Pith. sign in

REVIEW 4 major objections 5 minor 62 references

Quantum Filtering and Stabilization of Dissipative Quantum Systems via Augmented Neural Ordinary Differential Equations

T0 review · 4 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read An augmented neural ODE can reconstruct a dissipative qubit's state and time-dependent decoherence rates from partial weak measurement data, and can drive real-time feedback control to a target state.

desk verdict A serious method proposal for learning open-qubit dynamics with augmented neural ODEs, but the headline 'parameter inference' claim leaks the true initial Δ(0), γ(0) into the encoder, so the quantum-observer claim is not actually demonstrated. read the letter →

arxiv 2509.07196 v2 pith:37I66E5U submitted 2025-09-08 quant-ph

Shahid Qamar , Rana Imran Mushtaq , Bo Li , Ho-Kin Tang This is my paper
classification quant-ph
keywords quantumfilteringneuralordinarydifferentialequationsopensystemsnon-Markoviandynamicsweakmeasurementfeedbackcontrolstateestimationdissipativequbit
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 an augmented quantum neural ordinary differential equation (AQNODE) that learns to act as a data-driven quantum observer. From initial conditions, the initial dissipation parameters, and a continuous weak-measurement record, it reconstructs the full Bloch-vector trajectory and the time-dependent diffusion and dissipation coefficients, without needing the explicit master equation at inference time. The paper demonstrates this on simulated non-Markovian qubit dynamics, reporting state and parameter errors below about 1e-3 within the training distribution and around 1e-2 out of distribution. It then shows that the learned state estimate can feed proportional-derivative and time-varying LQR controllers to steer the qubit to a target state, with LQR reaching fidelities near 0.93. If correct, the framework offers a path to quantum filtering and control when the Hamiltonian and noise model are only partially known.

What carries the argument

The central object is the augmented state Y_aug(t) = [x(t), y(t), z(t), Delta(t), gamma(t)]^T, which lets a neural ODE represent both the measured Bloch trajectory and unmeasured environmental parameters. The mechanism is a measurement-conditioned latent evolution: an encoder initializes the latent trajectory from partial data, the derivative of the latent state is learned as MLP_theta([h(t), dY(t)]), and a decoder maps the latent trajectory back to physical observables and parameters. This construction allows the network to integrate measurement information over time, acting as a learned filter rather than a fixed analytical model.

What would settle it

Feed the trained model a simulated measurement trace with random shot noise added to the detector output rather than the clean trace of Eq. (7), then compare the reconstructed Bloch state to the true state; if the error rises well above the reported MSE and does not converge over time, the claim that AQNODE filters real measurement data fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that a single learned augmented neural ODE can replace explicit physical equations for both state estimation and feedback control of a dissipative qubit. The model uses an augmented state Y_aug(t) = [x(t), y(t), z(t), Delta(t), gamma(t)] that includes both the observable Bloch components and the hidden time-dependent environmental parameters. Initial conditions and partial measurement outputs are encoded into a latent state, evolved by a neural ODE whose derivative function is an MLP conditioned on the measurement record, and decoded back into the physical variables. Trained on trajectories generated from a non-Markovian Lindblad master equation, the model recons

Load-bearing premise

The whole demonstration assumes the weak-measurement signal is a clean, deterministic function of the true state and that the training labels come from the same master equations that define the task, so real measurement noise or a different environment model could break it.

Editorial extensions

If this is right

  • A trained AQNODE can act as an observer for dissipative qubits when the Hamiltonian or Liouvillian is unknown: from partial measurement records it outputs both the Bloch state and the hidden time-dependent decoherence parameters.
  • The learned state estimate is usable for real-time closed-loop control; the paper's LQR implementation solves a differential Riccati equation using AQNODE predictions and achieves high-fidelity transfer to the target state even out of distribution.
  • Initial-state perturbations decay over time as the measurement record is integrated, indicating behavior consistent with quantum filtering and suggesting practical robustness to uncertain starting conditions.
  • The differentiable, adjoint-trained model can in principle be retrained or fine-tuned for new device parameters or multi-qubit systems whenever suitable trajectory data become available.
  • The latent representation separates trajectories by control strategy and distributional regime, suggesting the model captures physically meaningful hidden structure rather than memorizing individual trajectories.

Reading between the lines

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

  • The paper's validation uses a clean, deterministic measurement record; a natural stronger test is to feed the model stochastic homodyne trajectories with explicit shot noise and compare its reconstructions against a standard Bayesian quantum filter.
  • Because the ground-truth parameters are generated from one specific spectral-density model, a sharper test of the 'no explicit physical equations' claim would train on that model and test on a different environment model; success would indicate the latent dynamics capture generic dissipative structure.
  • The PD and LQR gains are fixed rather than globally optimized; an implicit next step is to use AQNODE's differentiability to optimize control fields end-to-end, which could close the small predicted-versus-real energy and fidelity gaps.
  • AQNODE is currently a point estimator; extending it to output a posterior distribution over states and parameters would connect it more directly to established quantum filtering theory.
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

4 major / 5 minor

Summary. The paper proposes an Augmented Quantum Neural ODE (AQNODE) framework for a single dissipative qubit. The model encodes weak-measurement traces and an initial augmented state into a latent trajectory, evolves it with a neural ODE, and decodes it into the Bloch vector components and time-dependent diffusion/dissipation parameters Δ(t) and γ(t). The authors report low MSE for state and parameter prediction in within-distribution (WD) and out-of-distribution (OOD) numerical tests, and they combine AQNODE predictions with PD and time-varying LQR controllers to steer the qubit to a target state. The abstract and conclusion claim that the method enables state reconstruction and parameter inference without explicit physical equations and that it functions as a data-driven quantum observer.

Significance. The architecture is a reasonable and potentially useful extension of latent Neural ODEs to open quantum systems, and the control comparison (PD vs LQR) is clearly presented. If the claims were validated under realistic stochastic weak measurements and with hidden initial parameters not provided to the model, AQNODE could be an attractive data-driven observer for quantum filtering and feedback control. As it stands, however, the evidence supports a narrower claim: accurate supervised trajectory fitting for a single-qubit Lindblad model when the true initial Δ(0), γ(0) are given and the measurement record is a noise-free deterministic function of the state. The current experiments do not demonstrate parameter inference from the measurement trace alone, nor do they test robustness to measurement noise or model misspecification.

major comments (4)
  1. [Sec. II, Eq. (7)] The measurement model is deterministic and noiseless: dY/dt = sqrt(M) ζ tr(-σ_z ρ_t) = -sqrt(M) ζ z(t). This is a continuous noise-free observation of the Bloch component z(t), not a stochastic weak-measurement record. Standard quantum filtering requires a stochastic master equation with innovations noise and measurement backaction. Feedthis deterministic trace as the network input means the claimed 'robust quantum filtering' and 'partial continuous measurement data' performance is untested for realistic noisy measurements. The authors should either reformulate Eq. (7) as a proper stochastic measurement model or explicitly restrict the claims to noiseless continuous observation.
  2. [Sec. III.A.1, III.A.2, IV.B] The parameter-inference claim is undermined by initialization leakage. The encoder is h(t0) = Encoderψ([Yaug(t0), dY(t)0:tk]) with Yaug(t0) = [x0,y0,z0,Δ(0),γ(0)], and each training sample includes the true initial augmented state. The Phase-2 perturbation study perturbs the full 5D vector, including Δ0 and γ0, so the model is always given the true (or perturbed-but-known) initial values of the hidden parameters. The low MSEs for Δ(t) and γ(t) in Tables I and II can therefore be explained by learning a mapping from these initial values to their future values, not by inferring them from the measurement trace. This directly contradicts the abstract's 'parameter inference without explicit physical equations.' The manuscript itself acknowledges in Sec. III.A.4 that environmental parameters 'cannot be directly measured and may require additional steps or calibration.' The model must be tested
  3. [Tables I, II; Sec. IV.A] Tables I and II report single MSE values, although the text states that the reported values are 'mean and standard deviation' over all test trajectories. No error bars, standard deviations, or number of test trajectories are given. Given the large OOD degradation in Phase 2 (e.g., MSE_y from 1.16×10^-3 to 5.80×10^-2), statistical quantification is essential for the generalization claim. In addition, the text in Sec. IV.A says 'the use of a physics-informed loss function ensured that the model respected the underlying physical laws,' but Eqs. (12)–(14) define the training loss as a weighted MSE only; no physics residual term is defined. Either define the physics-informed term or remove that claim.
  4. [Table III; Sec. IV.C] The control evaluation metrics in Table III are not clearly defined. The column 'MSE' is ambiguous: if it is the error between predicted and true controlled trajectories, then the rows 'Real PD' and 'Real LQR' should not have an MSE. 'Energy Dev' is not defined. The text also states that the predicted LQR maintains 'high fidelity (≥0.94) even under OOD conditions,' but Table III reports OOD Pred LQR fidelity as 0.932. These inconsistencies need to be resolved, and the exact formulas for MSE, final-state deviation, and fidelity should be given.
minor comments (5)
  1. [Sec. II, Eq. (5)] Typo in the first equation: '−(∆(t) + M/2)(t)' should likely be '−(∆(t) + M/2)x(t)'.
  2. [Sec. II, Eq. (7) and Sec. III.A.1] The notation is confusing: Eq. (7) defines dY/dt, but the encoder and neural ODE use dY(t) as an input. Please clarify whether the network receives the measurement trace Y(t), its derivative dY/dt, or both, and specify the sign convention tr(-σ_z ρ_t) = -z(t).
  3. [Sec. IV.C, target state] The target is described as the 'pure excited state' Y_target = [0,0,1], but Fig. 12 calls it the eigenstate |0>. In the standard convention |0> is the ground state. Please align the notation.
  4. [General] The term 'out-of-distribution' is used for parameter values drawn from wider intervals of the same parametric family (Eqs. 2-3). This is extrapolation within a family, not distribution shift to a different physical model. Consider using 'extrapolation' or 'wider-range' to avoid overstatement.
  5. [Appendix B, Fig. 17] The convergence shown under initial-state perturbations reflects the intrinsic stability of the Lindblad dynamics, not necessarily the filtering capability of AQNODE. This point should be stated explicitly so the perturbation experiment is interpreted correctly.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the AQNODE is a supervised surrogate with a standard train/test split; parameter predictions are not equivalent to inputs by construction.

full rationale

The paper's claimed derivation chain is an ML training procedure: simulate trajectories from Eq. (1) with Δ(t),γ(t) from Eqs. (2)-(3), train a latent neural ODE to map initial conditions and measurement traces to those trajectories, and evaluate MSE on unseen WD/OOD parameter draws. This is a conventional generalization claim, not a derivation of Δ(t),γ(t) from the same quantities. The apparent leakage of 'true hidden initial values' into the encoder (Sec. III.A.1: h(t0)=Encoderψ([Yaug(t0), dY(t)0:tk]) with Yaug(t0)=[x0,y0,z0,Δ(0),γ(0)]) is not circular: Eqs. (2)-(3) give Δ(0)=γ(0)=0 for every trajectory, so these inputs carry no information about the environmental parameters α,r,ω0; the actual information source is the measurement trace dY(t). The perturbation study (Sec. IV.B) perturbs Δ0,γ0 artificially, but this is a robustness test, not a claim that the true initial values are inferred. The paper itself flags the practical limitation that environmental parameters may need calibration (Sec. III.A.4), which undercuts the abstract's 'without explicit physical equations' phrasing, but this is an overstatement about experimental readiness, not a circular step. No load-bearing self-citation or imported uniqueness theorem is used; ref. [42] is a standard citation for the Lindblad form. The in-sample evaluation on the same generative family limits external validity, but it is not circularity.

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

All results are generated by simulating the same Lindblad equations that provide the supervision and evaluation. The method's claims of model-free inference and quantum filtering therefore rest on the assumption that the simulated model is an adequate stand-in for real experimental conditions, including the absence of measurement noise.

free parameters (4)
  • Loss weights κ and β = unspecified
    Equation (14) combines state and parameter losses; the weights are hyperparameters chosen without stated values or sensitivity analysis.
  • PD controller gains = kxp=5, kyp=10, kxd=8, kyd=10
    Section IV.C.2 sets these gains by hand; no tuning procedure is given.
  • LQR weights Q and R = Q=diag(1000,1000,1000), R=diag(0.1,50)
    Section IV.C.1 specifies these weights without justification.
  • Network architecture sizes = encoder/ODE/decoder MLP sizes not fully specified; ~130k parameters in phase 1, 51,976 in phase 3
    The paper gives neuron counts (6 input, 128 hidden, 5 output for one network) but not layer counts or activation functions.
assumptions (4)
  • domain assumption The Lindblad master equation (Eq. 1) is the exact generator of the qubit dynamics, with time-dependent Δ(t) and γ(t) given by Eqs. (2)-(3).
    The data and ground truth are generated from this model; if the real environment differs, the learned mapping may not transfer.
  • ad hoc to paper The weak measurement output dY(t)/dt = sqrt(M) ζ tr(-σz ρ_t) is a noise-free deterministic function of the state (Eq. 7).
    Real continuous weak measurements are stochastic; the paper does not include measurement noise, which is essential for genuine quantum filtering.
  • domain assumption The latent state h(t) ∈ R^d initialized by an encoder and evolved by an MLP parameterizes the full augmented state [x,y,z,Δ,γ] through a decoder.
    The representational capacity of the neural ODE is assumed sufficient to capture the true dynamics from the given inputs.
  • domain assumption The analytical forms of Δ(t) and γ(t) (Eqs. B5-B7) are the only source of non-Markovianity; training and test trajectories are drawn from this family.
    Generalization is only tested within this parametric family, so the model is not validated against dynamics outside the assumed noise model.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quantum Filtering and Stabilization of Dissipative Quantum Systems via Augmented Neural Ordinary Differential Equations." pith.science (2026). https://pith.science/paper/37I66E5U

@misc{pith2026250907196,
  author       = {Pith},
  title        = {Pith review of: Quantum Filtering and Stabilization of Dissipative Quantum Systems via Augmented Neural Ordinary Differential Equations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/37I66E5U}},
  note         = {Machine review of arXiv:2509.07196}
}
read the original abstract

Modeling open quantum dynamics without full knowledge of the system Hamiltonian or noise model is a key challenge in quantum control and quantum state estimation. We introduce an Augmented Quantum Neural Ordinary Differential Equation (AQNODE) framework that learns quantum trajectories and dissipation parameters directly from partial continuous measurement data. By embedding the system into a latent space evolved via neural ODEs, AQNODE captures both observable and hidden non-Markovian dynamics with temporal smoothness and physical consistency. Our approach integrates weak measurement data to reconstruct qubit states and time-dependent decoherence rates, enabling accurate state prediction and parameter inference without explicit physical equations. Furthermore, we incorporate AQNODE-based feedback control techniques, including proportional-derivative and time-varying linear-quadratic regulator (LQR) strategies, to steer the quantum system toward target states in real time. Extensive numerical simulations demonstrate AQNODE's ability to generalize across system configurations, achieve low prediction errors, and perform robust quantum filtering and control. These results establish AQNODE as a scalable, differentiable, and experimentally compatible framework for real-time modeling and control of dissipative quantum systems.

Figures

Figures reproduced from arXiv: 2509.07196 by the authors.

Figure 1
Figure 1. FIG. 1: Schematic framework of the Augmented Neural ODE framework for open quantum state and parame [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. This figure demonstrates the capability of [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 2
Figure 2. FIG. 2: Augmented AQNODE prediction and evolution of the qubit state in Phase 1 under WD (a–e) and OOD [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figures from the paper (14 more)
Figure 3
Figure 3. Figure 3: FIG. 3: AQNODE prediction of qubit states and system parameters in Phase 1 under WD (a–b) and OOD (c–d) [PITH_FULL_IMAGE:figures/full_fig_p008_3.png]
Figure 4
Figure 4. Figure 4: FIG. 4: Time evolution of average MSE during Phase 1 [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5: Phase 1: The distribution exhibits a pronounced [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6: AQNODE prediction and evolution of the qubit state in Phase 2 under varying non-Markovian and system [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7: AQNODE prediction of qubit states and parameters in Phase 2 under WD (a–b) and OOD (c–d) conditions. [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8: Phase 2: Average MSE over time. The distribu [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 10
Figure 10. Figure 10: FIG. 10: Histogram of MSE distribution for Within [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 9
Figure 9. Figure 9: FIG. 9: Average MSE curve over time during Phase 2, [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]
Figure 11
Figure 11. Figure 11: FIG. 11: AQNODE predictions under initial perturba [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: FIG. 12 [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]
Figure 13
Figure 13. Figure 13: FIG. 13: Time evolution of population probabilities [PITH_FULL_IMAGE:figures/full_fig_p013_13.png]
Figure 14
Figure 14. Figure 14: FIG. 14 [PITH_FULL_IMAGE:figures/full_fig_p014_14.png]
Figure 15
Figure 15. Figure 15: FIG. 15: Time evolution of population probabilities [PITH_FULL_IMAGE:figures/full_fig_p014_15.png]
Figure 17
Figure 17. Figure 17: FIG. 17: Comparison of perturbed and unperturbed tra [PITH_FULL_IMAGE:figures/full_fig_p018_17.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

62 extracted references · 47 canonical work pages

  1. [1]

    Latent Dynamics:Once initialized, the latent state evolves via a neural ODE

    Dynamics of Augmented System Since direct access toY aug(t) is generally unavail- able, we initialize a latent trajectoryh(t)∈R d 5 from partial data using an encoder network:h(t 0) = Encoderψ ([Yaug(t0), dY(t)0:tk ]), where Encoder ψ is a learnable neural network parameterized byψ. Latent Dynamics:Once initialized, the latent state evolves via a neural O...

    work page

  2. [2]

    1) over a time intervalt∈[0, T) and with the continuous weak measurementY(t)

    Data Generation The training dataset is generated by simulating the evolution of a single qubit system under a non-Markovian open quantum dynamics governed by a time-dependent Lindblad equation (Eq. 1) over a time intervalt∈[0, T) and with the continuous weak measurementY(t). For each trajectory, the dissipation strengthα, memory pa- rameterr, system-envi...

    work page

  3. [3]

    The latent space dimension determines the representation’s complexity, which allows the model to describe compli- cated system dynamics

    Learning Process An MLP is employed to represent the dynamical rule of the latent state in the neural ODE framework. The latent space dimension determines the representation’s complexity, which allows the model to describe compli- cated system dynamics. The training procedure of AQN- ODE is: Initialization: The initial encoder latent state (h 0) and the m...

    work page

  4. [4]

    A decoder is used to translate the enhanced latent representation of the sys- tem’s state into observable values after the neural ODE model has evolved it using an ODE function

    Training The training procedure for learning quantum dynam- ics using Neural ODEs involves simulating the dissipative Bloch equations to generate ground truth trajectories 6 and training a neural network to reconstruct these trajec- tories from weak measurement data. A decoder is used to translate the enhanced latent representation of the sys- tem’s state...

    work page

  5. [5]

    Validation After training, the model is evaluated on two differ- ent test datasets: within-distribution (WD) testing and out-of-distribution (OOD) testing and with different per- turbations in the initial qubit state. The WD test set consists of qubit trajectories generated using parameter ranges similar to the training set, while the OOD test set include...

    work page

  6. [6]

    The evolution of the Bloch vec- tor is governed by the ˆY(t) = [ˆx(t),ˆy(t),ˆz(t)]T , and the control fieldsu x(t), uy(t) are applied via the Hamiltonian operatorsA x andA y

    LQR Control To achieve optimal control of the quantum state, LQR control dynamically adjusts feedback gains based on a cost function [49, 52]. The evolution of the Bloch vec- tor is governed by the ˆY(t) = [ˆx(t),ˆy(t),ˆz(t)]T , and the control fieldsu x(t), uy(t) are applied via the Hamiltonian operatorsA x andA y. Thus, the control matrixB(t) is constru...

    work page

  7. [7]

    Phase 3: PD control The PD control strategy computes the control fields based on the error ( ˆY(t)− Ytarget) and its derivative ˙ˆY(t) [53, 54]. Then the control fields are calculated as ux =−k x p ·e x(t)−k x d · ˙ˆx(t) (20) uy =−k y p ·e y(t)−k y d · ˙ˆy(t) (21) The errorse x(t) = ˆx(t)−x target(t) ande y(t) = ˆy(t)− ytarget(t) between the current estim...

    work page

  8. [8]

    Preliminaries of state transfer Coherence and purity are essential markers of the dy- namics of quantum systems. Pure (p(t) = 1) and mixed (p(t)<1) quantum states are distinguished by purity, i.e.,p(t) = 1+x2(t)+y2(t)+z2(t) 2 , which reflects the degree of decoherence or dissipation in the system. Its temporal evolution shows how energy splitting (∆(t)) a...

    work page

Show all 62 references
  1. [9]

    Gebhart, R

    V. Gebhart, R. Santagati, A. A. Gentile, E. M. Gauger, D. Craig, N. Ares, L. Banchi, F. Marquardt, L. Pezze, and C. Bonato, Learning quantum systems, Nature Re- views Physics5, 141 (2023)

    2023

  2. [10]

    Havl ´ ıˇ cek, A

    V. Havl ´ ıˇ cek, A. D. C´ orcoles, K. Temme, A. W. Harrow, A. Kandala, J. M. Chow, and J. M. Gambetta, Super- vised learning with quantum-enhanced feature spaces, Nature567, 209 (2019)

    2019

  3. [11]

    I. R. Petersen and D. Dong, Special section on estimation and control of quantum systems (2022)

    2022

  4. [12]

    Mart ´ ınez-Vargas, C

    E. Mart ´ ınez-Vargas, C. Pineda, F. Leyvraz, and P. Barberis-Blostein, Quantum estimation of unknown parameters, Physical Review A95, 012136 (2017)

    2017

  5. [13]

    Cortez, A

    L. Cortez, A. Chantasri, L. P. Garc ´ ıa-Pintos, J. Dressel, and A. N. Jordan, Rapid estimation of drifting parame- ters in continuously measured quantum systems, Physical Review A95, 012314 (2017)

    2017

  6. [14]

    H. Liao, D. S. Wang, I. Sitdikov, C. Salcedo, A. Seif, and Z. K. Minev, Machine learning for practical quantum error mitigation, Nature Machine Intelligence6, 1478 (2024)

    2024

  7. [15]

    J. B. Rigo and A. K. Mitchell, Machine learning effective models for quantum systems, Physical Review B101, 241105 (2020)

    2020

  8. [16]

    LeCun, Y

    Y. LeCun, Y. Bengio, and G. Hinton, Deep learning, Na- ture521, 436 (2015)

    2015

  9. [17]

    Huang, R

    H.-Y. Huang, R. Kueng, and J. Preskill, Predicting many properties of a quantum system from very few measure- ments, Nature Physics16, 1050 (2020)

    2020

  10. [18]

    H. Ma, B. Qi, I. R. Petersen, R.-B. Wu, H. Rabitz, and D. Dong, Machine learning for estimation and control of quantum systems, National Science Review , nwaf269 (2025)

    2025

  11. [19]

    M. Y. Niu, S. Boixo, V. N. Smelyanskiy, and H. Neven, Universal quantum control through deep reinforcement learning, npj Quantum Information5, 33 (2019)

    2019

  12. [20]

    Preskill, Quantum computing in the nisq era and be- yond, Quantum2, 79 (2018)

    J. Preskill, Quantum computing in the nisq era and be- yond, Quantum2, 79 (2018)

    2018

  13. [21]

    Callison and N

    A. Callison and N. Chancellor, Hybrid quantum-classical algorithms in the noisy intermediate-scale quantum era and beyond, Physical Review A106, 010101 (2022)

    2022

  14. [22]

    Bharti, A

    K. Bharti, A. Cervera-Lierta, T. H. Kyaw, T. Haug, S. Alperin-Lea, A. Anand, M. Degroote, H. Heimonen, J. S. Kottmann, T. Menke,et al., Noisy intermediate- scale quantum algorithms, Reviews of Modern Physics 94, 015004 (2022)

    2022

  15. [23]

    H. M. Wiseman and G. J. Milburn,Quantum measure- ment and control(Cambridge university press, 2009)

    2009

  16. [24]

    M. A. Nielsen and I. L. Chuang,Quantum computation and quantum information(Cambridge university press, 2010)

    2010

  17. [25]

    Domingo, G

    L. Domingo, G. Carlo, and F. Borondo, Optimal quan- tum reservoir computing for the noisy intermediate-scale quantum era, Physical Review E106, L043301 (2022)

    2022

  18. [26]

    M. J. Ranˇ ci´ c, Noisy intermediate-scale quantum comput- ing algorithm for solving an n-vertex maxcut problem with log (n) qubits, Physical Review Research5, L012021 (2023)

    2023

  19. [27]

    H. M. Wiseman and G. J. Milburn, Quantum theory of optical feedback via homodyne detection, Physical Re- view Letters70, 548 (1993)

    1993

  20. [28]

    T. H. Kyaw, T. Menke, S. Sim, A. Anand, N. P. Sawaya, W. D. Oliver, G. G. Guerreschi, and A. Aspuru-Guzik, Quantum computer-aided design: digital quantum sim- ulation of quantum processors, Physical Review Applied 16, 044042 (2021)

    2021

  21. [29]

    Iyer and D

    P. Iyer and D. Poulin, A small quantum computer is needed to optimize fault-tolerant protocols, Quantum Science and Technology3, 030504 (2018)

    2018

  22. [30]

    V. M. Bastidas, T. H. Kyaw, J. Tangpanitanon, G. Romero, L.-C. Kwek, and D. G. Angelakis, Flo- quet stroboscopic divisibility in non-markovian dynam- ics, New Journal of Physics20, 093004 (2018)

    2018

  23. [31]

    T. H. Kyaw, D. A. Herrera-Marti, E. Solano, G. Romero, and L.-C. Kwek, Creation of quantum error correcting codes in the ultrastrong coupling regime, Physical Re- view B91, 064503 (2015)

    2015

  24. [32]

    K. R. Brown, J. Kim, and C. Monroe, Co-designing a scalable quantum computer with trapped atomic ions, npj Quantum Information2, 1 (2016)

    2016

  25. [33]

    De Vega and D

    I. De Vega and D. Alonso, Dynamics of non-markovian open quantum systems, Reviews of Modern Physics89, 015001 (2017)

    2017

  26. [34]

    Flannigan, F

    S. Flannigan, F. Damanet, and A. J. Daley, Many-body quantum state diffusion for non-markovian dynamics in strongly interacting systems, Physical Review Letters 128, 063601 (2022)

    2022

  27. [35]

    V. R. Overbeck and H. Weimer, Time evolution of open quantum many-body systems, Physical Review A93, 012106 (2016)

    2016

  28. [36]

    Biamonte, P

    J. Biamonte, P. Wittek, N. Pancotti, P. Rebentrost, N. Wiebe, and S. Lloyd, Quantum machine learning, Na- ture549, 195 (2017)

    2017

  29. [37]

    R. Iten, T. Metger, H. Wilming, L. Del Rio, and R. Ren- ner, Discovering physical concepts with neural networks, Physical Review Letters124, 010508 (2020)

    2020

  30. [38]

    Flam-Shepherd, T

    D. Flam-Shepherd, T. C. Wu, X. Gu, A. Cervera-Lierta, M. Krenn, and A. Aspuru-Guzik, Learning interpretable representations of entanglement in quantum optics ex- periments using deep generative models, Nature Machine Intelligence4, 544 (2022)

    2022

  31. [39]

    R. T. Chen, Y. Rubanova, J. Bettencourt, and D. K. Du- venaud, Neural ordinary differential equations, Advances in Neural Information Processing Systems31(2018). 20

    2018

  32. [40]

    Dupont, A

    E. Dupont, A. Doucet, and Y. W. Teh, Augmented neu- ral odes, Advances in Neural Information Processing Sys- tems32(2019)

    2019

  33. [41]

    M. Choi, D. Flam-Shepherd, T. H. Kyaw, and A. Aspuru- Guzik, Learning quantum dynamics with latent neural ordinary differential equations, Physical Review A105, 042403 (2022)

    2022

  34. [42]

    Chen and Y

    L. Chen and Y. Wu, Learning quantum dissipation by the neural ordinary differential equation, Physical Review A 106, 022201 (2022)

    2022

  35. [43]

    J. Li, X. Yang, X. Peng, and C.-P. Sun, Hybrid quantum- classical approach to quantum optimal control, Physical Review Letters118, 150503 (2017)

    2017

  36. [44]

    S. C. Hou, M. A. Khan, X. X. Yi, D. Dong, and I. R. Petersen, Optimal lyapunov-based quantum control for quantum systems, Physical Review A86, 022321 (2012)

    2012

  37. [45]

    Campagne-Ibarcq, E

    P. Campagne-Ibarcq, E. Flurin, N. Roch, D. Darson, P. Morfin, M. Mirrahimi, M. H. Devoret, F. Mallet, and B. Huard, Persistent control of a superconducting qubit by stroboscopic measurement feedback, Physical Review X3, 021008 (2013)

    2013

  38. [46]

    Lloyd and J.-J

    S. Lloyd and J.-J. E. Slotine, Quantum feedback with weak measurements, Physical Review A62, 012307 (2000)

    2000

  39. [47]

    Zhang, Y

    J. Zhang, Y. xi Liu, R.-B. Wu, K. Jacobs, and F. Nori, Quantum feedback: Theory, experiments, and applica- tions, Physics Reports679, 1 (2017), quantum feedback: theory, experiments, and applications

    2017

  40. [48]

    Yang and Z

    L. Yang and Z. Shao, Neural controlled differen- tial equations with quantum hidden evolutions (2024), arXiv:2404.19673 [cs.LG]

    2024 arXiv

  41. [49]

    W. Cui, Z. R. Xi, and Y. Pan, Optimal decoherence control in non-markovian open dissipative quantum sys- tems, Physical Review A—Atomic, Molecular, and Opti- cal Physics77, 032117 (2008)

    2008

  42. [50]

    Qamar, S

    S. Qamar, S. Cong, K. Li, Z. Dong, and F. Shuang, Non- linear optimal feedback control of the two-level open non- markovian stochastic quantum system, Open Systems & Information Dynamics28, 2150013 (2021)

    2021

  43. [51]

    Breuer, E.-M

    H.-P. Breuer, E.-M. Laine, J. Piilo, and B. Vacchini, Col- loquium: Non-markovian dynamics in open quantum sys- tems, Reviews of Modern Physics88, 021002 (2016)

    2016

  44. [52]

    Maniscalco, S

    S. Maniscalco, S. Olivares, and M. G. Paris, Entan- glement oscillations in non-markovian quantum chan- nels, Physical Review A—Atomic, Molecular, and Op- tical Physics75, 062119 (2007)

    2007

  45. [53]

    Gardiner and P

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

    2004

  46. [54]

    G. E. Fux, E. P. Butler, P. R. Eastham, B. W. Lovett, and J. Keeling, Efficient exploration of hamiltonian pa- rameter space for optimal control of non-markovian open quantum systems, Physical Review Letters126, 200401 (2021)

    2021

  47. [55]

    Hwang and H.-S

    B. Hwang and H.-S. Goan, Optimal control for non- markovian open quantum systems, Physical Review A—Atomic, Molecular, and Optical Physics85, 032321 (2012)

    2012

  48. [56]

    Wu and W

    S. Wu and W. Ma, Trajectory tracking for non-markovian quantum systems, Physical Review A105, 012204 (2022)

    2022

  49. [57]

    S. C. Edwards and V. P. Belavkin, Optimal quantum filtering and quantum feedback control, arXiv preprint quant-ph/0506018 (2005)

    2005 arXiv

  50. [58]

    Langen, G

    T. Langen, G. Valtolina, D. Wang, and J. Ye, Quantum state manipulation and cooling of ultracold molecules, Nature Physics20, 702 (2024)

    2024

  51. [59]

    Ryan-Anderson, J

    C. Ryan-Anderson, J. G. Bohnet, K. Lee, D. Gresh, A. Hankin, J. Gaebler, D. Francois, A. Chernoguzov, D. Lucchetti, N. C. Brown,et al., Realization of real-time fault-tolerant quantum error correction, Physical Review X11, 041058 (2021)

    2021

  52. [60]

    A. C. Doherty and K. Jacobs, Feedback control of quan- tum systems using continuous state estimation, Physical Review A60, 2700 (1999)

    1999

  53. [61]

    H. Chen, H. Li, F. Motzoi, L. Martin, K. B. Whaley, and M. Sarovar, Quantum proportional-integral (pi) control, New Journal of Physics22, 113014 (2020)

    2020

  54. [62]

    J. E. Gough, A quantum kalman filter-based pid con- troller, Open Systems & Information Dynamics27, 2050014 (2020)

    2020

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.