Pith. sign in

REVIEW 3 major objections 4 minor 102 references

Reinforcement learning to learn quantum states for Heisenberg scaling accuracy

T0 review · 3 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A reinforcement-learning agent that tunes an optimizer's step size and learning rate can learn random quantum states with infidelity that falls as one over the total success count, and the three-qubit policy also works on five-qubit states.

desk verdict Useful RL meta-learning for quantum state learning, but the Heisenberg-scaling claim is overstated because the shot accounting omits failure shots and the exponent is largely built into the halting rule. read the letter →

arxiv 2412.02334 v2 pith:7QUEUUSX submitted 2024-12-03 quant-ph cs.AI

Jeongwoo Jae , Jeonghoon Hong , Jinho Choo , Yeong-Dae Kwon This is my paper · ORCID
classification quant-phcs.AI PACS 03.67.-a
keywords reinforcementlearningquantumstatemeta-learningevolutionstrategyHeisenbergscalingsingle-shotmeasurementhardware-efficientansatzactionrepetition
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 a meta-learning (learning-to-learn) scheme in which a reinforcement-learning (RL) agent learns to control an optimizer that trains a quantum circuit to reproduce an unknown state. The optimizer is an evolution strategy, and the agent's only observation is the number of consecutive success outcomes obtained before a failure when the circuit is measured in a fixed basis. The agent chooses the optimizer's sampling range and learning rate, and is rewarded for reaching a target success count quickly, so maximizing reward is the same as minimizing the total success count used. In simulated learning of 1-, 2-, and 3-qubit Haar-random (uniformly random) pure states, the trained agent lowers the average total success count by 14%, 28%, and 34% relative to the best fixed hyperparameters, and the paper reports an infidelity scaling $\bar f \sim O(C_{\mathrm{total}}^{-1})$ with fitted exponents from about 0.95 to 1.18. The same agent trained on three-qubit states is then applied to four- and five-qubit states, a test of whether the learned hyperparameter policy generalizes beyond its training dimension.

What carries the argument

The load-bearing machinery is the evolution-strategy update loop whose two hyperparameters are controlled by the RL agent. At each iteration, the ES samples $k$ parameter perturbations $\vec\theta + \sigma\vec\epsilon_i$ around the current ansatz parameters, measures the success count for each perturbed state, and updates $\vec\theta$ by a reparameterized gradient estimator. The observation fed to the agent is the geometric success count $C$ from the single-shot measurement, which ties the objective directly to measurement statistics rather than to expensive fidelity estimation. The action repetition strategy is the training device that makes RL tractable: repeating a chosen $(\sigma,\eta)$ for $t_{\mathrm{rep}}$ consecutive steps shortens the effective decision horizon early in training, producing a curriculum that lets the actor-critic learn to reduce the halting time.

What would settle it

Record every measurement outcome during a run, set the true shot budget to $S = C_{\mathrm{total}} + (\text{number of measurement sequences})$, and refit $\bar f = \alpha S^{-\beta}$ for 1-, 2-, and 3-qubit states; if the 2- and 3-qubit exponents fall below 1, or the 1-qubit exponent falls clearly below its reported 0.948, the Heisenberg-scaling claim in terms of actual shots does not hold.

Watch

Extended reading notes

Core claim

The central claim is that the meta-learned RL policy makes quantum state learning shot-efficient at the statistical limit. The environment is a single-shot measurement scheme: the hardware-efficient ansatz $\hat{U}(\vec\theta)$ is applied to the unknown state, a binary measurement is repeated until a failure outcome occurs, and the number of consecutive successes $C$ is recorded; because $C$ is geometrically distributed with mean $p_s/(1-p_s)$, maximizing $C$ aligns the ansatz with the success basis. The RL agent, an actor-critic pair, observes $C$ and selects the evolution-strategy hyperparameters $\sigma$ (sampling range) and $\eta$ (learning rate), receiving reward $-1$ for each time step before the halting step. With the action repetition strategy, each chosen action is held for $t_{\mathrm{rep}}$ steps, and $t_{\mathrm{rep}}$ is annealed downward during RL training as a curriculum on the depth of the decision process. After training at target success count $C_{\mathrm{target}}=10^4$, the policy reduces average total success counts by 14% (1 qubit), 28% (2 qubit), and 34% (3 qubit) relative to the best fixed action, and the fitted infidelity scaling is $\bar f = \alpha C_{\mathrm{total}}^{-\beta}$ with $\beta \approx 0.948, 1.161, 1.184$; the 3-qubit policy transfers to 4- and 5-qubit states with $\beta \approx -1.189$ and $-0.829$.

Load-bearing premise

The paper equates total success count with total measurement shots, even though each measurement sequence ends in an uncounted failure shot; the scaling claim depends on those failure shots being negligible or scaling at the same rate.

Editorial extensions

If this is right

  • If the scaling claim holds, reaching infidelity $\bar f$ costs $O(1/\bar f)$ total success counts, matching the Heisenberg limit and beating the $O(\bar f^{-4/3})$ scaling of standard quantum state tomography.
  • At target success count $10^4$, the RL-trained strategy saves 14%, 28%, and 34% of total success counts for 1-, 2-, and 3-qubit states relative to the best fixed hyperparameters, with average fidelities of about 0.99994, 0.99972, and 0.99946.
  • The three-qubit-trained policy generalizes to four- and five-qubit Haar-random states, achieving average infidelity around $6.6\times10^{-4}$ and $9.3\times10^{-4}$ with $7.3\times10^6$ and $1.29\times10^8$ success counts respectively.
  • The same policy learns a particular entangled five-qubit state to fidelity about 0.9989 with $6.62\times10^7$ success counts, indicating that the learned hyperparameter schedule is not tied to the Haar-random training distribution alone.
  • The hardware-efficient ansatz used here has fewer parameters than the general unitary for the same system size, and the ES uses fewer evaluation samples than parameters, so the optimizer-ansatz combination remains practical when gradients are expensive.

Reading between the lines

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

  • Editorial inference: the reported resource is total success count, not total measurement shots; every measurement sequence also ends in a failure shot, so the true shot budget is larger, and refitting exponents against the full shot count is the natural check on the Heisenberg claim.
  • Editorial inference: the learned policy seems to implement a simple 'large step size when C is small, small when C is large' schedule; if so, a closed-form adaptive schedule might capture much of the benefit, which could be tested by replacing the RL policy with a hand-coded rule.
  • Editorial inference: because the observation is a scalar success count, the same framework could be applied to other success/failure objectives such as gate fidelity or variational eigensolver energies; the open question is whether the scalar observation remains sufficient for those landscapes.
  • Editorial inference: the jump from 3-qubit training to 5-qubit states suggests the agent may have learned dimensionless features of the optimization landscape; simulating 6- to 8-qubit states would show whether the scaling exponents stay near 1 or degrade.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper proposes a reinforcement-learning-based meta-learning scheme that adapts the hyperparameters (sampling range sigma and learning rate eta) of an evolution strategy used to train a hardware-efficient ansatz for learning unknown pure quantum states via single-shot measurements. The authors report that the trained RL agent reduces the total success count needed to reach a target infidelity compared with fixed-hyperparameter baselines, achieves infidelity scaling close to the Heisenberg limit, and generalizes from 3-qubit training to 4- and 5-qubit state learning. They also introduce an action repetition strategy to make RL training tractable.

Significance. If the scaling claim were fully supported, the work would be a useful demonstration that meta-learned hyperparameter schedules can improve variational quantum state learning and that a hardware-efficient ansatz can be trained with fewer evaluation points than parameters. The manuscript is commendable for releasing code, for presenting the action repetition strategy clearly, and for comparing against maximum-likelihood QST. However, the central resource accounting and the interpretation of the scaling exponents need additional work before the Heisenberg-scaling claim can be accepted.

major comments (3)
  1. [Algorithm 1, line 15 and Fig. 4] The quantity plotted on the x-axis of Fig. 4 is the total success count C_total, not the total number of measurement shots. In each iteration of Algorithm 1, line 3 measures one sequence and line 8 measures k sequences, and every sequence terminates with a failure outcome. The true shot count is therefore C_total + (k+1)*t_H, where t_H is the number of ES iterations. The paper never reports t_H for the scaling runs. For the training runs at C_target=10^4, the reported t_H values (19, 660, 3409 for 1-, 2-, and 3-qubit states) with k=5, 10, 30 give omitted failure-shot offsets of 114, 7,260, and 105,679, respectively, which are up to about 4.5% of C_total and grow relative to C_total as C_target decreases. For the 5-qubit generalization (k=100, C_total≈1.29e8), t_H is not reported and the omitted failure shots could be comparable to C_total. Since the Heisenberg claim is a fitted exponent on a log-log plot, the authors must report t_H (or the total number of sequences) for every data point and refit the infidelity against C_true = C_total + (k+1)*t_H.
  2. [§II.A, Eqs. (2)-(4) and §III.C] The near-Heisenberg exponent is largely constrained by the halting rule, so the present data do not demonstrate that the RL agent discovers Heisenberg scaling. Because the infidelity is f = 1 - p_s for a pure state and Algorithm 1 halts when an observed success count satisfies C >= C_target (line 9), the final fidelity is forced to be of order 1/C_target up to geometric fluctuations. The fitted exponent in f = α C_total^{-β} therefore mostly measures how C_total scales with C_target, not an independent scaling law. To support the advertised claim, the authors should report C_total versus C_target for the RL and baseline runs, report the actual final p_s values, and show that the fitted exponent is robust when the x-axis is the true shot count rather than C_total.
  3. [§III.C, scaling exponents] The reported exponents β ≈ 1.161 and 1.184 for 2- and 3-qubit states exceed the range β ∈ [0.9, 1.0) that the paper itself cites for Heisenberg-limited one-qubit learning, and they also exceed the values one would expect if β=1 were the statistical limit. This inconsistency is likely a consequence of the missing failure shots in the resource accounting, but as written it makes the scaling claim internally questionable. The fits also use only four points per curve with no error bars. The authors should refit with the corrected shot count and provide confidence intervals for β.
minor comments (4)
  1. [§III.D] The text states that the scaling factor for 4- and 5-qubit generalization is “β = −1.189 and β = −0.829”; negative exponents contradict the fitted form f = α C^{-β} and the plotted decreasing trend. These should be positive values, or the sign convention should be clarified.
  2. [§II.B and Table I] The action space is defined as A = Aσ × Aη with Aσ, Aη ⊂ [0, ∞), but Table I specifies small discrete sets (e.g., 4 values each). Please state explicitly that the continuous spaces are discretized according to Table I and whether the 169-action case uses the geometric spacing of Eq. (10).
  3. [Appendix D and Fig. 4] The baseline selection uses “the action which gives the lowest total success count” from a grid of simulations; this is a post-selected baseline. Please state whether the same post-selection was used when computing the improvement percentages in Fig. 4(d), and clarify how simulations that fail within t_max are excluded from the averages.
  4. [Table I, footnote a] There is a typo in “We obtain similiar results for tl = 200”; it should read “similar.”

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the RL contribution is validated against fixed-action baselines, and the Heisenberg-scaling claim is transparently inherited from the cited SSML protocol.

full rationale

The paper proposes an RL agent that tunes the hyperparameters (sampling range and learning rate) of an evolution strategy used to train a hardware-efficient ansatz for quantum state learning. The reported improvement in shot efficiency is an empirical comparison against fixed-action baselines (Appendix D) chosen as the best grid points without RL; the RL agent is trained to minimize halting time t_H, not to match a pre-fitted target, so this comparison is not circular. The infidelity scaling f ~ O(C^{-1}) is presented as a simulation result obtained by curve-fitting the paper's own data. Although the geometric measurement model (Eq. 2), the reconstruction rule (Eq. 3), and the infidelity definition (Eq. 4) imply that, once a run of length C_target is observed, the success probability must be close to 1 and hence f ~ O(1/C_target), the paper explicitly attributes this single-shot measurement scheme to prior SSML work (Refs. [25,26]) and makes no claim to have derived this scaling from first principles. The actual novelty is the RL-based hyperparameter adaptation and the action repetition strategy, which are evaluated against baselines and shown to reduce total success counts. There is no load-bearing self-citation chain, no uniqueness argument imported from the authors' own prior work, and no fitted parameter relabeled as a prediction. The separate concern that C_total excludes the failure shot terminating each measurement sequence is a correctness and resource-accounting issue, not a circularity, and does not affect the derivation chain itself. Therefore no circular step is identified.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

No new physical entities are introduced; the RL agent, ES, and HEA are standard tools. The load-bearing choices are the hand-selected hyperparameters (action grids, k, curriculum schedule, t_max, HEA depth) and the resource definition that equates C_total with total measurement shots. The power-law exponents are fitted from the paper's own simulations rather than derived.

free parameters (6)
  • ES sample count k = k = 5, 10, 30, 100 for N = 1, 2, 3, 4, 5
    Chosen per qubit count in Table I; sets the gradient estimator quality and directly affects total measurement cost.
  • Action grid A_sigma x A_eta = 16 actions for N=1,2,3; 169 actions for one 2-qubit run
    Hand-selected ranges in Table I and Eq. (10); the RL agent can only choose from this discrete set, so performance depends on this choice.
  • Action repetition schedule (t_u, t_l, T_th) = (1,50,100), (80,800,100), (300,2000,500)
    Eq. (11) and Table I; the curriculum schedule is chosen manually per system and shapes the RL training difficulty.
  • Maximum ES steps t_max used in reward normalization = 3e3, 1e4, 2e4 for 1, 2, 3 qubits
    Appendix C Eq. (A1) divides cumulative reward by t_max; t_max is empirically chosen and affects the critic target and thus the learned policy.
  • Power-law scaling fit parameters (alpha, beta) = beta ~ 0.948, 1.161, 1.184 for 1, 2, 3 qubits; negative sign reported for 4/5 qubits
    The Heisenberg-scaling claim rests on curve fitting f = alpha C^{-beta} to only four target values in Fig. 4; the exponents are fitted outputs, not derived predictions.
  • HEA depth L = L = 1, 5, 10 for N = 2, 3, 4/5
    Chosen to make the HEA expressive enough to represent random states; the representational capacity is central to the fidelity results.
assumptions (5)
  • standard math Success count C is geometrically distributed with p(C) = p_s^C (1 - p_s).
    Eq. (2) assumes independent identical Bernoulli trials with success probability p_s, standard for repeated projective measurements until failure.
  • domain assumption The reconstructed state fidelity equals p_s.
    From Eq. (3), |psi_est> = U^dagger(theta_train)|s>, so |<psi|psi_est>|^2 = p_s; this identifies infidelity with 1 - p_s and connects the stopping rule to accuracy.
  • domain assumption The ES gradient estimator in Eq. (7) is useful with fewer samples than HEA parameters.
    The paper uses k = 10 for 12 parameters and k = 30 for 54 parameters; no convergence proof is given, and barren-plateau concerns are only cited, not resolved.
  • domain assumption Haar-random pure states are representative target states for the scaling claim.
    Random states are generated by Haar-distributed unitaries; the claim is about average performance over this ensemble, not worst-case or structured states.
  • domain assumption The HEA with the chosen depth can represent the target random states to the reported infidelity.
    The paper asserts expressivity based on parameter counts and empirical success; there is no rigorous guarantee for 4 and 5 qubits at f ~ 10^-4.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Reinforcement learning to learn quantum states for Heisenberg scaling accuracy." pith.science (2026). https://pith.science/paper/7QUEUUSX

@misc{pith2026241202334,
  author       = {Pith},
  title        = {Pith review of: Reinforcement learning to learn quantum states for Heisenberg scaling accuracy},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7QUEUUSX}},
  note         = {Machine review of arXiv:2412.02334}
}
read the original abstract

Learning quantum states is a crucial task for realizing quantum information technology. Recently, neural approaches have emerged as promising methods for learning quantum states. We propose a meta-learning model that utilizes reinforcement learning (RL) to optimize the process of learning quantum states. To improve the data efficiency of the RL, we introduce an action repetition strategy inspired by curriculum learning. The RL agent significantly improves the sample efficiency of learning random quantum states, and achieves infidelity scaling close to the Heisenberg limit. We also show that the RL agent trained using 3-qubit states can generalize to learning up to 5-qubit states. These results highlight the utility of RL-driven meta-learning to enhance the efficiency and generalizability of learning quantum states. Our approach can be applied to improve quantum control, quantum optimization, and quantum machine learning.

Figures

Figures reproduced from arXiv: 2412.02334 by the authors.

Figure 1
Figure 1. FIG. 1. Schematic of reinforcement learning to learn quantum states. (a) In the environment, the evolution strategy trains a [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Results of training RL agent for the target success count [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. (a) Scheme of action repetition strategy (ARS). Without the ARS ( [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Learning random pure states with the trained RL agent. We consider the target success counts from 10 [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Learning an entangled 5-qubit pure state ˆϱ [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Comparison between our meta-learning method and the quantum state tomography (QST). The results of learning a [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

102 extracted references · 57 canonical work pages

  1. [1]

    The state st is a quantum state ˆU (⃗θt) |ψ⟩ transformed by the HEA in Fig

    S is the Hilbert space of N -qubit states. The state st is a quantum state ˆU (⃗θt) |ψ⟩ transformed by the HEA in Fig. 1 (b)

  2. [2]

    A is the space of action at = (σt, ηt), where σ and η are hyperprameters of evolution strategy which de- termine sampling range and learning rate, respec- tively [43]

  3. [3]

    T is the space of transition probability defined by p(st+1|st, at) : S × A × S →[0, 1]

  4. [4]

    The reward rt = 0 if the quantum state learning is finished at time step t

    R is the space of reward rt. The reward rt = 0 if the quantum state learning is finished at time step t. Otherwise, rt = −1

  5. [5]

    To observe the state, we use a measurement of binary outcome; success and fail

    Ω is the space of observation ot. To observe the state, we use a measurement of binary outcome; success and fail. We perform the measurement until a fail outcome appears. The observation is defined as the number of consecutive success outcomes be- fore the fail outcome

  6. [6]

    O is the space of the transition probability of ob- servation defined by p(ot|st) : O × S →[0, 1]. 11 0.0010.010.11 0.0017.0027.0067.0067.006 0.017.0607.0057.0087.008 0.15.7336.4117.0507.014 17.0927.1097.0847.046 0.0010.010.11 0.001-0.014-0.035-0.038-0.034 0.01-0.302-0.032-0.047-0.052 0.1-3.558-3.620-0.269-0.087 1-1.201-1.205-1.034-0.542 0.0010.010.11 0.0...

    work page arXiv

  7. [7]

    Anshu and S

    A. Anshu and S. Arunachalam, A survey on the complex- ity of learning quantum states, Nature Reviews Physics 6, 59 (2024)

    2024

  8. [8]

    Gebhart, R

    V. Gebhart, R. Santagati, A. A. Gentile, E. M. Gauger, D. Craig, N. Ares, L. Banchi, F. Marquardt, L. Pezz` e, and C. Bonato, Learning quantum systems, Nature Reviews Physics 5, 141 (2023)

    2023

Show all 102 references
  1. [9]

    Vogel and H

    K. Vogel and H. Risken, Determination of quasiproba- bility distributions in terms of probability distributions for the rotated quadrature phase, Phys. Rev. A 40, 2847 (1989)

    1989

  2. [10]

    Hradil, Quantum-state estimation, Phys

    Z. Hradil, Quantum-state estimation, Phys. Rev. A 55, R1561 (1997)

    1997

  3. [11]

    Paris and J

    M. Paris and J. Rehacek, Quantum state estimation , Vol. 13 649 (Springer Science & Business Media, 2004)

    2004

  4. [12]

    Cramer, M

    M. Cramer, M. B. Plenio, S. T. Flammia, R. Somma, D. Gross, S. D. Bartlett, O. Landon-Cardinal, D. Poulin, and Y.-K. Liu, Efficient quantum state tomography, Na- ture Communications 1, 149 (2010)

    2010

  5. [13]

    B. P. Lanyon, C. Maier, M. Holz¨ apfel, T. Baumgratz, C. Hempel, P. Jurcevic, I. Dhand, A. S. Buyskikh, A. J. Daley, M. Cramer, M. B. Plenio, R. Blatt, and C. F. Roos, Efficient tomography of a quantum many-body sys- tem, Nature Physics 13, 1158 (2017)

    2017

  6. [14]

    Schwemmer, G

    C. Schwemmer, G. T´ oth, A. Niggebaum, T. Moroder, D. Gross, O. G¨ uhne, and H. Weinfurter, Experimental comparison of efficient tomography schemes for a six- qubit state, Phys. Rev. Lett. 113, 040503 (2014)

    2014

  7. [15]

    Aaronson, Shadow tomography of quantum states, in Proceedings of the 50th annual ACM SIGACT sympo- sium on theory of computing (2018) pp

    S. Aaronson, Shadow tomography of quantum states, in Proceedings of the 50th annual ACM SIGACT sympo- sium on theory of computing (2018) pp. 325–338

    2018

  8. [16]

    J. M. Lukens, K. J. H. Law, A. Jasra, and P. Lougov- ski, A practical and efficient approach for bayesian quan- tum state estimation, New Journal of Physics 22, 063038 (2020)

    2020

  9. [17]

    Park and M

    C.-Y. Park and M. J. Kastoryano, Geometry of learning neural quantum states, Phys. Rev. Res. 2, 023232 (2020)

    2020

  10. [18]

    Ahmed, C

    S. Ahmed, C. S´ anchez Mu˜ noz, F. Nori, and A. F. Kockum, Classification and reconstruction of optical quantum states with deep neural networks, Phys. Rev. Res. 3, 033278 (2021)

    2021

  11. [19]

    Ahmed, C

    S. Ahmed, C. S´ anchez Mu˜ noz, F. Nori, and A. F. Kockum, Quantum state tomography with conditional generative adversarial networks, Phys. Rev. Lett. 127, 140502 (2021)

    2021

  12. [20]

    P. Cha, P. Ginsparg, F. Wu, J. Carrasquilla, P. L. McMa- hon, and E.-A. Kim, Attention-based quantum tomog- raphy, Machine Learning: Science and Technology 3, 01LT01 (2021)

    2021

  13. [21]

    Lange, M

    H. Lange, M. Kebriˇ c, M. Buser, U. Schollw¨ ock, F. Grusdt, and A. Bohrdt, Adaptive Quantum State Tomography with Active Learning, Quantum 7, 1129 (2023)

    2023

  14. [22]

    Gaikwad, O

    A. Gaikwad, O. Bihani, Arvind, and K. Dorai, Neural- network-assisted quantum state and process tomography using limited data sets, Phys. Rev. A109, 012402 (2024)

    2024

  15. [23]

    A. M. Palmieri, G. M¨ uller-Rigat, A. K. Srivastava, M. Lewenstein, G. Rajchel-Mieldzio´ c, and M. P lodzie´ n, Enhancing quantum state tomography via resource- efficient attention-based neural networks, Phys. Rev. Res. 6, 033248 (2024)

    2024

  16. [24]

    H. Ma, D. Dong, I. R. Petersen, C.-J. Huang, and G.-Y. Xiang, Neural networks for quantum state tomography with constrained measurements, Quantum Information Processing 23, 317 (2024)

    2024

  17. [25]

    Q.-Q. Wang, S. Dong, X.-W. Li, X.-Y. Xu, C. Wang, S. Han, M.-H. Yung, Y.-J. Han, C.-F. Li, and G.-C. Guo, Efficient learning of mixed-state tomography for photonic quantum walk, Science Advances 10, eadl4871 (2024)

    2024

  18. [26]

    Torlai, G

    G. Torlai, G. Mazzola, J. Carrasquilla, M. Troyer, R. Melko, and G. Carleo, Neural-network quantum state tomography, Nature Physics 14, 447 (2018)

    2018

  19. [27]

    Rocchetto, E

    A. Rocchetto, E. Grant, S. Strelchuk, G. Carleo, and S. Severini, Learning hard quantum distributions with variational autoencoders, npj Quantum Information 4, 28 (2018)

    2018

  20. [28]

    Carrasquilla, G

    J. Carrasquilla, G. Torlai, R. G. Melko, and L. Aolita, Reconstructing quantum states with generative models, Nature Machine Intelligence 1, 155 (2019)

    2019

  21. [29]

    A. M. Palmieri, E. Kovlakov, F. Bianchi, D. Yudin, S. Straupe, J. D. Biamonte, and S. Kulik, Experimen- tal neural network enhanced quantum tomography, npj Quantum Information 6, 20 (2020)

    2020

  22. [30]

    Gao and L.-M

    X. Gao and L.-M. Duan, Efficient representation of quan- tum many-body states with deep neural networks, Na- ture Communications 8, 662 (2017)

    2017

  23. [31]

    S. M. Lee, J. Lee, and J. Bang, Learning unknown pure quantum states, Phys. Rev. A 98, 052302 (2018)

    2018

  24. [32]

    S. M. Lee, H. S. Park, J. Lee, J. Kim, and J. Bang, Quan- tum state learning via single-shot measurements, Phys. Rev. Lett. 126, 170504 (2021)

    2021

  25. [33]

    Y. Liu, D. Wang, S. Xue, A. Huang, X. Fu, X. Qiang, P. Xu, H.-L. Huang, M. Deng, C. Guo, X. Yang, and J. Wu, Variational quantum circuits for quantum state tomography, Phys. Rev. A 101, 052316 (2020)

    2020

  26. [34]

    S. Xue, Y. Liu, Y. Wang, P. Zhu, C. Guo, and J. Wu, Variational quantum process tomography of unitaries, Phys. Rev. A 105, 032427 (2022)

    2022

  27. [35]

    Innan, O

    N. Innan, O. I. Siddiqui, S. Arora, T. Ghosh, Y. P. Ko¸ cak, D. Paragas, A. A. O. Galib, M. A.-Z. Khan, and M. Ben- nai, Quantum state tomography using quantum machine learning, Quantum Machine Intelligence 6, 28 (2024)

    2024

  28. [36]

    K. H. Wan, O. Dahlsten, H. Kristj´ ansson, R. Gardner, and M. S. Kim, Quantum generalisation of feedforward neural networks, npj Quantum Information 3, 36 (2017)

    2017

  29. [37]

    K. Beer, D. Bondarenko, T. Farrelly, T. J. Osborne, R. Salzmann, D. Scheiermann, and R. Wolf, Training deep quantum neural networks, Nature Communications 11, 808 (2020)

    2020

  30. [38]

    Haghshenas, J

    R. Haghshenas, J. Gray, A. C. Potter, and G. K.-L. Chan, Variational power of quantum circuit tensor networks, Phys. Rev. X 12, 011047 (2022)

    2022

  31. [39]

    Abbas, D

    A. Abbas, D. Sutter, C. Zoufal, A. Lucchi, A. Figalli, and S. Woerner, The power of quantum neural networks, Nature Computational Science 1, 403 (2021)

    2021

  32. [40]

    Farhi, J

    E. Farhi, J. Goldstone, and S. Gutmann, A quan- tum approximate optimization algorithm, arXiv preprint arXiv:1411.4028 (2014)

    2014 arXiv

  33. [41]

    Verdon, M

    G. Verdon, M. Broughton, J. R. McClean, K. J. Sung, R. Babbush, Z. Jiang, H. Neven, and M. Mohseni, Learn- ing to learn with quantum neural networks via classical neural networks (2019), arXiv:1907.05415 [quant-ph]

    2019 arXiv

  34. [42]

    Wilson, R

    M. Wilson, R. Stromswold, F. Wudarski, S. Hadfield, N. M. Tubman, and E. G. Rieffel, Optimizing quantum heuristics with meta-learning, Quantum Machine Intelli- gence 3, 13 (2021)

    2021

  35. [43]

    Khairy, R

    S. Khairy, R. Shaydulin, L. Cincio, Y. Alexeev, and P. Balaprakash, Learning to optimize variational quan- tum circuits to solve combinatorial problems, in Pro- ceedings of the AAAI conference on artificial intelligence, Vol. 34 (2020) pp. 2367–2375

    2020

  36. [44]

    Li and J

    K. Li and J. Malik, Learning to optimize, in International Conference on Learning Representations (2017)

    2017

  37. [45]

    Y. Chen, M. W. Hoffman, S. G. Colmenarejo, M. Denil, T. P. Lillicrap, M. Botvinick, and N. Freitas, Learning to learn without gradient descent by gradient descent, in International Conference on Machine Learning (PMLR,

  38. [46]

    A. W. R. Smith, J. Gray, and M. S. Kim, Efficient quan- tum state sample tomography with basis-dependent neu- ral networks, PRX Quantum 2, 020348 (2021)

    2021

  39. [47]

    Y. Quek, S. Fort, and H. K. Ng, Adaptive quantum state tomography with neural networks, npj Quantum Infor- 14 mation 7, 105 (2021)

    2021

  40. [48]

    Kandala, A

    A. Kandala, A. Mezzacapo, K. Temme, M. Takita, M. Brink, J. M. Chow, and J. M. Gambetta, Hardware- efficient variational quantum eigensolver for small molecules and quantum magnets, Nature 549, 242 (2017)

    2017

  41. [49]

    Salimans, J

    T. Salimans, J. Ho, X. Chen, S. Sidor, and I. Sutskever, Evolution strategies as a scalable alternative to reinforce- ment learning, arXiv preprint arXiv:1703.03864 (2017)

    2017 arXiv

  42. [50]

    Narvekar, B

    S. Narvekar, B. Peng, M. Leonetti, J. Sinapov, M. E. Taylor, and P. Stone, Curriculum learning for reinforce- ment learning domains: a framework and survey, J. Mach. Learn. Res. 21 (2020)

    2020

  43. [51]

    Ostaszewski, L

    M. Ostaszewski, L. M. Trenkwalder, W. Masarczyk, E. Scerri, and V. Dunjko, Reinforcement learning for op- timization of variational quantum circuit architectures, Advances in Neural Information Processing Systems 34, 18182 (2021)

    2021

  44. [52]

    Y. J. Patel, A. Kundu, M. Ostaszewski, X. Bonet- Monroig, V. Dunjko, and O. Danaci, Curriculum rein- forcement learning for quantum architecture search un- der hardware errors, in The Twelfth International Con- ference on Learning Representations (2024)

    2024

  45. [53]

    J. Haah, A. W. Harrow, Z. Ji, X. Wu, and N. Yu, Sample- optimal tomography of quantum states, IEEE Transac- tions on Information Theory 63, 5628 (2017)

    2017

  46. [54]

    R. S. Sutton and A. G. Barto, Reinforcement learning: An introduction (MIT press, 2018)

    2018

  47. [55]

    Bellman, The theory of dynamic programming, Bul- letin of the American Mathematical Society 60, 503 (1954)

    R. Bellman, The theory of dynamic programming, Bul- letin of the American Mathematical Society 60, 503 (1954)

    1954

  48. [56]

    K. J. ˚Astr¨ om, Optimal control of markov processes with incomplete state information ii: The convexity of the loss- function, Journal of Mathematical Analysis and Applica- tions 26, 403 (1969)

    1969

  49. [57]

    Hausknecht and P

    M. Hausknecht and P. Stone, Deep recurrent q-learning for partially observable mdps, in 2015 aaai fall sympo- sium series (2015)

    2015

  50. [58]

    Barry, D

    J. Barry, D. T. Barry, and S. Aaronson, Quantum par- tially observable markov decision processes, Phys. Rev. A 90, 032311 (2014)

    2014

  51. [59]

    V. V. Sivak, A. Eickbusch, H. Liu, B. Royer, I. Tsioutsios, and M. H. Devoret, Model-free quantum control with re- inforcement learning, Phys. Rev. X 12, 011059 (2022)

    2022

  52. [60]

    Bukov, A

    M. Bukov, A. G. R. Day, D. Sels, P. Weinberg, A. Polkovnikov, and P. Mehta, Reinforcement learning in different phases of quantum control, Phys. Rev. X 8, 031086 (2018)

    2018

  53. [61]

    F¨ osel, P

    T. F¨ osel, P. Tighineanu, T. Weiss, and F. Marquardt, Re- inforcement learning with neural networks for quantum feedback, Phys. Rev. X 8, 031084 (2018)

    2018

  54. [62]

    M. Y. Niu, S. Boixo, V. N. Smelyanskiy, and H. Neven, Universal quantum control through deep reinforcement learning, npj Quantum Information 5, 33 (2019)

    2019

  55. [63]

    Z. T. Wang, Y. Ashida, and M. Ueda, Deep reinforcement learning control of quantum cartpoles, Phys. Rev. Lett. 125, 100401 (2020)

    2020

  56. [64]

    Porotti, A

    R. Porotti, A. Essig, B. Huard, and F. Marquardt, Deep Reinforcement Learning for Quantum State Preparation with Weak Nonlinear Measurements, Quantum 6, 747 (2022)

    2022

  57. [65]

    Metz and M

    F. Metz and M. Bukov, Self-correcting quantum many- body control using reinforcement learning with tensor networks, Nature Machine Intelligence 5, 780 (2023)

    2023

  58. [66]

    Vent, Rechenberg, ingo, evolutionsstrategie — opti- mierung technischer systeme nach prinzipien der biologis- chen evolution

    W. Vent, Rechenberg, ingo, evolutionsstrategie — opti- mierung technischer systeme nach prinzipien der biologis- chen evolution. 170 s. mit 36 abb. frommann-holzboog- verlag. stuttgart 1973. broschiert, Feddes Repertorium 86, 337 (1975)

    1975

  59. [67]

    Hansen, The cma evolution strategy: A tutorial, arXiv preprint arXiv:1604.00772 (2016)

    N. Hansen, The cma evolution strategy: A tutorial, arXiv preprint arXiv:1604.00772 (2016)

    2016 arXiv

  60. [68]

    Mitarai, M

    K. Mitarai, M. Negoro, M. Kitagawa, and K. Fujii, Quan- tum circuit learning, Phys. Rev. A 98, 032309 (2018)

    2018

  61. [69]

    Schuld, V

    M. Schuld, V. Bergholm, C. Gogolin, J. Izaac, and N. Kil- loran, Evaluating analytic gradients on quantum hard- ware, Phys. Rev. A 99, 032331 (2019)

    2019

  62. [70]

    J. R. McClean, S. Boixo, V. N. Smelyanskiy, R. Bab- bush, and H. Neven, Barren plateaus in quantum neural network training landscapes, Nature Communications 9, 4812 (2018)

    2018

  63. [71]

    Cerezo and P

    M. Cerezo and P. J. Coles, Higher order derivatives of quantum neural networks with barren plateaus, Quan- tum Science and Technology 6, 035006 (2021)

    2021

  64. [72]

    Arrasmith, M

    A. Arrasmith, M. Cerezo, P. Czarnik, L. Cincio, and P. J. Coles, Effect of barren plateaus on gradient-free opti- mization, Quantum 5, 558 (2021)

    2021

  65. [73]

    Anand, M

    A. Anand, M. Degroote, and A. Aspuru-Guzik, Natu- ral evolutionary strategies for variational quantum com- putation, Machine Learning: Science and Technology 2, 045012 (2021)

    2021

  66. [74]

    J. Xie, C. Xu, C. Yin, Y. Dong, and Z. Zhang, Nat- ural evolutionary gradient descent strategy for varia- tional quantum algorithms, Intelligent Computing 2, 0042 (2023)

    2023

  67. [75]

    Mezzadri, How to generate random matrices from the classical compact groups (2007), arXiv:math-ph/0609050 [math-ph]

    F. Mezzadri, How to generate random matrices from the classical compact groups (2007), arXiv:math-ph/0609050 [math-ph]

    2007 arXiv

  68. [76]

    Paszke, S

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, et al. , Pytorch: An imperative style, high-performance deep learning library, Advances in neural information processing systems 32 (2019)

    2019

  69. [77]

    See https://github.com/quantum-jwjae/RL2LQS

  70. [78]

    S. J. Pan and Q. Yang, A survey on transfer learning, IEEE Transactions on Knowledge and Data Engineering 22, 1345 (2010)

    2010

  71. [79]

    R. Zen, L. My, R. Tan, F. H´ ebert, M. Gattobigio, C. Miniatura, D. Poletti, and S. Bressan, Transfer learn- ing for scalability of neural-network quantum states, Phys. Rev. E 101, 053301 (2020)

    2020

  72. [80]

    A. Mari, T. R. Bromley, J. Izaac, M. Schuld, and N. Killo- ran, Transfer learning in hybrid classical-quantum neural networks, Quantum 4, 340 (2020)

    2020

  73. [81]

    Bagan, M

    E. Bagan, M. Baig, R. Mu˜ noz Tapia, and A. Rodriguez, Collective versus local measurements in a qubit mixed- state estimation, Phys. Rev. A 69, 010304 (2004)

    2004

  74. [82]

    D. H. Mahler, L. A. Rozema, A. Darabi, C. Ferrie, R. Blume-Kohout, and A. M. Steinberg, Adaptive quan- tum state tomography improves accuracy quadratically, Phys. Rev. Lett. 111, 183601 (2013)

    2013

  75. [83]

    K. S. Kravtsov, S. S. Straupe, I. V. Radchenko, N. M. T. Houlsby, F. Husz´ ar, and S. P. Kulik, Experimental adap- tive bayesian tomography, Phys. Rev. A 87, 062122 (2013)

    2013

  76. [84]

    Ferrie, Self-guided quantum tomography, Phys

    C. Ferrie, Self-guided quantum tomography, Phys. Rev. Lett. 113, 190404 (2014)

    2014

  77. [85]

    Shen and S

    J. Shen and S. Castan, An optimal linear operator for step edge detection, CVGIP: Graphical Models and Im- 15 age Processing 54, 112 (1992)

    1992

  78. [86]

    D. N. Page, Average entropy of a subsystem, Phys. Rev. Lett. 71, 1291 (1993)

    1993

  79. [87]

    Sharma, A

    S. Sharma, A. S. Lakshminarayanan, and B. Ravindran, Learning to repeat: Fine grained action repetition for deep reinforcement learning, in International Conference on Learning Representations (2017)

    2017

  80. [88]

    Silver, A

    D. Silver, A. Huang, C. J. Maddison, A. Guez, L. Sifre, G. van den Driessche, J. Schrittwieser, I. Antonoglou, V. Panneershelvam, M. Lanctot, S. Dieleman, D. Grewe, J. Nham, N. Kalchbrenner, I. Sutskever, T. Lillicrap, M. Leach, K. Kavukcuoglu, T. Graepel, and D. Hassabis, Mas...

    2016

  81. [89]

    Porotti, V

    R. Porotti, V. Peano, and F. Marquardt, Gradient- ascent pulse engineering with feedback, PRX Quantum 4, 030305 (2023)

    2023

  82. [90]

    I. S. Maria Schuld and F. Petruccione, An introduction to quantum machine learn- ing, Contemporary Physics 56, 172 (2015), https://doi.org/10.1080/00107514.2014.964942

    2015

  83. [91]

    Abbas, A

    A. Abbas, A. Ambainis, B. Augustino, A. B¨ artschi, H. Buhrman, C. Coffrin, G. Cortiana, V. Dunjko, D. J. Egger, B. G. Elmegreen, N. Franco, F. Fratini, B. Fuller, J. Gacon, C. Gonciulea, S. Gribling, S. Gupta, S. Had- field, R. Heese, G. Kircher, T. Kleinert, T. Koch, G. Ko- ...

    2024

  84. [92]

    Y. Chen, M. W. Hoffman, S. G. Colmenarejo, M. Denil, T. P. Lillicrap, M. Botvinick, and N. de Freitas, Learn- ing to learn without gradient descent by gradient de- scent, in Proceedings of the 34th International Conference on Machine Learning , Proceedings of Machine Learning ...

    2017

  85. [93]

    V. TV, P. Malhotra, J. Narwariya, L. Vig, and G. Shroff, Meta-learning for black-box optimization, in Joint Eu- ropean Conference on Machine Learning and Knowledge Discovery in Databases (Springer, 2019) pp. 366–381

    2019

  86. [94]

    Shala, A

    G. Shala, A. Biedenkapp, N. Awad, S. Adriaensen, M. Lindauer, and F. Hutter, Learning step-size adapta- tion in cma-es, in Parallel Problem Solving from Nature – PPSN XVI , edited by T. B¨ ack, M. Preuss, A. Deutz, H. Wang, C. Doerr, M. Emmerich, and H. Trautmann (Springer Int...

    2020

  87. [95]

    R. T. Lange, T. Schaul, Y. Chen, T. Zahavy, V. Dal- ibard, C. Lu, S. Singh, and S. Flennerhag, Discovering evolution strategies via meta-black-box optimization, in The Eleventh International Conference on Learning Rep- resentations (2023)

    2023

  88. [96]

    Konda and J

    V. Konda and J. Tsitsiklis, Actor-critic algorithms, Ad- vances in Neural Information Processing Systems 12 (1999)

    1999

  89. [97]

    Z. Wang, V. Bapst, N. Heess, V. Mnih, R. Munos, K. Kavukcuoglu, and N. De Freitas, Sample efficient actor-critic with experience replay, Proceedings of the International Conference on Learning Representations (ICLR) (2017)

    2017

  90. [98]

    V. Mnih, K. Kavukcuoglu, D. Silver, A. A. Rusu, J. Veness, M. G. Bellemare, A. Graves, M. Riedmiller, A. K. Fidjeland, G. Ostrovski, S. Petersen, C. Beat- tie, A. Sadik, I. Antonoglou, H. King, D. Kumaran, D. Wierstra, S. Legg, and D. Hassabis, Human-level con- trol through de...

    2015

  91. [99]

    V. Mnih, A. P. Badia, M. Mirza, A. Graves, T. Lillicrap, T. Harley, D. Silver, and K. Kavukcuoglu, Asynchronous methods for deep reinforcement learning, in Proceedings of The 33rd International Conference on Machine Learn- ing, Proceedings of Machine Learning Research, Vol. 48...

    2016

  92. [100]

    Kocsis and C

    L. Kocsis and C. Szepesv´ ari, Bandit based monte-carlo planning, in Machine Learning: ECML 2006 , edited by J. F¨ urnkranz, T. Scheffer, and M. Spiliopoulou (Springer Berlin Heidelberg, Berlin, Heidelberg, 2006) pp. 282–293

    2006

  93. [101]

    D. P. Kingma and J. Ba, Adam: A method for stochastic optimization, arXiv preprint arXiv:1412.6980 (2014)

    2014 arXiv

  94. [102]

    ˇReh´ aˇ cek, Z

    J. ˇReh´ aˇ cek, Z. c. v. Hradil, E. Knill, and A. I. Lvovsky, Diluted maximum-likelihood algorithm for quantum to- mography, Phys. Rev. A 75, 042108 (2007)

    2007

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.