–1108 (1996)

1996
[17]

Gambetta and H

J. Gambetta and H. M. Wiseman, State and dynamical parameter estimation for open quantum systems, Physi- cal Review A64, 042105 (2001)

2001
[18]

Negretti and K

A. Negretti and K. Mølmer, Estimation of classical pa- rameters via continuous probing of complementary quan- tum observables, New Journal of Physics15, 125002 (2013)

2013
[19]

Gammelmark and K

S. Gammelmark and K. Mølmer, Bayesian parameter in- ference from continuously monitored quantum systems, Physical Review A87, 032115 (2013)

2013
[20]

P. Six, P. Campagne-Ibarcq, L. Bretheau, B. Huard, and P. Rouchon, Parameter estimation from measurements along quantum trajectories, in2015 54th IEEE Confer- ence on Decision and Control (CDC)(IEEE, 2015)

2015
[21]

Guilmin, A

P. Guilmin, A. Bocquet, ´E. Genois, D. Weiss, and R. Gautier, Dynamiqs: an open-source Python library for GPU-accelerated and differentiable simulation of quan- tum systems (2025)

2025
[22]

Bradbury, R

J. Bradbury, R. Frostig, P. Hawkins, M. J. Johnson, C. Leary, D. Maclaurin, G. Necula, A. Paszke, J. Van- derPlas, S. Wanderman-Milne, and Q. Zhang, JAX: com- posable transformations of Python+NumPy programs (2018)

2018
[23]

Kidger, On neural differential equations, arXiv preprint 10.48550/ARXIV.2202.02435 (2022)

P. Kidger, On neural differential equations, arXiv preprint 10.48550/ARXIV.2202.02435 (2022). 7 Generalization The general formula for the jump and/or diffusive monitoring ofmjump operators (L 1, . . . , Lm) with mea- surement recordJ k = (I (1) k , . . . , I(m) k ) at stepkis...

2022 doi arXiv

Pith tools

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

[1] [1]

H. M. Wiseman and G. J. Milburn,Quantum Measure- ment and Control(Cambridge University Press, 2009)

2009

[2] [2]

binnedρ” in improper French, which gives “rho binn´ e

In French, robinet means faucet / tap. The name comes 6 from pronouncing “binnedρ” in improper French, which gives “rho binn´ e”, which sounds like robinet

work page

[3] [3]

Jacobs and D

K. Jacobs and D. A. Steck, A straightforward introduc- tion to continuous quantum measurement, Contempo- rary Physics47, 279 (2006)

2006

[4] [4]

Guilmin, P

P. Guilmin, P. Rouchon, and A. Tilloy, Param- eters estimation by fitting correlation functions of continuous quantum measurement, arXiv preprint 10.48550/ARXIV.2410.11955 (2024)

arXiv doi:10.48550/arxiv.2410.11955 2024

[5] [5]

Rouchon, A tutorial introduction to quantum stochas- tic master equations based on the qubit/photon system, Annual Reviews in Control54, 252 (2022)

P. Rouchon, A tutorial introduction to quantum stochas- tic master equations based on the qubit/photon system, Annual Reviews in Control54, 252 (2022)

work page 2022

[6] [6]

Rouchon and J

P. Rouchon and J. F. Ralph, Efficient quantum filtering for quantum feedback control, Physical Review A91, 012118 (2015)

work page 2015

[7] [8]

Wonglakhon, H

N. Wonglakhon, H. M. Wiseman, and A. Chantasri, Completely positive trace-preserving maps for higher- order unraveling of lindblad master equations, Physical Review A110, 10.1103/physreva.110.062207 (2024)

work page doi:10.1103/physreva.110.062207 2024

[8] [9]

K. W. Murch, S. J. Weber, C. Macklin, and I. Siddiqi, Observing single quantum trajectories of a superconduct- ing quantum bit, Nature502, 211–214 (2013)

work page 2013