Pith. sign in

REVIEW 3 major objections 3 minor 43 references

Tracking Quantum Dynamics in an Optical Cavity for Recovering Purity and Squeezing via Quantum State Smoothing

T0 review · 3 major / 3 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read Using both past and future measurement records, quantum state smoothing experimentally recovers more purity and squeezing in a continuously monitored optical cavity than conventional filtering, and gives a better estimate of the hidden true

desk verdict First experimental realization of quantum state smoothing in a linear Gaussian optical system; the data strongly support the theory, and the central result is solid, though the absolute recovery percentages inherit an unquantified systematic caveat from the approximate 'true' state. read the letter →

arxiv 2509.04754 v1 pith:FTEW7GVR submitted 2025-09-05 quant-ph

Shota Yokoyama , Kiarn T. Laverick , David McManus , Qi Yu , Areeya Chantasri , Warit Asavanant , Daoyi Dong , Howard M. Wiseman
show 1 more author
Hidehiro Yonezawa
This is my paper · ORCID
classification quant-ph MSC 81P1681P1581V80 PACS 03.65.Ta42.50.Dv42.50.Lc
keywords quantumstatesmoothingfilteringlinearGaussiansystemsopticalparametricoscillatorcontinuousmeasurementsqueezingpurityrecoveryhomodynedetection
topics The Measurement Problem
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

Quantum state filtering, the standard way to track a monitored quantum system, uses only past records and leaves the estimated state impure when information escapes into unobserved channels. This paper experimentally shows that quantum state smoothing, which also uses future records and assumes a specific hidden-observer measurement on the lost channel, retroactively recovers some of that lost purity and squeezing. In an optical parametric oscillator whose output is split between two homodyne detectors, the smoothed state is purer, more squeezed, and on average closer to the dual-detector hidden state than the filtered state is. The largest measured improvements are 10.3%±1.6% of the recoverable purity and 7.6%±2.6% of the recoverable squeezing. The result makes smoothing a practical post-processing tool for quantum information tasks that can tolerate delayed estimates.

What carries the argument

The load-bearing mechanism is the linear-Gaussian smoothing identity that combines a filtered estimate from the past record with a retrofiltered element from the future record: V_S = V_T + (V_F - V_T)[I + Λ_R(V_F - V_T)]^{-1}, with Λ_R = (V_R + V_T)^{-1} built from the future record and the assumption that the hidden channel was a homodyne measurement at a specified angle. This identity converts a causal trajectory into a retroactive estimate whose covariance is smaller than the filtered covariance, which is why purity and squeezing recover, and why the smoothed mean lands closer to the hidden true state.

What would settle it

In a repeat run, deliberately mis-set the hidden detector's angle by about 20 degrees and see whether the purity recovery drops in the direction and magnitude predicted by the smoothing equations. If a mis-set angle leaves the recovery unchanged—or if smoothing stops beating filtering when the residual loss is assigned to a different measurement model—the assumption-dependence of the central claim collapses.

Watch

Extended reading notes

Core claim

The paper's central claim is that the acausal estimator called quantum state smoothing outperforms causal quantum filtering on three useful metrics in a real, linear-Gaussian optical system. The experiment continuously monitors an optical parametric oscillator through two homodyne detectors; one measurement record is used for estimation, while the other detector's record is withheld. The state conditioned on both records plays the role of the true state. Against that benchmark, the smoothed state—formed from the past record, the future record, and the assumption that the hidden channel was a homodyne measurement at a known angle—has greater purity, greater squeezing, and a smaller average tr

Load-bearing premise

The 'true' state used to score the estimates is itself only an estimate: it conditions on both homodyne records but still assumes the residual 14% loss can be neglected, so a different model of that loss could shift the reported recovery percentages.

Editorial extensions

If this is right

  • Stored measurement records can be reprocessed after the fact to retroactively purify squeezed states, so applications that keep records—such as measurement-based quantum computing—can use smoothing as a post-processing resource-recovery step.
  • At every tested efficiency and angle combination, smoothing performs at least as well as filtering, so switching from filtering to smoothing carries no penalty when real-time tracking is not required.
  • The amount of recovered purity and squeezing is controlled jointly by both measurement settings, and the optima for purity and squeezing are not reached simultaneously, so applications must choose which resource to favour.
  • Smoothed estimates are not merely cleaner-looking states: they reduce the average squared deviation from the hidden true state, giving a statistically better estimate of the system's underlying dynamics.

Reading between the lines

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

  • If the claim generalises, a natural test the paper does not report is to vary the total efficiency downward and check that the recovered-fraction numbers move as the 0.86-efficiency benchmark predicts; if they do not, the absolute percentages reflect the benchmark rather than a universal feature.
  • If the claim transfers to other linear-Gaussian platforms—mechanical resonators, circuit-QED readout, atomic ensembles—the same covariance equations would make this demonstration a generic recipe for recovering purity from stored continuous-measurement records.
  • Beyond the paper's stated scope, the experiment implies a diagnostic use: the hidden-observer assumption that yields the best smoothed estimate is the unraveling actually operative in the loss channel, so smoothing could be used to characterise loss channels rather than only to compensate for them.
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 / 3 minor

Summary. The paper reports an experimental demonstration of quantum state smoothing for a continuous-variable linear-Gaussian system: an optical parametric oscillator whose output is split into two homodyne channels, one observed by Alice (the estimator) and the other by a hidden observer Bob. Alice's smoothed estimate uses her past and future records, and is compared with her causal filtered estimate. The 'true' state is defined as Bob's state conditioned on both records, with total efficiency η_tot≈0.86. The authors measure purity, average trace-squared deviation (TrSD) from the true state, squeezing, and anti-squeezing as functions of Alice's efficiency and of both homodyne angles. Parameter-free theory curves (from independently measured efficiencies, loss, gain, and angles) agree with the data across four metrics and two parameter scans. The headline results are a maximum purity recovery of 0.016±0.003 (10.3%±1.6% of the total recoverable purity) and a maximum squeezing recovery of 0.006±0.002 (7.6%±2.6% of the total recoverable conditional squeezing). The authors conclude that smoothing yields purer, more squeezed states that are better estimates of the true state than filtering.

Significance. If the results hold, this is the first experimental realization of quantum state smoothing for a linear-Gaussian continuous-variable system, a milestone that had been awaited since the theoretical proposals. The paper's strengths are notable: the theory curves are parameter-free predictions using independently measured quantities; four complementary metrics (purity, TrSD, squeezing, anti-squeezing) are compared; the TrSD is computed independently of the reconstructed covariances as an internal consistency check; and the optimal measurement settings predicted by theory are tested rather than fitted. The qualitative conclusion that smoothing outperforms filtering is robust and well supported. The main weakness is that the absolute recovery percentages in the abstract are defined relative to an approximate 'true' state whose residual impurity is not propagated into systematic uncertainties.

major comments (3)
  1. [Results – Experimental setup; Discussion] The headline quantitative claims (abstract: 10.3%±1.6% purity recovery, 7.6%±2.6% squeezing recovery) are defined relative to the 'true' state ρ_T, which is conditioned on both Alice's and Bob's homodyne records with η_tot≈0.86. The paper concedes that η_tot is not unity, so ρ_T is not the ideal pure true state of quantum state smoothing. The reported error bars are statistical only and do not include the systematic effect of the residual ~14% loss, the choice of unraveling for that loss, or the ±2° phase noise. A different modeling of the unobserved loss would shift P(ρ_T) and therefore both the numerator and denominator of the recovery fractions. This does not undermine the qualitative smoothing-vs-filtering comparison, which is robust, but it does affect the absolute percentages featured in the abstract. Please quantify this systematic uncertainty or rephrase the headline claim to avo
  2. [Methods – Reconstruction of covariance, Eq. (7)] In Eq. (7), the reconstructed covariances \tilde V_F and \tilde V_S are obtained by adding the theoretically calculated V_T to the experimental mean-square-error matrices. This choice protects the filtering-vs-smoothing comparison from common noise in V_T, but the absolute purities and squeezing values used in Figs. 4–6, and especially the ratios R_P/[P(ρ_T)-P(ρ_F)] and R_S/[S(ρ_F)-S(ρ_T)], inherit any model error in V_T. Since V_T also enters the denominators, the percentage recoveries are not independent of the theoretical model. A systematic propagation of uncertainties in η_T, η_B, ξ, and V_T should be reported, or the claims should be restricted to the directly measured differences rather than the normalized percentages.
  3. [Data acquisition (phase scanning)] In the phase-scan measurements (Fig. 5 and Fig. 6), θ_B is swept at ~140 Hz and the data are binned into 1° blocks, with a stated control-loop phase noise of ±2°. These uncertainties are not propagated into the error bars shown in Fig. 5. Given that the theoretical predictions are evaluated at nominal angles, a brief statement that the agreement is robust to these phase uncertainties—or inclusion of them in the error bars—would strengthen the quantitative comparison.
minor comments (3)
  1. [Fig. 6 caption and axis] The x-axis label reads 'Measurement efficiency, η_A', but the tick values are 0, 45, 90, 135, 180 degrees and the caption describes a scan over θ_A. This appears to be a copy-paste from Fig. 4 and should be changed to 'Measurement angle, θ_A (deg)'.
  2. [Main text near Fig. 3; Methods Eq. (8)] The covariance decomposition is misstated. The text says V_unc = V_S + Cov(⟨x̂⟩_unc, ⟨x̂⟩_S) = V_S + Cov(⟨x̂⟩_S,0) since ⟨x̂⟩_unc=0. As written, Cov(⟨x̂⟩_S,0)=0, which would imply V_unc=V_S and contradict the data. The intended relation is V_unc = V_S + Cov(⟨x̂⟩_S,⟨x̂⟩_S). Similarly, Eq. (8) should be \tilde V_T = V_unc − Cov(⟨x̂⟩_T,⟨x̂⟩_T), not Cov(⟨x̂⟩_T,0).
  3. [Fig. 4 legend] The legend entry 'Unconditioning' should probably be 'Unconditional' or 'Unconditional state' for consistency with the text.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; smoothing predictions are tested against external homodyne records and no parameter is fitted.

full rationale

I walked the derivation chain from the linear-Gaussian model (Methods Eqs. 1-6) to the reported recoveries. The experimental records y_A(t) and y_B(t) are external data; the efficiencies eta_A, eta_B, angles theta_A, theta_B, and pump amplitude xi are independently characterized, not fitted to the purity/squeezing results. The theory curves in Figs. 4-6 are predictions from the Riccati equations, and the data agree without adjustment. The reconstructed covariances in Eq. (7) add the theoretically calculated V_T to an experimentally measured mean-square-error matrix; this is an application of the law of total covariance rather than a fit, and it does not force P(rho_S) > P(rho_F) because the experimental MSE term is free to disagree with the theory. The TrSD is additionally computed independently via Eq. (9), so the central comparison is not obtained only from the theoretical purity identity D = P(T) - P(C). The smoothing estimator is, by construction, the Bayesian estimate of the true state from past-future records, so its optimality is a theorem from Refs. [26,37]; the experimental contribution is to verify that the real OPO conforms to that theorem. Self-citations to the same authors' prior theory are present but are parameter-free mathematical results with stated assumptions, and this experiment provides external falsifiability; no uniqueness claim is imported to forbid alternatives. The acknowledged approximation that the 'true' state has eta_tot ~ 0.86 and the use of theoretical V_T in the headline absolute numbers introduce systematic uncertainty, which is a correctness and robustness caveat, not circularity. I found no step where a fitted parameter is renamed a prediction, and no equation reduces to its input by definition.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No free parameters were fitted to the target result; all experimental parameters (parametric gain, cavity linewidth, efficiencies, angles, beam-splitter transmittance) were independently measured or controlled. The central claim rests on standard linear Gaussian quantum optics assumptions and on the designed hidden-observer scenario, not on new postulated entities.

assumptions (4)
  • domain assumption The OPO with homodyne detection is a linear Gaussian quantum system, so all states are Gaussian and fully described by mean and covariance.
    Invoked in Results, Estimation of quantum dynamics: 'since the OPO with homodyne detection is a linear Gaussian quantum system, all states ... are fully described by the mean vector and the covariance matrix.'
  • standard math The quantum state smoothing equations from Laverick, Chantasri, Wiseman (Ref. 26) are correct for this system.
    The Methods section adopts the linear Gaussian smoothing equations (Eqs. 3-6) from the cited theoretical paper without re-deriving them.
  • domain assumption The hidden observer Bob's channel is the only information lost to Alice, and its unravelling is a homodyne measurement at angle theta_B.
    Central to quantum state smoothing; the experiment engineers this by splitting the OPO output and giving Bob a homodyne detector. Residual loss is treated as unobserved vacuum noise.
  • domain assumption The state conditioned on both Alice's and Bob's records is close enough to pure (eta_tot ~ 0.86) to be called the 'true' state.
    Stated in Results, Experimental setup: 'we are justified in using the term true for these states.' This is the benchmark used in TrSD and recovery metrics.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Tracking Quantum Dynamics in an Optical Cavity for Recovering Purity and Squeezing via Quantum State Smoothing." pith.science (2026). https://pith.science/paper/FTEW7GVR

@misc{pith2026250904754,
  author       = {Pith},
  title        = {Pith review of: Tracking Quantum Dynamics in an Optical Cavity for Recovering Purity and Squeezing via Quantum State Smoothing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FTEW7GVR}},
  note         = {Machine review of arXiv:2509.04754}
}
read the original abstract

Tracking the dynamics of a quantum system is conventionally achieved by monitoring the system continuously in time and filtering the information contained in measurement records via the causal quantum trajectory approach. However, in practical scenarios there is often loss of information to the environment, leading to filtered states that are impure because of decoherence. If real-time tracking is not required, the lost information can be maximally extracted via acausal quantum state smoothing, which has been theoretically proven to better restore the system's coherence (purity) than causal filtering. Interestingly, quantum state smoothing requires assumptions of how any lost quantum information (unobserved by the experimenter) was turned into classical information by the environment. In this work, we experimentally demonstrate smoothing scenarios, using an optical parametric oscillator and introducing `observed' and `unobserved' channels by splitting the output beam into two independent homodyne detectors. We achieve improvement in state purification of 10.3% +/- 1.6%, squeezing restoration of 7.6% +/- 2.6%, and show that smoothed states are better estimates of hidden true states than those from conventional filtering. The estimation techniques used in this paper are promising for many applications in quantum information that incorporate post-processing.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 43 canonical work pages

  1. [1]

    Nielsen, M. A. & Chuang, I. L.Quantum Computation and Quantum Information: 10th Anniversary Edition(Cambridge University Press, 2010)

    work page 2010

  2. [2]

    Wiseman, H. M. & Milburn, G. J.Quantum Measurement and Control(Cambridge University Press, 2009)

    work page 2009

  3. [3]

    W., Weber, S., Macklin, C

    Murch, K. W., Weber, S., Macklin, C. & Siddiqi, I. Observing single quantum trajectories of a superconducting quantum bit.Nature502, 211–214 (2013)

    work page 2013

  4. [4]

    & Schliesser, A

    Rossi, M., Mason, D., Chen, J. & Schliesser, A. Observing and verifying the quantum trajectory of a mechanical resonator.Physical Review Letters123, 163601 (2019)

    work page 2019

  5. [5]

    Weber, S.et al.Mapping the optimal route between two quantum states.Nature511, 570–573 (2014)

    work page 2014

  6. [6]

    E., Roch, N., Siddiqi, I

    Chantasri, A., Kimchi-Schwartz, M. E., Roch, N., Siddiqi, I. & Jordan, A. N. Quantum trajectories and their statistics for remotely entangled quantum bits.Physical Review X6, 041052 (2016). 11

    work page 2016

  7. [7]

    Naghiloo, M.et al.Quantum caustics in resonance-fluorescence trajectories.Physical Review A96, 053807 (2017)

    work page 2017

  8. [8]

    & Murch, K

    Foroozani, N., Naghiloo, M., Tan, D., Mølmer, K. & Murch, K. W. Correlations of the time dependent signal and the state of a continuously monitored quantum system.Physical Review Letters116, 110401 (2016)

    work page 2016

Show all 43 references
  1. [9]

    Tan, D.et al.Homodyne monitoring of postselected decay.Physical Review A96, 022104 (2017)

    2017

  2. [10]

    Wieczorek, W.et al.Optimal state estimation for cavity optomechanical systems.Physical Review Letters 114, 223601 (2015)

    2015

  3. [11]

    A., Bennett, J

    Meng, C., Brawley, G. A., Bennett, J. S., Vanner, M. R. & Bowen, W. P. Mechanical squeezing via fast continuous measurement.Physical Review Letters125, 043604 (2020)

    2020

  4. [12]

    Carmichael, H.An open systems approach to quantum optics(Springer, Berlin, 1993)

    1993

  5. [13]

    Measurement, filtering and control in quantum open dynamical systems.Reports on Mathematical Physics43, A405–A425 (1999)

    Belavkin, V. Measurement, filtering and control in quantum open dynamical systems.Reports on Mathematical Physics43, A405–A425 (1999)

    1999

  6. [14]

    Doherty, A. C. & Jacobs, K. Feedback control of quantum systems using continuous state estimation. Physical Review A60, 2700 (1999)

    1999

  7. [15]

    & Nori, F

    Zhang, J., Liu, Y.-x., Wu, R.-B., Jacobs, K. & Nori, F. Quantum feedback: Theory, experiments, and applications.Physics Reports679, 1–60 (2017)

    2017

  8. [16]

    Haykin, S.Kalman Filtering and Neural Networks(Wiley, New York, 2001)

    2001

  9. [17]

    L.Fixed Interval Smoothing for State Space Models(Kluwer Academic, New York, 2001)

    Weinert, H. L.Fixed Interval Smoothing for State Space Models(Kluwer Academic, New York, 2001)

    2001

  10. [18]

    Van Trees, H. L. & Bell, K. L.Detection, Estimation, and Modulation Theory, Part I: Detection, Estimation, and Filtering Theory, 2nd ed.(Wiley, 2013)

    2013

  11. [19]

    Brown, R. G. & Hwang, P. Y. C.Introduction to Random Signals and Applied Kalman Filtering4th edn (Wiley, 2012)

    2012

  12. [20]

    Einicke, G.Smoothing, Filtering and Prediction: Estimating the Past, Present and Future(InTech, 2012)

    2012

  13. [21]

    Friedland, B.Control System Design: An Introduction to State-Space Methods(Courier Corporation, New York, 2012)

    2012

  14. [22]

    Wheatley, T.et al.Adaptive optical phase estimation using time-symmetric quantum smoothing.Physical Review Letters104, 093601 (2010)

    2010

  15. [23]

    Yonezawa, H.et al.Quantum-enhanced optical-phase tracking.Science337, 1514–1517 (2012)

    2012

  16. [24]

    Iwasawa, K.et al.Quantum-limited mirror-motion estimation.Physical Review Letters111, 163602 (2013)

    2013

  17. [25]

    & Wiseman, H

    Guevara, I. & Wiseman, H. Quantum state smoothing.Physical Review Letters115, 180407 (2015)

    2015

  18. [26]

    T., Chantasri, A

    Laverick, K. T., Chantasri, A. & Wiseman, H. M. Quantum state smoothing for linear Gaussian systems. Physical Review Letters122, 190402 (2019)

    2019

  19. [27]

    T., Chantasri, A

    Laverick, K. T., Chantasri, A. & Wiseman, H. M. Linear Gaussian quantum state smoothing: Under- standing the optimal unravelings for Alice to estimate Bob’s state.Physical Review A103, 012213 (2021)

    2021

  20. [28]

    V., Guo, X., Breum, C

    Larsen, M. V., Guo, X., Breum, C. R., Neergaard-Nielsen, J. S. & Andersen, U. L. Deterministic multi- mode gates on a scalable photonic quantum computing platform.Nature Physics17, 1018–1023 (2021). 12

    2021

  21. [29]

    Asavanant, W.et al.Time-domain-multiplexed measurement-based quantum operations with 25-MHz clock frequency.Physical Review Applied16, 034005 (2021)

    2021

  22. [30]

    & Marandi, A

    Roy, A., Jahani, S., Langrock, C., Fejer, M. & Marandi, A. Spectral phase transitions in optical parametric oscillators.Nature Communications12, 835 (2021)

    2021

  23. [31]

    Aasi, J.et al.Enhanced sensitivity of the LIGO gravitational wave detector by using squeezed states of light.Nature Photonics7, 613–619 (2013)

    2013

  24. [32]

    Walk, N.et al.Experimental demonstration of Gaussian protocols for one-sided device-independent quantum key distribution.Optica3, 634–642 (2016)

    2016

  25. [33]

    Chen, M., Menicucci, N. C. & Pfister, O. Experimental realization of multipartite entanglement of 60 modes of a quantum optical frequency comb.Physical Review Letters112, 120505 (2014)

    2014

  26. [34]

    Weedbrook, C.et al.Gaussian quantum information.Reviews of Modern Physics84, 621–669 (2012)

    2012

  27. [35]

    & Gardiner, C

    Collett, M. & Gardiner, C. Squeezing of intracavity and traveling-wave light fields produced in parametric amplification.Physical Review A30, 1386 (1984)

    1984

  28. [36]

    Chantasri, A., Guevara, I., Laverick, K. T. & Wiseman, H. M. Unifying theory of quantum state estimation using past and future information.Physics Reports930, 1–40 (2021)

    2021

  29. [37]

    T., Guevara, I

    Laverick, K. T., Guevara, I. & Wiseman, H. M. Quantum state smoothing as an optimal Bayesian estimation problem with three different cost functions.Physical Review A104, 032213 (2021)

    2021

  30. [38]

    W., Mensen, L

    Walshe, B. W., Mensen, L. J., Baragiola, B. Q. & Menicucci, N. C. Robust fault tolerance for continuous- variable cluster states with excess antisqueezing.Physical Review A100, 010301 (2019)

    2019

  31. [39]

    & Leuchs, G

    Gl¨ ockl, O., Andersen, U., Filip, R., Bowen, W. & Leuchs, G. Squeezed-state purification with linear optics and feedforward.Physical Review Letters97, 053601 (2006)

    2006

  32. [40]

    Optics Express28, 30784–30796 (2020)

    Dirmeier, T.et al.Distillation of squeezing using an engineered pulsed parametric down-conversion source. Optics Express28, 30784–30796 (2020)

    2020

  33. [41]

    W.et al.Laser phase and frequency stabilization using an optical resonator.Applied Physics B31, 97–105 (1983)

    Drever, R. W.et al.Laser phase and frequency stabilization using an optical resonator.Applied Physics B31, 97–105 (1983)

    1983

  34. [42]

    & Zoller, P.Quantum noise: a handbook of Markovian and non-Markovian quantum stochastic methods with applications to quantum optics(Springer Science & Business Media, 2004)

    Gardiner, C. & Zoller, P.Quantum noise: a handbook of Markovian and non-Markovian quantum stochastic methods with applications to quantum optics(Springer Science & Business Media, 2004)

    2004

  35. [43]

    Alice" HD2

    Walls, D. F. & Milburn, G. J.Quantum Optics3rd edn (Springer, 2025). 13 Supplementary Information for Tracking Quantum Dynamics in an Optical Cavity for Recovering Purity and Squeezing via Quantum State Smoothing 1 Experimental details The schematic of our optical system and c...

    2025

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.