pith. sign in

arxiv: 2606.24932 · v1 · pith:WE32SLGZnew · submitted 2026-06-22 · 🪐 quant-ph · cs.AI· cs.ET· cs.LG· cs.NE

Recursive QLSTM with Dynamic Variational Quantum Circuit Adaptation

Samuel Yen-Chi Chen , Yifeng Peng , Jiun-Cheng Jiang , Chun-Hua Lin , Kuo-Chung Peng , Junghoon Justin Park , Huan-Hsin Tseng , Hsin-Yi Lin
show 3 more authors
Kuan-Cheng Chen Chen-Yu Liu Shinjae Yoo
This is my paper

Pith reviewed 2026-06-26 08:10 UTC · model grok-4.3

classification 🪐 quant-ph cs.AIcs.ETcs.LGcs.NE
keywords recursive QLSTMquantum machine learningvariational quantum circuitstime series processingsequential dataquantum recurrent modelsmetacore constructionstemporal information propagation
0
0 comments X

The pith

Recursive QLSTM uses metacore constructions to process time series of varying lengths more effectively than fixed QLSTM.

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

The paper proposes Recursive QLSTM as an extension of QLSTM that applies metacore-based recursive constructions to handle sequential data. Numerical tests across different sequence lengths, metacore designs, and recursive rules identify the strongest variant among them. Theoretical arguments then show how the recursive structure improves propagation of temporal information through the network. A reader would care because many real-world time series have inconsistent lengths, and a flexible quantum recurrent model could adapt without redesign for each case.

Core claim

Recursive QLSTM extends QLSTM through metacore-based recursive constructions. Numerical experiments under varying input lengths, metacore designs, and recursive rules select the best architecture, while theoretical arguments establish that the recursive structure improves temporal information propagation and enhances learning performance for sequences of different lengths.

What carries the argument

Metacore-based recursive constructions that allow dynamic adaptation of variational quantum circuits within the QLSTM framework.

If this is right

  • The selected recursive architecture improves learning on tested sequence lengths compared with other variants.
  • Different recursive rules can be chosen to tune temporal information flow for specific tasks.
  • The model supplies a single flexible framework that covers input time series of many lengths without separate redesigns.
  • Theoretical reasoning links the recursive structure directly to better propagation of temporal features.

Where Pith is reading between the lines

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

  • The same metacore recursion pattern could be tried on other quantum recurrent architectures such as quantum RNNs.
  • If noise scaling stays favorable, the approach might support practical deployment on longer sequences using current quantum hardware.
  • It points toward a general route for making quantum sequence models scale with input size through structural repetition rather than added parameters.

Load-bearing premise

The variational quantum circuits realizing the recursive constructions keep depth and noise manageable as sequence length grows.

What would settle it

Numerical runs on longer sequences showing that required circuit depth produces noise levels that erase any performance gain over non-recursive QLSTM.

Figures

Figures reproduced from arXiv: 2606.24932 by Chen-Yu Liu, Chun-Hua Lin, Hsin-Yi Lin, Huan-Hsin Tseng, Jiun-Cheng Jiang, Junghoon Justin Park, Kuan-Cheng Chen, Kuo-Chung Peng, Samuel Yen-Chi Chen, Shinjae Yoo, Yifeng Peng.

Figure 1
Figure 1. Figure 1: General variational quantum neural network architecture. The input is encoded by U(x) and then processed by a learnable quantum circuit W(Θ). QQ q=1 RY (Θℓq)  EQ|ψ (ℓ−1) t ⟩ for ℓ = 1, . . . , L. The VQC output is the vector of Pauli-Z expectation values on the first H qubits, V(ut; Θ) = [⟨Z1⟩ψ (L) t , . . . ,⟨ZH⟩ψ (L) t ] ⊤ ∈ [−1, 1]H, where ⟨Zh⟩ψ (L) t = ⟨ψ (L) t |Zh|ψ (L) t ⟩. |0⟩ H Ry(x1) Ry(θ1) |0⟩ … view at source ↗
Figure 2
Figure 2. Figure 2: , which has been investigated in prior studies [57]. For an input ut and VQC parameter matrix Θ ∈ R L×Q, the circuit first applies Hadamard gates and input-encoding rotations, giving |ψ (0) t ⟩ = QQ q=1 RY (ut,q)Hq  |0⟩ ⊗Q. Each variational layer applies a nearest-neighbor CNOT entangling layer EQ followed by trainable rotations, i.e., |ψ (ℓ) t ⟩ = U(⃗x) W(Θ) |0⟩ |0⟩ |0⟩ . . . . . . Encoding Circuit Lear… view at source ↗
Figure 3
Figure 3. Figure 3: Quantum LSTM cell. The input xt and previous hidden state ht−1 are concatenated and processed by four VQC-based gates to update the cell state ct and hidden state ht. V. RECURSIVE QLSTM The Recursive QLSTM introduces a trainable MetaCore network that dynamically generates or adapts VQC parameters from the current recurrent context. Define zt = [ht−1; ct−1; xt] ∈ R 2H+D. (9) The MetaCore produces four gate-… view at source ↗
Figure 5
Figure 5. Figure 5: MetaCore architecture for Enc1NNGate. A shared encoder extracts the latent representation et from zt, while gate-specific heads produce parameter updates for the four QLSTM gates. B. Recursive Rules 1) Base Parameter plus MetaCore Delta (base+delta): This rule keeps a trainable base parameter matrix for each gate and adds the dynamic MetaCore output, i.e., Θa t = Θ¯ a + ∆a t for a ∈ {i, f, g, o}, where Θ¯ … view at source ↗
Figure 4
Figure 4. Figure 4: Recursive Quantum LSTM cell. The MetaCore generates time￾dependent VQC parameter updates from zt = [ht−1; ct−1; xt], enabling adaptive recurrent quantum gates. A. MetaCore Variants Let R = LQ and m denote the latent dimension. 1) SingleNN: The single-linear MetaCore directly gener￾ates all gate-wise updates as rt = W zt + b ∈ R 4R, followed by (∆i t , ∆ f t , ∆ g t , ∆o t ) = reshape(rt, 4, L, Q), making i… view at source ↗
Figure 6
Figure 6. Figure 6: Recursive parameter update rules for the effective VQC param￾eters. The model considers base-plus-delta (base+delta), MetaCore-only (meta-only), and recurrent delta-based (delta) parameter modulation. of QNN layers L = 5, learning_rate=10−3 , Adam, and random seeds 0, 1, and 2. Numerical results are orga￾nized by task, with three complementary views per dataset. First, epoch-wise prediction snapshots at se… view at source ↗
Figure 7
Figure 7. Figure 7: compares QLSTM with rQLSTM Enc1NNGate base+delta at sequence length 16. At epoch 1, both models fail to reproduce the decaying oscillatory structure. By epoch 15, the recursive model already matches the waveform across the train and test regions, whereas QLSTM still shows visible amplitude mismatch. The gap narrows at epoch 30, and both achieve high accuracy by epoch 100, indicating that recursion mainly a… view at source ↗
Figure 8
Figure 8. Figure 8: reports the phase-0 train/test convergence results at sequence length 16, comparing all candidate recursive variants against the QLSTM baseline. A consistent trend emerges in [PITH_FULL_IMAGE:figures/full_fig_p004_8.png] view at source ↗
Figure 10
Figure 10. Figure 10: Epoch-wise prediction comparison between QLSTM and rQL￾STM Enc1NNGate base+delta on the damped_shm dataset under the seq_len=16 setting. Rows show selected training epochs. Blue curves denote the mean prediction over seeds, red dashed curves denote the ground truth, and the orange dashed line marks the train/test split. b) Damped-SHM.: We next consider the damped_shm task [PITH_FULL_IMAGE:figures/full_fi… view at source ↗
Figure 9
Figure 9. Figure 9: Summary performance comparison versus sequence length on the bessel_j2 dataset for sequence lengths 4, 8, 16, 32, and 64. We compare the baseline QLSTM with selected rQLSTM variants under different MetaCore designs and recursive update rules. The three panels report final test loss, AUC@20 test loss, and t95 test epoch, respectively [PITH_FULL_IMAGE:figures/full_fig_p005_9.png] view at source ↗
Figure 11
Figure 11. Figure 11: Train and test loss convergence comparison on the damped_shm dataset with seq_len=16. We compare the baseline QLSTM against all rQLSTM variants with different MetaCore designs and recursive rules. Curves show the mean over seeds with shaded regions indicating standard deviation, and the inset zooms into the first five epochs [PITH_FULL_IMAGE:figures/full_fig_p006_11.png] view at source ↗
Figure 13
Figure 13. Figure 13: Epoch-wise prediction comparison between QLSTM and rQL￾STM Enc1NNGate base+delta on the delayed_quantum_control dataset under the seq_len=16 setting. Rows show selected training epochs. Blue curves denote the mean prediction over seeds, red dashed curves denote the ground truth, and the orange dashed line marks the train/test split [PITH_FULL_IMAGE:figures/full_fig_p006_13.png] view at source ↗
Figure 15
Figure 15. Figure 15: summarizes the phase-1 results across sequence lengths 4, 8, 16, 32, and 64 for the retained variants, with the three panels reporting final test loss, AUC@20 test loss, and t95 test epoch, respectively. The recursive models remain advantageous in optimization-oriented metrics at short and medium sequence lengths. However, the delta rule becomes noticeably less stable as the sequence length increases: in … view at source ↗
Figure 16
Figure 16. Figure 16: Epoch-wise prediction comparison between QLSTM and rQLSTM Enc1NNGate base+delta on the narma_5 dataset under the seq_len=16 setting. Rows show selected training epochs. Blue curves denote the mean prediction over seeds, red dashed curves denote the ground truth, and the orange dashed line marks the train/test split [PITH_FULL_IMAGE:figures/full_fig_p008_16.png] view at source ↗
Figure 20
Figure 20. Figure 20: Train and test loss convergence comparison on the narma_10 dataset with seq_len=16. We compare the baseline QLSTM against all rQLSTM variants with different MetaCore designs and recursive rules. Curves show the mean over seeds with shaded regions indicating standard deviation, and the inset zooms into the first five epochs. in particular, delta-based variants degrade noticeably at longer sequence lengths … view at source ↗
Figure 19
Figure 19. Figure 19: Epoch-wise prediction comparison between QLSTM and rQL￾STM Enc1NNGate base+delta on the narma_10 dataset under the seq_len=16 setting. Rows show selected training epochs. Blue curves denote the mean prediction over seeds, red dashed curves denote the ground truth, and the orange dashed line marks the train/test split [PITH_FULL_IMAGE:figures/full_fig_p009_19.png] view at source ↗
Figure 21
Figure 21. Figure 21: Summary performance comparison versus sequence length on the narma_10 dataset for sequence lengths 4, 8, 16, 32, and 64. We compare the baseline QLSTM with selected rQLSTM variants under different MetaCore designs and recursive update rules. The three panels report final test loss, AUC@20 test loss, and t95 test epoch, respectively. gates are first-order context-dependent corrections of the base QLSTM gat… view at source ↗
read the original abstract

Recent advances in quantum computing and machine learning have motivated the development of quantum models for sequential data processing. In this paper, we propose a Recursive Quantum Long Short-Term Memory model, or Recursive QLSTM, which extends QLSTM through metacore-based recursive constructions. We numerically test the model under different input sequence lengths, metacore designs, and recursive rules, and identify the best-performing architecture among these variants. For this selected model, we further provide theoretical arguments explaining why its recursive structure improves temporal information propagation and enhances learning performance. Our results suggest that Recursive QLSTM offers a flexible and effective framework for quantum recurrent learning over input time series of various lengths.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

1 major / 0 minor

Summary. The manuscript proposes Recursive QLSTM, extending standard QLSTM via metacore-based recursive constructions for processing input time series of varying lengths. It reports numerical tests over different sequence lengths, metacore designs, and recursive rules to select the best variant, followed by theoretical arguments that the recursive structure improves temporal information propagation and learning performance.

Significance. If the numerical results and theoretical arguments hold under realistic conditions, the work could supply a flexible quantum recurrent architecture. The explicit comparison of multiple metacore and recursion variants is a strength, as is the attempt to link recursion to better information flow. However, the central claim of practicality for sequences of various lengths rests on unverified scaling behavior.

major comments (1)
  1. [Abstract] Abstract: the numerical tests are described as covering 'different input sequence lengths' yet give no indication that depth scaling, error accumulation, or hardware noise models were included. This directly affects the load-bearing assumption that metacore recursion remains NISQ-feasible as sequence length grows, as noted in the stress-test concern.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback. We address the concern about the abstract's description of the numerical tests below and will revise the manuscript accordingly.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the numerical tests are described as covering 'different input sequence lengths' yet give no indication that depth scaling, error accumulation, or hardware noise models were included. This directly affects the load-bearing assumption that metacore recursion remains NISQ-feasible as sequence length grows, as noted in the stress-test concern.

    Authors: We agree that the abstract does not explicitly state the ideal nature of the simulations. Our numerical experiments test the Recursive QLSTM variants on classical emulators of the quantum circuits for varying sequence lengths, metacore designs, and recursive rules, but these are noiseless simulations that do not include hardware noise models, explicit depth scaling studies, or error accumulation analysis. The central claims rest on these ideal-case results plus theoretical arguments for improved temporal information propagation via the recursive structure; we do not claim verified NISQ feasibility or scaling behavior for growing sequence lengths under realistic noise. To address the concern, we will revise the abstract to specify that the tests are performed under ideal conditions without noise models. This clarification will be incorporated in the revised version. revision: yes

Circularity Check

0 steps flagged

No circularity in derivation chain

full rationale

The paper introduces Recursive QLSTM via metacore-based recursion, performs numerical tests across sequence lengths and variants to select a best architecture, and then supplies separate theoretical arguments for improved temporal propagation in that architecture. No equations, predictions, or uniqueness claims are shown to reduce by construction to fitted inputs, self-citations, or prior ansatzes from the same authors. The numerical experiments and theoretical explanations remain independent of each other, with the former serving as empirical selection and the latter as explanatory support rather than tautological restatement.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated in the provided text.

pith-pipeline@v0.9.1-grok · 5689 in / 964 out tokens · 13836 ms · 2026-06-26T08:10:14.242724+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

71 extracted references · 38 canonical work pages

  1. [1]

    M. A. Nielsen and I. L. Chuang,Quantum Computation and Quantum Information, 10th ed. Cambridge University Press, 2010. [Online]. Available: https://doi.org/10.1017/CBO9780511976667

    work page doi:10.1017/cbo9780511976667 2010

  2. [2]

    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, W.-K. Mok, S. Sim, L.-C. Kwek, and A. Aspuru-Guzik, “Noisy intermediate-scale quantum algorithms,”Reviews of Modern Physics, vol. 94, no. 1, p. 015004, 2022. [Online]. Available: https://doi.org/10.1103/RevModPhys.94.015004

    work page doi:10.1103/revmodphys.94.015004 2022

  3. [3]

    Cerezo, Andrew Arrasmith, Ryan Babbush, Simon C

    M. Cerezo, A. Arrasmith, R. Babbush, S. C. Benjamin, S. Endo, K. Fujii, J. R. McClean, K. Mitarai, X. Yuan, L. Cincio, and P. J. Coles, “Variational quantum algorithms,”Nature Reviews Physics, vol. 3, no. 9, pp. 625–644, 2021. [Online]. Available: https://doi.org/10.1038/s42254-021-00348-9

    work page doi:10.1038/s42254-021-00348-9 2021

  4. [4]

    Mitarai, M

    K. Mitarai, M. Negoro, M. Kitagawa, and K. Fujii, “Quantum circuit learning,”Physical Review A, vol. 98, no. 3, p. 032309, 2018. [Online]. Available: https://doi.org/10.1103/PhysRevA.98.032309

    work page doi:10.1103/physreva.98.032309 2018

  5. [5]

    Cavity electro-optic circuit for microwave-to-optical conversion in the quantum ground state

    M. Schuld, A. Bocharov, K. M. Svore, and N. Wiebe, “Circuit- centric quantum classifiers,”Physical Review A, vol. 101, no. 3, p. 032308, 2020. [Online]. Available: https://doi.org/10.1103/PhysRevA. 101.032308

    work page doi:10.1103/physreva 2020

  6. [6]

    Quantum convolutional neural network for classical data classification,

    T. Hur, L. Kim, and D. K. Park, “Quantum convolutional neural network for classical data classification,”Quantum Machine Intelligence, vol. 4, no. 1, p. 3, 2022

    2022

  7. [7]

    Validating large-scale quantum ma- chine learning: Efficient simulation of quantum support vector machines using tensor networks,

    K.-C. Chen, T.-Y . Li, Y .-Y . Wang, S. See, C.-C. Wang, R. Wille, N.-Y . Chen, A.-C. Yang, and C.-Y . Lin, “Validating large-scale quantum ma- chine learning: Efficient simulation of quantum support vector machines using tensor networks,”Machine Learning: Science and Technology, vol. 6, no. 1, p. 015047, 2025

    2025

  8. [8]

    Transformer-based multi-aspect multi-granularity non-native english speaker pronunciation assessment,

    S. Y .-C. Chen, S. Yoo, and Y .-L. L. Fang, “Quantum long short-term memory,” inICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2022, pp. 8622–8626. [Online]. Available: https://doi.org/10.1109/ICASSP43922. 2022.9747369

    work page doi:10.1109/icassp43922 2022

  9. [9]

    Quantum recurrent neural networks for sequential learning,

    Y . Li, Z. Wang, R. Han, S. Shi, J. Li, R. Shang, H. Zheng, G. Zhong, and Y . Gu, “Quantum recurrent neural networks for sequential learning,” Neural Networks, vol. 166, pp. 148–161, 2023. [Online]. Available: https://doi.org/10.1016/j.neunet.2023.07.003

    work page doi:10.1016/j.neunet.2023.07.003 2023

  10. [10]

    Enhancing and Adversarial: Improve ASR with Speaker Labels

    S. S. Li, X. Zhang, S. Zhou, H. Shu, R. Liang, H. Liu, and L. P. Garcia, “Pqlm-multilingual decentralized portable quantum language model,” in ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2023, pp. 1–5. [Online]. Available: https://doi.org/10.1109/ICASSP49357.2023.10095215

    work page doi:10.1109/icassp49357.2023.10095215 2023

  11. [11]

    Variational quantum circuits for deep reinforcement learning,

    S. Y .-C. Chen, C.-H. H. Yang, J. Qi, P.-Y . Chen, X. Ma, and H.-S. Goan, “Variational quantum circuits for deep reinforcement learning,” IEEE access, vol. 8, pp. 141 007–141 024, 2020. [Online]. Available: https://doi.org/10.1109/ACCESS.2020.3010470

    work page doi:10.1109/access.2020.3010470 2020

  12. [12]

    Enhancing and Adversarial: Improve ASR with Speaker Labels

    S. Y .-C. Chen, “Quantum deep recurrent reinforcement learning,” in ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2023, pp. 1–5. [Online]. Available: https://doi.org/10.1109/ICASSP49357.2023.10096981

    work page doi:10.1109/icassp49357.2023.10096981 2023

  13. [13]

    A survey on quantum reinforcement learning,

    N. Meyer, C. Ufrecht, M. Periyasamy, D. D. Scherer, A. Plinge, and C. Mutschler, “A survey on quantum reinforcement learning,” arXiv preprint arXiv:2211.03464, 2022. [Online]. Available: https: //doi.org/10.48550/arXiv.2211.03464

    work page doi:10.48550/arxiv.2211.03464 2022

  14. [14]

    Quantum-train- based distributed multi-agent reinforcement learning,

    K.-C. Chen, S. Y .-C. Chen, C.-Y . Liu, and K. K. Leung, “Quantum-train- based distributed multi-agent reinforcement learning,” in2025 IEEE Symposium for Multidisciplinary Computational Intelligence Incubators (MCII Companion). IEEE, 2025, pp. 1–5

    2025

  15. [15]

    The dawn of quantum natural language processing,

    R. Di Sipio, J.-H. Huang, S. Y .-C. Chen, S. Mangini, and M. Worring, “The dawn of quantum natural language processing,” inICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2022, pp. 8612–8616. [Online]. Available: https://doi.org/10.1109/ICASSP43922.2022.9747675

    work page doi:10.1109/icassp43922.2022.9747675 2022

  16. [16]

    QAOA with n·p≥ 200

    J. Stein, I. Christ, N. Kraus, M. B. Mansky, R. M ¨uller, and C. Linnhoff-Popien, “Applying qnlp to sentiment analysis in finance,” in2023 IEEE International Conference on Quantum Computing and Engineering (QCE), vol. 2. IEEE, 2023, pp. 20–25. [Online]. Available: https://doi.org/10.1109/QCE57702.2023.10178

    work page doi:10.1109/qce57702.2023.10178 2023

  17. [17]

    Lstm-qgan: Scalable nisq generative adversarial network,

    C. Chu, A. Hastak, and F. Chen, “Lstm-qgan: Scalable nisq generative adversarial network,” inICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2025, pp. 1–5. [Online]. Available: https://doi.org/10.1109/ ICASSP49660.2025.10888847

    arXiv 2025

  18. [18]

    Quantum kernel-based long short-term memory for climate time-series forecasting,

    Y .-C. Hsu, N.-Y . Chen, T.-Y . Li, P.-H. H. Lee, and K.-C. Chen, “Quantum kernel-based long short-term memory for climate time-series forecasting,” in2025 International Conference on Quantum Communica- tions, Networking, and Computing (QCNC). IEEE, 2025, pp. 421–426

    2025

  19. [19]

    Toward large-scale distributed quantum long short-term memory with modular quantum computers,

    K.-C. Chen, S. Y .-C. Chen, C.-Y . Liu, and K. K. Leung, “Toward large-scale distributed quantum long short-term memory with modular quantum computers,” in2025 International Wireless Communications and Mobile Computing (IWCMC). IEEE, 2025, pp. 337–342

    2025

  20. [20]

    Recurrent networks and NARMA modeling,

    J. Connor, L. E. Atlas, and D. R. Martin, “Recurrent networks and NARMA modeling,” inAdvances in Neural Information Processing Systems, vol. 4, 1991, pp. 301–308. [Online]. Available: https://proceedings.neurips.cc/paper files/paper/ 1991/hash/5ef0b4eba35ab2d6180b0bca7e46b6f9-Abstract.html

    1991

  21. [21]

    Neural Computation 9(8), 1735–1780 (1997) https://doi.org/10.1162/neco.1997.9.8.1735

    S. Hochreiter and J. Schmidhuber, “Long short-term memory,”Neural Computation, vol. 9, no. 8, pp. 1735–1780, 1997. [Online]. Available: https://doi.org/10.1162/neco.1997.9.8.1735

    work page doi:10.1162/neco.1997.9.8.1735 1997

  22. [22]

    Time-series forecasting with deep learning: a survey,

    B. Lim and S. Zohren, “Time-series forecasting with deep learning: a survey,”Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, vol. 379, no. 2194, p. 20200209, 2021. [Online]. Available: https://doi.org/10.1098/rsta. 2020.0209

    work page doi:10.1098/rsta 2021

  23. [23]

    and Lloyd, S

    J. Biamonte, P. Wittek, N. Pancotti, P. Rebentrost, N. Wiebe, and S. Lloyd, “Quantum machine learning,”Nature, vol. 549, no. 7671, pp. 195–202, 2017. [Online]. Available: https://doi.org/10.1038/nature23474

    work page doi:10.1038/nature23474 2017

  24. [24]

    Parameterized quantum circuits as machine learning models , volume=

    M. Benedetti, E. Lloyd, S. Sack, and M. Fiorentini, “Parameterized quantum circuits as machine learning models,”Quantum Science and Technology, vol. 4, no. 4, p. 043001, 2019. [Online]. Available: https://doi.org/10.1088/2058-9565/ab4eb5

    work page doi:10.1088/2058-9565/ab4eb5 2019

  25. [25]

    McClean, Sergio Boixo, Vadim N

    J. R. McClean, S. Boixo, V . N. Smelyanskiy, R. Babbush, and H. Neven, “Barren plateaus in quantum neural network training landscapes,”Nature Communications, vol. 9, no. 1, p. 4812, 2018. [Online]. Available: https://doi.org/10.1038/s41467-018-07090-4

    work page doi:10.1038/s41467-018-07090-4 2018

  26. [26]

    Expressibility and entangling capability of parameterized quantum circuits for hybrid quantum-classical algorithms,

    S. Sim, P. D. Johnson, and A. Aspuru-Guzik, “Expressibility and entangling capability of parameterized quantum circuits for hybrid quantum-classical algorithms,”Advanced Quantum Technologies, vol. 2, no. 12, p. 1900070, 2019. [Online]. Available: https: //doi.org/10.1002/qute.201900070

    work page doi:10.1002/qute.201900070 2019

  27. [27]

    Abbas, D

    A. Abbas, D. Sutter, C. Zoufal, A. Lucchi, A. Figalli, and S. Woerner, “The power of quantum neural networks,”Nature Computational Science, vol. 1, no. 6, pp. 403–409, 2021. [Online]. Available: https://doi.org/10.1038/s43588-021-00084-1

    work page doi:10.1038/s43588-021-00084-1 2021

  28. [28]

    Connecting ansatz expressibility to gradient magnitudes and barren plateaus,

    Z. Holmes, K. Sharma, M. Cerezo, and P. J. Coles, “Connecting ansatz expressibility to gradient magnitudes and barren plateaus,” PRX Quantum, vol. 3, no. 1, p. 010313, 2022. [Online]. Available: https://doi.org/10.1103/PRXQuantum.3.010313

    work page doi:10.1103/prxquantum.3.010313 2022

  29. [29]

    Data re-uploading for a universal quantum classifier,

    A. P ´erez-Salinas, A. Cervera-Lierta, E. Gil-Fuster, and J. I. Latorre, “Data re-uploading for a universal quantum classifier,”Quantum, vol. 4, p. 226, 2020. [Online]. Available: https://doi.org/10.22331/ q-2020-02-06-226

    2020

  30. [30]

    Effect of data encoding on the expressive power of variational quantum-machine-learning models,

    M. Schuld, R. Sweke, and J. J. Meyer, “Effect of data encoding on the expressive power of variational quantum-machine-learning models,” Physical Review A, vol. 103, no. 3, p. 032430, 2021. [Online]. Available: https://doi.org/10.1103/PhysRevA.103.032430

    work page doi:10.1103/physreva.103.032430 2021

  31. [31]

    Neural quantum embedding via deterministic quantum computation with one qubit,

    H. Liu, T. Hur, S. Zhang, L. Che, X. Long, X. Wang, K. Huang, Y .-a. Fan, Y . Zheng, Y . Feng, Y . Zhou, J. Ng, X. Nie, D. K. Park, and D. Lu, “Neural quantum embedding via deterministic quantum computation with one qubit,”Physical Review Letters, vol. 135, no. 8, p. 080603,

  32. [32]

    Available: https://doi.org/10.1103/y8wr-yml4

    [Online]. Available: https://doi.org/10.1103/y8wr-yml4

    work page doi:10.1103/y8wr-yml4

  33. [33]

    Recurrent quantum neural networks,

    J. Bausch, “Recurrent quantum neural networks,” inAd- vances in Neural Information Processing Systems, vol. 33. Curran Associates, Inc., 2020, pp. 1368–1379. [On- line]. Available: https://proceedings.neurips.cc/paper files/paper/2020/ file/0ec96be397dd6d3cf2fecb4a2d627c1c-Paper.pdf

    2020

  34. [34]

    Learning temporal data with a variational quantum recurrent neural network,

    Y . Takaki, K. Mitarai, M. Negoro, K. Fujii, and M. Kitagawa, “Learning temporal data with a variational quantum recurrent neural network,”Physical Review A, vol. 103, no. 5, p. 052414, 2021. [Online]. Available: https://doi.org/10.1103/PhysRevA.103.052414

    work page doi:10.1103/physreva.103.052414 2021

  35. [35]

    Harnessing disordered-ensemble quantum dynamics for machine learning,

    K. Fujii and K. Nakajima, “Harnessing disordered-ensemble quantum dynamics for machine learning,”Physical Review Applied, vol. 8, no. 2, p. 024030, 2017. [Online]. Available: https://doi.org/10.1103/ PhysRevApplied.8.024030

    2017

  36. [36]

    Rapid training of quantum recurrent neural networks,

    M. Siemaszko, A. Buraczewski, B. Le Saux, and M. Stobi ´nska, “Rapid training of quantum recurrent neural networks,”Quantum Machine Intelligence, vol. 5, no. 2, p. 31, 2023. [Online]. Available: https://doi.org/10.1007/s42484-023-00117-0

    work page doi:10.1007/s42484-023-00117-0 2023

  37. [37]

    Quantum reservoir computing using arrays of rydberg atoms,

    R. Araiza Bravo, K. Najafi, X. Gao, and S. F. Yelin, “Quantum reservoir computing using arrays of rydberg atoms,”PRX Quantum, vol. 3, no. 3, p. 030325, 2022. [Online]. Available: https://doi.org/10. 1103/PRXQuantum.3.030325

    2022

  38. [38]

    Time-series quantum reservoir computing with weak and projective measurements,

    P. Mujal, R. Mart ´ınez-Pe˜na, G. L. Giorgi, M. C. Soriano, and R. Zambrini, “Time-series quantum reservoir computing with weak and projective measurements,”npj Quantum Information, vol. 9, no. 1, p. 16,

  39. [39]

    Available: https://doi.org/10.1038/s41534-023-00682-z

    [Online]. Available: https://doi.org/10.1038/s41534-023-00682-z

    work page doi:10.1038/s41534-023-00682-z

  40. [40]

    Overcoming the coherence time barrier in quantum machine learning on temporal data,

    F. Hu, S. A. Khan, N. T. Bronn, G. Angelatos, G. E. Rowlands, G. J. Ribeill, and H. E. T ¨ureci, “Overcoming the coherence time barrier in quantum machine learning on temporal data,”Nature Communications, vol. 15, no. 1, p. 7491, 2024. [Online]. Available: https://doi.org/10.1038/s41467-024-51162-7

    work page doi:10.1038/s41467-024-51162-7 2024

  41. [41]

    From quantum feature maps to quantum reservoir computing: an applicative perspective,

    C. Gyurik, F. Wudarski, E. J. Philip, A. Sannia, H. Sadeghi, O. Kyriienko, D. Venturelli, and A. A. Gentile, “From quantum feature maps to quantum reservoir computing: an applicative perspective,” Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, vol. 384, no. 2315, p. 20250085,

  42. [42]

    Available: https://doi.org/10.1098/rsta.2025.0085

    [Online]. Available: https://doi.org/10.1098/rsta.2025.0085

    work page doi:10.1098/rsta.2025.0085 2025

  43. [43]

    Density matrix emulation of quantum recurrent neural networks for multivariate time series prediction,

    J. D. Viqueira, D. Fa ´ılde, M. M. Juane, A. G ´omez, and D. Mera, “Density matrix emulation of quantum recurrent neural networks for multivariate time series prediction,”Machine Learning: Science and Technology, vol. 6, no. 1, p. 015023, 2025

    2025

  44. [44]

    Hybrid quantum-classical recurrent neural networks for time series prediction,

    A. Ceschini, A. Rosato, and M. Panella, “Hybrid quantum-classical recurrent neural networks for time series prediction,” in2022 International Joint Conference on Neural Networks (IJCNN). IEEE, 2022, pp. 1–8. [Online]. Available: https://doi.org/10.1109/IJCNN55064. 2022.9892441

    work page doi:10.1109/ijcnn55064 2022

  45. [45]

    Quantum long short-term memory (QLSTM) vs. classical LSTM in time series forecasting: a comparative study in solar power forecasting,

    S. Z. Khan, N. Muzammil, S. Ghafoor, H. Khan, S. M. H. Zaidi, A. J. Aljohani, and I. Aziz, “Quantum long short-term memory (QLSTM) vs. classical LSTM in time series forecasting: a comparative study in solar power forecasting,”Frontiers in Physics, vol. 12, p. 1439180,

  46. [46]

    Available: https://doi.org/10.3389/fphy.2024.1439180

    [Online]. Available: https://doi.org/10.3389/fphy.2024.1439180

    work page doi:10.3389/fphy.2024.1439180 2024

  47. [47]

    A hybrid quantum- classical model for stock price prediction using quantum-enhanced long short-term memory,

    K. Kea, D. Kim, C. Huot, T.-K. Kim, and Y . Han, “A hybrid quantum- classical model for stock price prediction using quantum-enhanced long short-term memory,”Entropy, vol. 26, no. 11, p. 954, 2024. [Online]. Available: https://doi.org/10.3390/e26110954

    work page doi:10.3390/e26110954 2024

  48. [48]

    Quantum recurrent neural networks: Predicting the dynamics of oscillatory and chaotic systems,

    Y . Chen and A. Khaliq, “Quantum recurrent neural networks: Predicting the dynamics of oscillatory and chaotic systems,”Algorithms, vol. 17, no. 4, p. 163, 2024. [Online]. Available: https://doi.org/10.3390/ a17040163

    2024

  49. [49]

    Liu, Y .-H

    Y .-C. Hsu, J.-C. Jiang, C.-H. Lin, W.-T. Chen, K.-C. Peng, P. Tiwari, S. Y .-C. Chen, and E.-J. Kuo, “Federated quantum kernel-based long short-term memory for human activity recognition,” in2025 IEEE International Conference on Quantum Computing and Engineering (QCE), vol. 2. IEEE, 2025, pp. 54–58. [Online]. Available: https://doi.org/10.1109/QCE65121.2...

    work page doi:10.1109/qce65121.2025.10293 2025

  50. [50]

    QKAN-LSTM: Quantum- inspired Kolmogorov-Arnold long short-term memory,

    Y .-C. Hsu, J.-C. Jiang, C.-H. Lin, K.-C. Peng, N.-Y . Chen, S. Y .-C. Chen, E.-J. Kuo, and H.-S. Goan, “QKAN-LSTM: Quantum- inspired Kolmogorov-Arnold long short-term memory,” 2025. [Online]. Available: https://arxiv.org/abs/2512.05049

    arXiv 2025

  51. [51]

    Quantum variational activation functions empower Kolmogorov-Arnold networks,

    J.-C. Jiang, Y .-C. Huang, T. Chen, and H.-S. Goan, “Quantum variational activation functions empower Kolmogorov-Arnold networks,”arXiv preprint arXiv:2509.14026, 2025. [Online]. Available: https://arxiv.org/ abs/2509.14026

    arXiv 2025

  52. [52]

    Learning to control fast-weight memories: An alternative to dynamic recurrent networks,

    J. Schmidhuber, “Learning to control fast-weight memories: An alternative to dynamic recurrent networks,”Neural Computation, vol. 4, no. 1, pp. 131–139, 1992. [Online]. Available: https: //doi.org/10.1162/neco.1992.4.1.131

    work page doi:10.1162/neco.1992.4.1.131 1992

  53. [53]

    Using fast weights to attend to the recent past,

    J. Ba, G. E. Hinton, V . Mnih, J. Z. Leibo, and C. Ionescu, “Using fast weights to attend to the recent past,” inAdvances in Neural Information Processing Systems, vol. 29, 2016. [Online]. Available: https://proceedings.neurips.cc/paper files/paper/ 2016/hash/9f44e956e3a2b7b5598c625fcc802c36-Abstract.html

    2016

  54. [54]

    Learning to learn by gradient descent by gradient descent,

    M. Andrychowicz, M. Denil, S. G ´omez, M. W. Hoffman, D. Pfau, T. Schaul, B. Shillingford, and N. de Freitas, “Learning to learn by gradient descent by gradient descent,” inAdvances in Neural Information Processing Systems, vol. 29, 2016, pp. 3981–3989. [Online]. Available: https://papers.nips.cc/paper files/paper/2016/hash/ fb87582825f9d28a8d42c5e5e5e8b2...

    2016

  55. [55]

    Optimization as a model for few-shot learning,

    S. Ravi and H. Larochelle, “Optimization as a model for few-shot learning,” inInternational Conference on Learning Representations,

  56. [56]

    Available: https://openreview.net/forum?id=rJY0-Kcll

    [Online]. Available: https://openreview.net/forum?id=rJY0-Kcll

  57. [57]

    Hypernetworks,

    D. Ha, A. M. Dai, and Q. V . Le, “Hypernetworks,” inInternational Conference on Learning Representations, 2017. [Online]. Available: https://openreview.net/forum?id=rkpACe1lx

    2017

  58. [58]

    Meta networks,

    T. Munkhdalai and H. Yu, “Meta networks,” inProceedings of the 34th International Conference on Machine Learning, ser. Proceedings of Machine Learning Research, vol. 70. PMLR, 2017, pp. 2554–2563. [Online]. Available: https://proceedings.mlr.press/v70/munkhdalai17a. html

    2017

  59. [59]

    Learning feed-forward one-shot learners,

    L. Bertinetto, J. F. Henriques, J. Valmadre, P. Torr, and A. Vedaldi, “Learning feed-forward one-shot learners,” inAdvances in Neural Information Processing Systems, vol. 29, 2016, pp. 523–531. [Online]. Available: https://papers.nips.cc/paper files/paper/2016/hash/ 839ab46820b524afda05122893c2fe8e-Abstract.html

    2016

  60. [60]

    Dynamic filter networks,

    X. Jia, B. De Brabandere, T. Tuytelaars, and L. Van Gool, “Dynamic filter networks,” inAdvances in Neural In- formation Processing Systems, vol. 29, 2016. [Online]. Available: https://proceedings.neurips.cc/paper files/paper/2016/hash/ 8bf1211fd4b7b94528899de0a43b9fb3-Abstract.html

    2016

  61. [61]

    Condconv: Conditionally parameterized convolutions for efficient inference,

    B. Yang, G. Bender, Q. V . Le, and J. Ngiam, “Condconv: Conditionally parameterized convolutions for efficient inference,” inAdvances in Neural Information Processing Systems, vol. 32, 2019, pp. 1307–1318. [Online]. Available: https://papers.nips.cc/paper files/paper/2019/hash/ f2201f5191c4e92cc5af043eebfd0946-Abstract.html

    2019

  62. [62]

    Learning to program variational quantum circuits with fast weights,

    S. Y .-C. Chen, “Learning to program variational quantum circuits with fast weights,” in2024 International Joint Conference on Neural Networks (IJCNN), 2024, pp. 1–9

    2024

  63. [63]

    Quantum artificial intelligence: From quantum neural networks to self-programming architectures [feature],

    ——, “Quantum artificial intelligence: From quantum neural networks to self-programming architectures [feature],”IEEE Circuits and Systems Magazine, vol. 26, no. 1, pp. 41–66, 2026

    2026

  64. [64]

    Differentiable quantum architecture search,

    S.-X. Zhang, C.-Y . Hsieh, S. Zhang, and H. Yao, “Differentiable quantum architecture search,”Quantum Science and Technology, vol. 7, no. 4, p. 045023, 2022. [Online]. Available: https://doi.org/10.1088/ 2058-9565/ac87cd

    2022

  65. [65]

    Quantum architecture search via deep reinforcement learning,

    E.-J. Kuo, Y .-L. L. Fang, and S. Y .-C. Chen, “Quantum architecture search via deep reinforcement learning,” 2021. [Online]. Available: https://arxiv.org/abs/2104.07715

    arXiv 2021

  66. [66]

    Quantum circuit architecture search for variational quantum algorithms,

    Y . Du, T. Huang, S. You, M.-H. Hsieh, and D. Tao, “Quantum circuit architecture search for variational quantum algorithms,”npj Quantum Information, vol. 8, p. 62, 2022. [Online]. Available: https://doi.org/10.1038/s41534-022-00570-y

    work page doi:10.1038/s41534-022-00570-y 2022

  67. [67]

    QuantumDARTS: Differentiable quantum architecture search for variational quantum algorithms,

    W. Wu, G. Yan, X. Lu, K. Pan, and J. Yan, “QuantumDARTS: Differentiable quantum architecture search for variational quantum algorithms,” inProceedings of the 40th International Conference on Machine Learning, ser. Proceedings of Machine Learning Research, vol. 202. PMLR, 2023, pp. 37 745–37 764. [Online]. Available: https://proceedings.mlr.press/v202/wu23v.html

    2023

  68. [68]

    In: IEEE International Symposium on High-Performance Computer Archi- tecture, HPCA 2022, Seoul, South Korea, April 2-6, 2022

    H. Wang, Y . Ding, J. Gu, Z. Li, Y . Lin, D. Z. Pan, F. T. Chong, and S. Han, “QuantumNAS: Noise-adaptive search for robust quantum circuits,” in2022 IEEE International Symposium on High-Performance Computer Architecture (HPCA). IEEE, 2022, pp. 692–708. [Online]. Available: https://doi.org/10.1109/HPCA53966.2022.00057

    work page doi:10.1109/hpca53966.2022.00057 2022

  69. [69]

    Martyniuk, J

    S. Y .-C. Chen, “Differentiable quantum architecture search in asynchronous quantum reinforcement learning,” in2024 IEEE International Conference on Quantum Computing and Engineering (QCE), vol. 1, 2024, pp. 1516–1524. [Online]. Available: https://doi.org/10.1109/QCE60285.2024.00178

    work page doi:10.1109/qce60285.2024.00178 2024

  70. [70]

    Reservoir computing via quantum recurrent neural networks,

    S. Y .-C. Chen, D. Fry, A. Deshmukh, V . Rastunkov, and C. Stefanski, “Reservoir computing via quantum recurrent neural networks,” arXiv preprint arXiv:2211.02612, 2022. [Online]. Available: https: //doi.org/10.48550/arXiv.2211.02612

    work page doi:10.48550/arxiv.2211.02612 2022

  71. [71]

    Liu, Y .-H

    S. Y .-C. Chen, C.-Y . Liu, K.-C. Chen, W.-J. Huang, Y .-J. Chang, and W.-H. Huang, “Differentiable quantum architecture search in quantum-enhanced neural network parameter generation,” in2025 IEEE International Conference on Quantum Computing and Engineering (QCE), vol. 1. IEEE, 2025, pp. 1525–1535. [Online]. Available: https://doi.org/10.1109/QCE65121.2...

    work page doi:10.1109/qce65121.2025.00170 2025