pith. machine review for the scientific record. sign in

arxiv: 2605.13833 · v1 · submitted 2026-05-13 · 💻 cs.LG · cs.CV

Recognition: unknown

QLAM: A Quantum Long-Attention Memory Approach to Long-Sequence Token Modeling

Hoang-Quan Nguyen , Sankalp Pandey , Khoa Luu
Authors on Pith no claims yet

Pith reviewed 2026-05-14 19:07 UTC · model grok-4.3

classification 💻 cs.LG cs.CV
keywords quantum machine learningstate-space modelssequence modelinglong-range dependencieshybrid quantum-classicalsuperposition statesrecurrent models
0
0 comments X

The pith

QLAM represents sequence memory as a quantum superposition state evolved by input-conditioned circuits to capture global dependencies in linear time.

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

The paper introduces QLAM to address long-range dependencies in sequential data by extending state-space models with quantum mechanics. Instead of classical additive updates to a hidden state, it maintains a quantum state whose amplitudes encode superposed historical token information and evolves that state through parameterized circuits. This preserves the linear-time recurrent structure of SSMs while allowing non-classical global interactions that attention mechanisms compute explicitly. Evaluation on flattened sequential image tasks shows consistent gains over recurrent and transformer baselines.

Core claim

QLAM maintains the hidden state as a quantum state whose amplitudes encode a superposition of historical information and evolves the state through parameterized quantum circuits conditioned on each input token. This yields a non-classical global update mechanism that captures complex dependencies implicitly, retrieves relevant information via query-dependent measurements, and retains the linear-time recurrent computation of state-space models.

What carries the argument

Parameterized quantum circuits that evolve a quantum superposition state representing the superposition of all prior token information.

If this is right

  • Delivers higher accuracy than recurrent baselines and transformers on sequential image classification while using linear computation.
  • Implicitly models global token dependencies through quantum-state evolution rather than explicit pairwise attention.
  • Preserves the recurrent structure and linear scaling of state-space models.
  • Retrieves task-relevant information from the quantum memory via input-dependent measurements.

Where Pith is reading between the lines

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

  • The same quantum-state memory could be tested on language-modeling sequences where context length exceeds typical transformer limits.
  • Classical simulation cost grows with qubit count, so any advantage will depend on whether the circuit depth and width remain feasible for target sequence lengths.
  • Different ansatz circuits or measurement strategies could be swapped in to target specific data modalities without changing the overall recurrent framework.

Load-bearing premise

That the parameterized quantum circuits can evolve the superposition state to capture complex global token interactions more effectively than classical additive or linear transitions at practical simulation cost.

What would settle it

A side-by-side run on sMNIST, sFashion-MNIST or sCIFAR-10 in which a classical linear state update matches or exceeds the accuracy of the quantum-circuit version.

Figures

Figures reproduced from arXiv: 2605.13833 by Hoang-Quan Nguyen, Khoa Luu, Sankalp Pandey.

Figure 1
Figure 1. Figure 1: Overview framework of the proposed Quantum Long-Attention Memory. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The quantum long-attention memory is evolved across input features. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Computation flow of quantum long-attention readout. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The observable O(qt) is formed as a weighted combination of multiple Pauli matrices based on the query qt. previous token s < t, we compute an attention through a measurement operator: αt,s = ⟨ψt|O(qt)|ψt⟩, (13) where O(qt) ∈ C 2 n×2 n is a learnable Hermitian observable operator parameterized by the query qt. In detail, given a set of Pauli matrices P = {Oi} p i=1, we project the query qt via a small neur… view at source ↗
Figure 5
Figure 5. Figure 5: Visualization of sample inputs from the benchmark datasets used in our experiments, including (a) MNIST [57], (b) Fashion-MNIST [58], and (c) [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Evaluation protocol for sequential modeling with images. [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗
read the original abstract

Modeling long-range dependencies in sequential data remains a central challenge in machine learning. Transformers address this challenge through attention mechanisms, but their quadratic complexity with respect to sequence length limits scalability to long contexts. State-space models (SSMs) provide an efficient alternative with linear-time computation by evolving a latent state through recurrent updates, but their memory is typically formed via additive or linear transitions, which can limit their ability to capture complex global interactions across tokens. In this work, we introduce one of the first studies to leverage the superposition property of quantum systems to enhance state-based sequence modeling. In particular, we propose Quantum Long-Attention Memory (QLAM), a hybrid quantum-classical memory mechanism that can be viewed as a quantum extension of state-space models. Instead of maintaining a classical latent state updated through additive dynamics, QLAM represents the hidden state as a quantum state whose amplitudes encode a superposition of historical information. The state evolves through parameterized quantum circuits conditioned on the input, enabling a non-classical, globally update mechanism. In this way, QLAM preserves the recurrent and linear-time structure of SSMs while fundamentally enriching the memory representation through quantum superposition. Unlike attention mechanisms that explicitly compute pairwise interactions, QLAM implicitly captures global dependencies through the evolution of the quantum state, and retrieves task-relevant information via query-dependent measurements. We evaluate QLAM on sequential variants of standard image classification benchmarks, including sMNIST, sFashion-MNIST, and sCIFAR-10, where images are flattened into token sequences. Across all tasks, QLAM consistently improves over recurrent baselines and transformer-based models.

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

3 major / 2 minor

Summary. The manuscript introduces Quantum Long-Attention Memory (QLAM) as a hybrid quantum-classical extension of state-space models for long-sequence token modeling. It replaces classical additive latent-state updates with a quantum state whose amplitudes encode a superposition of historical tokens; this state evolves via input-conditioned parameterized quantum circuits, enabling implicit global dependency capture through non-classical dynamics while retaining the recurrent linear-time structure of SSMs. Retrieval occurs via query-dependent measurements. The model is evaluated on flattened sequential image benchmarks (sMNIST, sFashion-MNIST, sCIFAR-10) and is claimed to outperform recurrent baselines and transformer models across all tasks.

Significance. If the quantum-superposition mechanism can be realized with practical linear scaling, the work would offer a novel route to enriching SSM memory representations beyond linear or additive transitions, potentially improving long-range dependency modeling without quadratic attention cost. The hybrid framing and emphasis on preserving recurrence are strengths; however, the absence of any circuit specification, qubit count, or simulation method prevents evaluation of whether the claimed non-classical advantage is achievable or merely an artifact of extra parameters.

major comments (3)
  1. [Abstract / Method] Abstract and method description: the assertion that QLAM 'preserves the recurrent and linear-time structure of SSMs' while using parameterized quantum circuits is unsupported, because no qubit number, circuit depth, entanglement structure, or efficient simulation technique (e.g., matrix-product states or tensor networks) is specified. General quantum-circuit simulation is exponential in qubit count, which directly contradicts the linear-time claim for sequence lengths such as 1024 on sCIFAR-10.
  2. [Experiments] Experiments section: the central empirical claim that 'across all tasks, QLAM consistently improves over recurrent baselines and transformer-based models' is presented without any numerical metrics, error bars, tables, or ablation studies. This absence makes it impossible to verify whether reported gains arise from quantum superposition or from additional classical parameters.
  3. [Method] Method description: the weakest assumption—that parameterized quantum circuits evolve the superposition state to capture complex global token interactions more effectively than classical additive or linear transitions—is stated but never tested. No derivation, complexity analysis, or controlled comparison isolating the quantum component is provided.
minor comments (2)
  1. [Introduction] The phrase 'one of the first studies' would benefit from explicit citations to prior quantum sequence-modeling or quantum-SSM literature to clarify novelty.
  2. [Method] Notation for the quantum state and measurement operators is introduced only descriptively; explicit equations would improve reproducibility.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. We value the recognition of QLAM's potential to enrich SSM memory via quantum superposition and agree that greater specificity on implementation, empirical reporting, and validation is required. We address each major comment below and indicate the corresponding revisions.

read point-by-point responses

  1. Referee: [Abstract / Method] Abstract and method description: the assertion that QLAM 'preserves the recurrent and linear-time structure of SSMs' while using parameterized quantum circuits is unsupported, because no qubit number, circuit depth, entanglement structure, or efficient simulation technique (e.g., matrix-product states or tensor networks) is specified. General quantum-circuit simulation is exponential in qubit count, which directly contradicts the linear-time claim for sequence lengths such as 1024 on sCIFAR-10.

    Authors: We acknowledge the need for explicit implementation details. In the revised manuscript we specify a fixed qubit register of 6 qubits whose amplitudes encode a superposition over historical tokens. Each recurrent step applies a constant-depth parameterized circuit (alternating RY rotations and CZ entangling gates whose angles are linear functions of the current input token). Because the qubit count is independent of sequence length, the state can be simulated classically with matrix-product-state tensor networks whose bond dimension remains modest for the low-entanglement regimes encountered in these tasks, yielding overall linear scaling in sequence length. A new subsection 'Quantum Implementation and Complexity' now provides the qubit count, gate set, and tensor-network simulation argument that supports the linear-time claim. revision: yes

  2. Referee: [Experiments] Experiments section: the central empirical claim that 'across all tasks, QLAM consistently improves over recurrent baselines and transformer-based models' is presented without any numerical metrics, error bars, tables, or ablation studies. This absence makes it impossible to verify whether reported gains arise from quantum superposition or from additional classical parameters.

    Authors: We agree that the original presentation omitted quantitative detail. The revised experiments section now includes a table reporting mean accuracy and standard deviation over five independent runs on sMNIST, sFashion-MNIST, and sCIFAR-10, together with direct comparisons against LSTM, GRU, and Transformer baselines of matched parameter count. Additional ablation columns isolate the contribution of the quantum superposition by replacing the quantum circuit with a classical linear or additive update of identical dimensionality; the performance gap remains statistically significant, indicating that the observed gains are not explained by parameter count alone. revision: yes

  3. Referee: [Method] Method description: the weakest assumption—that parameterized quantum circuits evolve the superposition state to capture complex global token interactions more effectively than classical additive or linear transitions—is stated but never tested. No derivation, complexity analysis, or controlled comparison isolating the quantum component is provided.

    Authors: We have strengthened the method section with an explicit derivation: the unitary evolution on the superposition state induces amplitude interference that realizes a non-linear mixing of all prior tokens within a single recurrent step, an operation outside the span of classical linear or additive state updates. We supply a formal complexity argument showing that, under the tensor-network simulation described above, each step remains O(1) with respect to sequence length. Finally, the revised experiments contain a controlled comparison that replaces the quantum circuit with a classical feed-forward layer of equivalent expressivity; the quantum variant retains a consistent advantage on long-range dependency probes, supporting the modeling claim. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected in derivation chain

full rationale

The paper presents QLAM as an independent architectural proposal: a hybrid quantum-classical extension of state-space models that replaces additive latent-state transitions with evolution of a quantum superposition state via parameterized circuits. No equations, parameter-fitting procedures, or self-citations appear in the provided text that reduce the claimed linear-time global-interaction advantage to a tautological redefinition of inputs, a fitted quantity renamed as prediction, or a load-bearing uniqueness theorem imported from the authors' prior work. The central claims rest on the explicit design choice of quantum-state representation and measurement-based retrieval rather than on any self-referential reduction. This is the most common honest outcome for a modeling paper that introduces a new mechanism without deriving its performance from its own fitted parameters or self-citations.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 1 invented entities

The central claim rests on standard quantum mechanics for superposition and measurement, plus trainable parameters inside the quantum circuits that must be fitted to data; no independent evidence for the new memory entity is supplied beyond the reported benchmark gains.

free parameters (1)
  • quantum circuit parameters
    Parameterized quantum circuits contain trainable angles or gates whose values are fitted during learning to achieve the reported performance.
axioms (1)
  • domain assumption Quantum superposition and unitary evolution hold for the hidden state representation
    The model assumes standard quantum mechanics applies to the latent state without decoherence or hardware noise in the abstract description.
invented entities (1)
  • Quantum Long-Attention Memory state no independent evidence
    purpose: To encode a superposition of historical token information that evolves globally via quantum circuits
    A new postulated memory representation introduced to replace classical additive states; no external falsifiable prediction is given in the abstract.

pith-pipeline@v0.9.0 · 5591 in / 1337 out tokens · 66909 ms · 2026-05-14T19:07:13.496227+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

62 extracted references · 15 canonical work pages · 10 internal anchors

  1. [1]

    Long short-term memory,

    S. Hochreiter and J. Schmidhuber, “Long short-term memory,”Neural computation, vol. 9, no. 8, pp. 1735–1780, 1997

    1997

  2. [2]

    Learning long-term dependencies with gradient descent is difficult,

    Y . Bengio, P. Simard, and P. Frasconi, “Learning long-term dependencies with gradient descent is difficult,”IEEE transactions on neural networks, vol. 5, no. 2, pp. 157–166, 1994

    1994

  3. [3]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,”Advances in neural information processing systems, vol. 30, 2017

    2017

  4. [4]

    Generating Long Sequences with Sparse Transformers

    R. Child, S. Gray, A. Radford, and I. Sutskever, “Generating long sequences with sparse transformers,”arXiv preprint arXiv:1904.10509, 2019

    work page internal anchor Pith review Pith/arXiv arXiv 1904

  5. [5]

    Transformers are rnns: Fast autoregressive transformers with linear attention,

    A. Katharopoulos, A. Vyas, N. Pappas, and F. Fleuret, “Transformers are rnns: Fast autoregressive transformers with linear attention,” in International conference on machine learning. PMLR, 2020, pp. 5156– 5165

    2020

  6. [6]

    Rethinking Attention with Performers

    K. Choromanski, V . Likhosherstov, D. Dohan, X. Song, A. Gane, T. Sar- los, P. Hawkins, J. Davis, A. Mohiuddin, L. Kaiseret al., “Rethinking attention with performers,”arXiv preprint arXiv:2009.14794, 2020

    work page internal anchor Pith review Pith/arXiv arXiv 2009

  7. [7]

    Efficient transformers: A survey,

    Y . Tay, M. Dehghani, D. Bahri, and D. Metzler, “Efficient transformers: A survey,”ACM Computing Surveys, vol. 55, no. 6, pp. 1–28, 2022

    2022

  8. [8]

    Flashattention: Fast and memory-efficient exact attention with io-awareness,

    T. Dao, D. Fu, S. Ermon, A. Rudra, and C. R ´e, “Flashattention: Fast and memory-efficient exact attention with io-awareness,”Advances in neural information processing systems, vol. 35, pp. 16 344–16 359, 2022

    2022

  9. [9]

    Efficiently Modeling Long Sequences with Structured State Spaces

    A. Gu, K. Goel, and C. R ´e, “Efficiently modeling long sequences with structured state spaces,”arXiv preprint arXiv:2111.00396, 2021

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  10. [10]

    Combining recurrent, convolutional, and continuous-time models with linear state space layers,

    A. Gu, I. Johnson, K. Goel, K. Saab, T. Dao, A. Rudra, and C. R ´e, “Combining recurrent, convolutional, and continuous-time models with linear state space layers,”Advances in neural information processing systems, vol. 34, pp. 572–585, 2021

    2021

  11. [11]

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

    2010

  12. [12]

    Quantum computing in the nisq era and beyond,

    J. Preskill, “Quantum computing in the nisq era and beyond,”Quantum, vol. 2, p. 79, 2018

    2018

  13. [13]

    Big bird: Transformers for longer sequences,

    M. Zaheer, G. Guruganesh, K. A. Dubey, J. Ainslie, C. Alberti, S. Ontanon, P. Pham, A. Ravula, Q. Wang, L. Yanget al., “Big bird: Transformers for longer sequences,”Advances in neural information processing systems, vol. 33, pp. 17 283–17 297, 2020

    2020

  14. [14]

    Learning internal representations by error propagation,

    D. E. Rumelhart, G. E. Hinton, and R. J. Williams, “Learning internal representations by error propagation,” Tech. Rep., 1985

    1985

  15. [15]

    Finding structure in time,

    J. L. Elman, “Finding structure in time,”Cognitive science, vol. 14, no. 2, pp. 179–211, 1990

    1990

  16. [16]

    Learning phrase representations using rnn encoder–decoder for statistical machine translation,

    K. Cho, B. Van Merri ¨enboer, C ¸ . Gulc ¸ehre, D. Bahdanau, F. Bougares, H. Schwenk, and Y . Bengio, “Learning phrase representations using rnn encoder–decoder for statistical machine translation,” inProceedings of the 2014 conference on empirical methods in natural language processing (EMNLP), 2014, pp. 1724–1734

    2014

  17. [17]

    Bert: Pre-training of deep bidirectional transformers for language understanding,

    J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, “Bert: Pre-training of deep bidirectional transformers for language understanding,” inPro- ceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human language technologies, volume 1 (long and short papers), 2019, pp. 4171–4186

    2019

  18. [18]

    Language mod- els are few-shot learners,

    T. Brown, B. Mann, N. Ryder, M. Subbiah, J. D. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askellet al., “Language mod- els are few-shot learners,”Advances in neural information processing systems, vol. 33, pp. 1877–1901, 2020

    1901

  19. [19]

    Language models are unsupervised multitask learners,

    A. Radford, J. Wu, R. Child, D. Luan, D. Amodei, I. Sutskeveret al., “Language models are unsupervised multitask learners,”OpenAI blog, vol. 1, no. 8, p. 9, 2019

    2019

  20. [20]

    An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gellyet al., “An image is worth 16x16 words: Transformers for image recognition at scale,”arXiv preprint arXiv:2010.11929, 2020

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  21. [21]

    Swin transformer: Hierarchical vision transformer using shifted windows,

    Z. Liu, Y . Lin, Y . Cao, H. Hu, Y . Wei, Z. Zhang, S. Lin, and B. Guo, “Swin transformer: Hierarchical vision transformer using shifted windows,” inProceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 10 012–10 022

    2021

  22. [22]

    Simpli- fied state space layers for sequence modeling,

    J. T. Smith, A. Warrington, and S. W. Linderman, “Simplified state space layers for sequence modeling,”arXiv preprint arXiv:2208.04933, 2022

    work page arXiv 2022

  23. [23]

    Mamba: Linear-Time Sequence Modeling with Selective State Spaces

    A. Gu and T. Dao, “Mamba: Linear-time sequence modeling with selective state spaces,”arXiv preprint arXiv:2312.00752, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  24. [24]

    Quantum machine learning,

    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

    2017

  25. [25]

    An introduction to quantum machine learning,

    M. Schuld, I. Sinayskiy, and F. Petruccione, “An introduction to quantum machine learning,”Contemporary Physics, vol. 56, no. 2, pp. 172–185, 2015

    2015

  26. [26]

    Quantum machine learning in feature hilbert spaces,

    M. Schuld and N. Killoran, “Quantum machine learning in feature hilbert spaces,”Physical review letters, vol. 122, no. 4, p. 040504, 2019

    2019

  27. [27]

    Quantum algorithms for supervised and unsupervised machine learning

    S. Lloyd, M. Mohseni, and P. Rebentrost, “Quantum algorithms for supervised and unsupervised machine learning,”arXiv preprint arXiv:1307.0411, 2013

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  28. [28]

    Quantum vision clustering,

    X. B. Nguyen, H. Churchill, K. Luu, and S. U. Khan, “Quantum vision clustering,”arXiv preprint arXiv:2309.09907, 2023

    work page arXiv 2023

  29. [29]

    Qclusformer: A quantum transformer-based framework for unsupervised visual clustering,

    X.-B. Nguyen, H.-Q. Nguyen, S. Y .-C. Chen, S. U. Khan, H. Churchill, and K. Luu, “Qclusformer: A quantum transformer-based framework for unsupervised visual clustering,” in2024 IEEE International Conference on Quantum Computing and Engineering (QCE), vol. 2. IEEE, 2024, pp. 347–352

    2024

  30. [30]

    Quantum principal compo- nent analysis,

    S. Lloyd, M. Mohseni, and P. Rebentrost, “Quantum principal compo- nent analysis,”Nature Physics, vol. 10, no. 9, pp. 631–633, 2014

    2014

  31. [31]

    Prediction by linear regression on a quantum computer,

    M. Schuld, I. Sinayskiy, and F. Petruccione, “Prediction by linear regression on a quantum computer,”Physical Review A, vol. 94, no. 2, p. 022342, 2016

    2016

  32. [32]

    Quantum gradient descent for linear systems and least squares,

    I. Kerenidis and A. Prakash, “Quantum gradient descent for linear systems and least squares,”Physical Review A, vol. 101, no. 2, p. 022316, 2020

    2020

  33. [33]

    Quantum support vector machine for big data classification,

    P. Rebentrost, M. Mohseni, and S. Lloyd, “Quantum support vector machine for big data classification,”Physical review letters, vol. 113, no. 13, p. 130503, 2014

    2014

  34. [34]

    Variational quantum algorithms,

    M. Cerezo, A. Arrasmith, R. Babbush, S. C. Benjamin, S. Endo, K. Fujii, J. R. McClean, K. Mitarai, X. Yuan, L. Cincioet al., “Variational quantum algorithms,”Nature Reviews Physics, vol. 3, no. 9, pp. 625– 644, 2021

    2021

  35. [35]

    Parameterized quantum circuits as machine learning models,

    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

    2019

  36. [36]

    Circuit-centric quantum classifiers,

    M. Schuld, A. Bocharov, K. M. Svore, and N. Wiebe, “Circuit-centric quantum classifiers,”Physical Review A, vol. 101, no. 3, p. 032308, 2020

    2020

  37. [37]

    Hierarchical quantum control gates for functional mri understanding,

    X.-B. Nguyen, H.-Q. Nguyen, H. Churchill, S. U. Khan, and K. Luu, “Hierarchical quantum control gates for functional mri understanding,” in2024 IEEE Workshop on Signal Processing Systems (SiPS). IEEE, 2024, pp. 159–164

    2024

  38. [38]

    Qmoe: A quantum mixture of experts framework for scalable quantum neural networks,

    H.-Q. Nguyen, X.-B. Nguyen, S. Pandey, S. U. Khan, I. Safro, and K. Luu, “Qmoe: A quantum mixture of experts framework for scalable quantum neural networks,” in2025 IEEE International Conference on Quantum Computing and Engineering (QCE), vol. 2. IEEE, 2025, pp. 223–228

    2025

  39. [39]

    A Quantum Approximate Optimization Algorithm

    E. Farhi, J. Goldstone, and S. Gutmann, “A quantum approximate optimization algorithm,”arXiv preprint arXiv:1411.4028, 2014

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  40. [40]

    Quantum approximate optimization algorithm: Performance, mechanism, and im- plementation on near-term devices,

    L. Zhou, S.-T. Wang, S. Choi, H. Pichler, and M. D. Lukin, “Quantum approximate optimization algorithm: Performance, mechanism, and im- plementation on near-term devices,”Physical Review X, vol. 10, no. 2, p. 021067, 2020

    2020

  41. [41]

    Quadro: A hybrid quantum optimization framework for drone delivery,

    J. B. Holliday, D. Blount, H. Q. Nguyen, S. U. Khan, and K. Luu, “Quadro: A hybrid quantum optimization framework for drone delivery,” in2025 IEEE International Conference on Quantum Computing and Engineering (QCE), vol. 1. IEEE, 2025, pp. 2090–2100

    2025

  42. [42]

    A generative modeling approach for benchmark- ing and training shallow quantum circuits,

    M. Benedetti, D. Garcia-Pintos, O. Perdomo, V . Leyton-Ortega, Y . Nam, and A. Perdomo-Ortiz, “A generative modeling approach for benchmark- ing and training shallow quantum circuits,”npj Quantum information, vol. 5, no. 1, p. 45, 2019

    2019

  43. [43]

    Experimental quantum generative adversarial networks for image generation,

    H.-L. Huang, Y . Du, M. Gong, Y . Zhao, Y . Wu, C. Wang, S. Li, F. Liang, J. Lin, Y . Xuet al., “Experimental quantum generative adversarial networks for image generation,”Physical Review Applied, vol. 16, no. 2, p. 024051, 2021

    2021

  44. [44]

    Diffusion-inspired quantum noise mitigation in parameterized quantum circuits,

    H.-Q. Nguyen, X. B. Nguyen, S. Y .-C. Chen, H. Churchill, N. Borys, S. U. Khan, and K. Luu, “Diffusion-inspired quantum noise mitigation in parameterized quantum circuits,”Quantum Machine Intelligence, vol. 7, no. 1, p. 55, 2025

    2025

  45. [45]

    Variational quantum reinforcement learning via evolutionary optimiza- tion,

    S. Y .-C. Chen, C.-M. Huang, C.-W. Hsing, H.-S. Goan, and Y .-J. Kao, “Variational quantum reinforcement learning via evolutionary optimiza- tion,”Machine Learning: Science and Technology, vol. 3, no. 1, p. 015025, 2022

    2022

  46. [46]

    Quantum convolutional neural networks,

    I. Cong, S. Choi, and M. D. Lukin, “Quantum convolutional neural networks,”Nature Physics, vol. 15, no. 12, pp. 1273–1278, 2019

    2019

  47. [47]

    Quantum autoencoders for efficient compression of quantum data,

    J. Romero, J. P. Olson, and A. Aspuru-Guzik, “Quantum autoencoders for efficient compression of quantum data,”Quantum Science and Technology, vol. 2, no. 4, p. 045001, 2017

    2017

  48. [48]

    Neural networks with quantum architec- ture and quantum learning,

    M. Panella and G. Martinelli, “Neural networks with quantum architec- ture and quantum learning,”International Journal of Circuit Theory and Applications, vol. 39, no. 1, pp. 61–77, 2011

    2011

  49. [49]

    Quantum circuit learning,

    K. Mitarai, M. Negoro, M. Kitagawa, and K. Fujii, “Quantum circuit learning,”Physical Review A, vol. 98, no. 3, p. 032309, 2018

    2018

  50. [50]

    Quantum visual feature encoding revisited,

    X.-B. Nguyen, H.-Q. Nguyen, H. Churchill, S. U. Khan, and K. Luu, “Quantum visual feature encoding revisited,”Quantum Machine Intelli- gence, vol. 6, no. 2, p. 61, 2024

    2024

  51. [51]

    Quantum-brain: Quantum-inspired neu- ral network approach to vision-brain understanding,

    H.-Q. Nguyen, X.-B. Nguyen, H. Churchill, A. K. Choudhary, P. Sinha, S. U. Khan, and K. Luu, “Quantum-brain: Quantum-inspired neu- ral network approach to vision-brain understanding,”arXiv preprint arXiv:2411.13378, 2024

    work page arXiv 2024

  52. [52]

    Phi-adapt: A physics-informed adaptation learning approach to 2d quantum material discovery,

    H.-Q. Nguyen, X. B. Nguyen, S. Pandey, T. Faltermeier, N. Borys, H. Churchill, and K. Luu, “Phi-adapt: A physics-informed adaptation learning approach to 2d quantum material discovery,”arXiv preprint arXiv:2507.05184, 2025

    work page arXiv 2025

  53. [53]

    Openqlaw: An agentic ai assistant for analysis of 2d quantum materials,

    S. Pandey, X.-B. Nguyen, H.-Q. Nguyen, T. Faltermeier, N. Borys, H. Churchill, and K. Luu, “Openqlaw: An agentic ai assistant for analysis of 2d quantum materials,”arXiv preprint arXiv:2603.17043, 2026

    work page arXiv 2026

  54. [54]

    Qupaint: Physics-aware instruction tuning ap- proach to quantum material discovery,

    X.-B. Nguyen, H.-Q. Nguyen, S. Pandey, T. Faltermeier, N. Borys, H. Churchill, and K. Luu, “Qupaint: Physics-aware instruction tuning ap- proach to quantum material discovery,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2026

    2026

  55. [55]

    Training deep quantum neural networks,

    K. Beer, D. Bondarenko, T. Farrelly, T. J. Osborne, R. Salzmann, D. Scheiermann, and R. Wolf, “Training deep quantum neural networks,” Nature communications, vol. 11, no. 1, p. 808, 2020

    2020

  56. [56]

    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

    2023

  57. [57]

    Mnist handwritten digit database,

    Y . LeCun, C. Cortes, and C. Burges, “Mnist handwritten digit database,” ATT Labs [Online]. Available: http://yann.lecun.com/exdb/mnist, vol. 2, 2010

    2010

  58. [58]

    Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms

    H. Xiao, K. Rasul, and R. V ollgraf, “Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms,”arXiv preprint arXiv:1708.07747, 2017

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  59. [59]

    Learning multiple layers of features from tiny images,

    A. Krizhevsky, G. Hintonet al., “Learning multiple layers of features from tiny images,” 2009

    2009

  60. [60]

    Pytorch: An imperative style, high-performance deep learning library,

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

    2019

  61. [61]

    PennyLane: Automatic differentiation of hybrid quantum-classical computations

    V . Bergholm, J. Izaac, M. Schuld, C. Gogolin, S. Ahmed, V . Ajith, M. S. Alam, G. Alonso-Linaje, B. AkashNarayanan, A. Asadiet al., “Pennylane: Automatic differentiation of hybrid quantum-classical com- putations,”arXiv preprint arXiv:1811.04968, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  62. [62]

    Adam: A Method for Stochastic Optimization

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

    work page internal anchor Pith review Pith/arXiv arXiv 2014