Pith. sign in

REVIEW 2 major objections 4 minor 82 references

This paper shows that measuring a quantum reservoir can supply the dissipation and non-unitality needed for temporal computing, even when the unmonitored dynamics is unitary — and proves a criterion for when that works.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-02 02:29 UTC pith:L5D4PO66

load-bearing objection The unification and the contractivity criterion are real contributions, but the blanket claim that AD monitoring gives input separability for any input-dependent unitary is false and should be fixed before publication. the 2 major comments →

arxiv 2607.14311 v1 pith:L5D4PO66 submitted 2026-07-15 quant-ph

General theory of monitored Quantum Reservoir Computing

Oriol Morgu\'i-Sancho , Gian Luca Giorgi , Gonzalo Manzano , Roberta Zambrini (Instituto de F\'isica Interdisciplinar y Sistemas Complejos (IFISC) , UIB-CSIC) This is my paper
classification quant-ph
keywords quantum reservoir computingmeasurement back-actiononline monitoringecho-state propertyfading memoryinput separabilitystrict contractivityindirect measurements
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper sets out to show that measurement back-action is not merely a disturbance in online quantum reservoir computing: it can be engineered as a resource. The central claim is that monitored dynamics can acquire the echo-state property, fading memory, and input separability even when the unmonitored evolution is unitary, and therefore unsuitable for in-memory processing. To establish this, the authors build a unified framework based on indirect measurements — projective, weak, partial, and dissipative — and derive a necessary and sufficient condition under which the composition of two non-contractive maps becomes strictly contractive. If correct, this gives a systematic route to designing quantum reservoirs by engineering measurements, and explains why non-unital monitoring schemes such as amplitude-damping or partial measurement with reset can render unitary reservoirs viable while unital ones cannot.

Core claim

The core discovery is that the non-selective, ensemble-averaged map T(s) = M ∘ Λ of a monitored reservoir can satisfy the three requirements for reservoir computing — echo-state property, fading memory, and input separability — even when the unmonitored map Λ is unitary. The mathematical engine is Theorem 1: the composition T2 ∘ T1 of two CPTP maps is strictly contractive if and only if T1(F(T1)) ∩ F(T2) = ∅, where F(T) is the set of state pairs whose trace distance T preserves. This identifies exactly when repeated monitoring creates contractivity that neither map possesses alone. The paper then classifies monitoring schemes: unital protocols such as dephasing or partial projective measurem

What carries the argument

The central object is the non-selective measurement map M, obtained by averaging over all outcomes, composed with the unmonitored dynamics Λ; viability of the reservoir is assessed through the contraction coefficient κ_max and the non-unitality of T = M ∘ Λ. The load-bearing theorem is the disjointness criterion: two CPTP maps compose to a strictly contractive map if and only if the image under the first map of its saturating set F(T1) — the state pairs whose trace distance is preserved — is disjoint from the second map's saturating set F(T2). This criterion is what lets contractivity 'emerge' from repeated monitoring, and it is used to prove Corollaries 1 and 2 that separate echo-state beha

Load-bearing premise

The framework equates a monitored reservoir's computational power with the ensemble-averaged, non-selective map T = M ∘ Λ and its contraction and unitality properties; if information carried by individual measurement records or by non-Markovian probe memory contributes to computation, the classification based on this averaged map would be incomplete.

What would settle it

Run a monitored unitary reservoir with dephasing-only (unital) monitoring on a task that requires distinguishing input histories beyond the fading-memory time. The paper predicts the asymptotic state is independent of input, so performance on such a task should saturate at zero. Measurable long-range input separability under unital monitoring would refute the central claim; alternatively, finding a single trajectory, record-dependent processing that succeeds where the averaged map fails would show the averaged-map criterion is insufficient.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Unitary reservoirs, otherwise excluded from in-memory QRC, become viable when monitored with non-unital schemes such as amplitude-damping measurement or partial measurement with reset.
  • Unital monitoring (dephasing, partial projective) can enforce the echo-state property but leaves an input-independent maximally mixed fixed point, so it cannot separate inputs — a clear design rule.
  • Measurement strength tunes a trade-off: strong back-action improves short-term memory and shot efficiency, weak back-action preserves long-term memory, and the optimal choice depends on the task and the number of experimental shots.
  • Time-multiplexing in monitored QRC creates an optimal number of virtual nodes: more measurements extract more features but eventually freeze the dynamics via the quantum Zeno effect.
  • The framework unifies previously separate proposals, so performance comparisons can be made under a common reference dynamics rather than protocol by protocol.

Where Pith is reading between the lines

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

  • Although the paper analyzes the ensemble-averaged map, individual measurement records may carry extra information; a trajectory-conditioned readout could in principle exceed the averaged-map prediction.
  • Theorem 1 is stated for CPTP maps generally; if it holds as broadly as it appears, it provides a general mechanism for emergent dissipation in any repeated concatenation of non-contractive quantum channels, not just QRC reservoirs.
  • The classification criterion could be tested experimentally by comparing dephasing-monitored versus amplitude-damping-monitored unitary reservoirs on a task with long-range temporal correlations: the former should saturate at the maximally mixed fixed point, the latter should not.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper develops a general framework for online monitored quantum reservoir computing, unifying projective, weak, amplitude-damping, and partial measurements through indirect-measurement theory. The central claim is that measurement back-action can supply the dissipation and non-unital dynamics needed for QRC even when the unmonitored evolution is unitary. The mathematical core is Theorem 1 (Appendix C), a necessary and sufficient condition for the composition of two CPTP maps to be strictly contractive, used to establish the echo-state property and fading memory for monitored unitary dynamics. The authors benchmark the framework on STM and NARMA tasks for Ising and Haar-random reservoirs, study the effect of finite measurement shots, and analyze time-multiplexing. The paper also claims to derive general criteria for input separability, and asserts in Sec. IIIC that amplitude-damping monitoring makes any input-dependent unitary satisfy the ESP and input separability.

Significance. If the claims are correct, the paper would be a valuable unification of a fragmented literature, providing a principled criterion for when monitoring can replace engineered dissipation and a systematic comparison of measurement protocols. The proof of Theorem 1 in Appendix C is clean and appears sound, and the benchmarks are standard, reproducible in principle, and include error bars. The uncertainty formulas in Appendix A are internally consistent. However, the paper's strongest general claim about input separability is overstated, and the missing sufficient condition for input separability is a load-bearing gap, because input separability is one of the three properties advertised in the abstract.

major comments (2)
  1. [Sec. IIIC, text near Eq. (22) and Fig. 4] The assertion that AD-monitored unitary dynamics 'satisfies the ESP and input separability for any input-dependent unitary' is false. For a single qubit, take U(s)=diag(e^{iφ(s)}, e^{-iφ(s)}), which leaves the AD fixed point |0><0| invariant. Since the amplitude-damping channel has unique fixed point |0><0|, the composed map D_AD∘Λ(s) has the same fixed point |0><0| for every s; all input histories converge to the same state and input separability fails. More generally, any unitary that maps the AD fixed point into a state from which the AD channel returns to that fixed point will produce an input-independent asymptotic state. The paper itself notes (Sec. IIIA) that non-unitality is not sufficient for input separability, using the fully damped channel as a counterexample; the same issue appears here. This is not a cosmetic point: the abstract promises general criteria for input separabil
  2. [Secs. IIIA–IIIB, and Sec. IIIC] The paper does not actually derive a sufficient condition for input separability; it only gives non-unitality as necessary and not sufficient, and then asserts separability for the AD-monitored unitary map. The necessary-and-sufficient criterion in Theorem 1 concerns strict contractivity (ESP/FMP), not separability. For the claim of a 'general theory' and 'general criteria' in the abstract, a rigorous condition under which T(s)=D_AD∘Λ(s) has an input-dependent fixed point—and why that implies distinguishability of distinct input histories in the long-time limit—is required. Without it, the classification in Table I and the central conclusion that AD monitoring makes unitary dynamics viable are only supported for the specific numerically tested unitaries. The authors should either prove a sufficient condition (e.g., based on the unitary not permuting the AD fixed-point subspace) or explic
minor comments (4)
  1. [Abstract and Sec. IIIC] The abstract says 'general criteria ... input separability' but the body only provides a necessary condition. After correcting the major issue, align the abstract with what is actually proven.
  2. [Sec. IIIE, Eq. (27)] The finite-shot noise model is an approximation: Eq. (27) models shot noise as additive Gaussian with state-independent variance. This is a practical simplification, but the paper should state more explicitly that trajectory-to-trajectory correlations and the full measurement-record statistics are not modeled, and that the stated 'general theory' applies to the ensemble-averaged dynamics in the infinite-shot limit.
  3. [Table I] The row for 'Partial with reset' under the Unitary column is marked viable, but the text in Sec. IIIC correctly says the reduced map must be strictly contractive and that this depends on the unitary. The table could be annotated to indicate this conditional viability.
  4. [Appendix C, Corollary 2] The proof says 'Since both sub-maps are unital, their only common fixed point is the maximally mixed state.' This is correct, but the sentence would be clearer if it noted that the unique fixed point of the composition is I/d because the composition is unital and strictly contractive.

Circularity Check

0 steps flagged

No significant circularity: the central contractivity criterion is proved in-house from definitions, benchmarks are standard, and self-cited criteria are independent mathematical facts.

full rationale

The derivation chain is self-contained for its central new result. Theorem 1 (Appendix C) is proved from the definitions of the trace-norm contraction coefficient and saturating sets (Prop. 1, Def. 1, Thm. 1), with no fitted input or external result used as the load-bearing premise. The monitored QRC map T=(M∘Λ)(s) is used only to formulate the classification; the conditions ESP/FMP via strict contractivity and the blocking role of unitality are cited to prior peer-reviewed works, including some from the same group, but these are parameter-free mathematical theorems (valid for any CPTP map) and therefore independent support under the stated rules. Numerical results use standard STM and NARMA benchmarks with stated parameters (Table II), and θ is scanned, not fitted to the reported capacities; no fitted quantity is renamed as a prediction. The only in-scope weakness is a correctness gap, not circularity: Sec. IIIC asserts AD monitoring gives input separability 'for any input-dependent unitary' from strict contractivity plus non-unitality, although the paper earlier notes non-unitality is not sufficient (fully damped channel counterexample). This overclaim is a validity issue, so per rule 5 it is a correctness risk, not a circularity, and does not raise the score.

Axiom & Free-Parameter Ledger

2 free parameters · 5 axioms · 0 invented entities

The central contribution rests on standard QRC sufficiency conditions, the non-selective ensemble model, and finite-dimensional CPTP contraction facts. The only hand-tunable knobs in simulations are measurement strengths and reservoir parameters, which are scanned rather than fitted to the conclusions. No new physical entities are introduced; the monitoring schemes are implemented with existing qubit-probe couplings.

free parameters (2)
  • Measurement strength θ (dephasing and AD monitoring) = scanned; e.g., θ=0.9π/2 for weak dephasing, θ=π/2 for AD in Fig. 8, θ=0 or π for projective/reset limits
    Controls the balance of information extraction and back-action. It is a tunable physical knob, not fitted to a target; the qualitative resource claim holds across the scan, but the numerical performance optima depend on θ and the task.
  • Reservoir and protocol hyperparameters (N=5, h=1, J=1, Δt=2, L=3000/14000, N_P=3–4) = as in Table II
    Chosen to lie in the ergodic phase and to keep simulations tractable; not fitted to the central theoretical claims, but the numerical benchmarks do not prove parameter-independent generality.
axioms (5)
  • domain assumption Strict contractivity of a CPTP map is sufficient for the echo-state property and fading memory, and unital maps cannot yield input-dependent fixed points.
    Invoked in Secs. IIIA–IIIB as the foundation for all viability criteria; the properties are cited from [44,46,54] rather than rederived.
  • domain assumption The computational dynamics is represented by the non-selective ensemble-averaged map T(s)=(M∘Λ)(s), with features from outcome probabilities of the same measurements.
    Eqs. (16), (20), (27); this is the modeling choice that lets the authors use CPTP-map criteria. Single-shot/trajectory-resolved processing is not analyzed.
  • standard math CPTP maps are contractive in trace norm, and the contraction-coefficient supremum is attained on the compact set of normalized traceless Hermitian operators.
    Used in the proof of Proposition 1 and Theorem 1, Appendix C.
  • domain assumption Each measurement uses a fresh probe reset to |0>⟨0|, so the monitored ensemble evolution is Markovian and divisible per step.
    Sec. II, Eqs. (4)–(7); partial measurements without reset are explicitly non-Markovian and are treated only with reset in the QRC map.
  • domain assumption Input encoding via a single-qubit RY rotation and reservoir unitaries (Haar-random or transverse-field Ising) captures representative dynamics.
    Sec. IIIC and Appendix B; numerical generality beyond these encodings is not demonstrated.

pith-pipeline@v1.3.0-alltime-deepseek · 26864 in / 17759 out tokens · 189044 ms · 2026-08-02T02:29:10.161487+00:00 · methodology

0 comments
read the original abstract

Quantum reservoir computing (QRC) provides a powerful framework for processing temporal data using quantum dynamics, but incorporating measurements into the reservoir remains a fundamental challenge and distinctive feature with respect to classical settings. The induced back-action can vary from a source of disturbance to a computational resource, as measurement deeply modifies the dynamics underlying temporal processing. Existing approaches have treated specific monitoring schemes independently, missing the common physical principles governing online quantum reservoirs. Here we develop a general theory of monitored quantum reservoir computing based on indirect quantum measurements, which unifies projective, weak, partial, and dissipative monitoring protocols within a single operational framework. Measurement back-action can serve as a controllable resource, providing the effective dissipation and non-unital dynamics required for successful QRC, even when the underlying unmonitored evolution is unsuitable. We derive general criteria under which monitored dynamics satisfy the echo-state property, fading memory, and input separability, including a necessary and sufficient condition for emergent strict contractivity. By comparing different monitoring schemes under a common reference dynamics, we show that these protocols are not interchangeable parameterizations to be optimized for peak performance, but rather constitute qualitatively distinct routes to computational capability, each enabled by the interplay between information extraction and measurement-induced disturbance -- a trade-off that can be further shaped through time multiplexing. Our results provide a unified theoretical foundation for online monitored quantum reservoir computing and establish quantum measurement engineering as a systematic approach for designing reservoir architectures across different quantum platforms.

Figures

Figures reproduced from arXiv: 2607.14311 by Gian Luca Giorgi, Gonzalo Manzano, Oriol Morgu\'i-Sancho, Roberta Zambrini (Instituto de F\'isica Interdisciplinar y Sistemas Complejos (IFISC), UIB-CSIC).

Figure 1
Figure 1. Figure 1: FIG. 1. (a) The monitored dynamics of a quantum system can [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Using the equivalence between open quantum dynamics and nonselective measurement maps, one can engineer indirect [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. (a) QRC paradigm. At each step, an input [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Dynamics of [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Correlation between the predicted and target time series [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7. Effect of the finite resources on the Short-Term Mem [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Correlation between the predicted and target time series for [PITH_FULL_IMAGE:figures/full_fig_p011_6.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8. (a-c) Dynamics of [PITH_FULL_IMAGE:figures/full_fig_p013_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9. Effect of the finite resources on the Short-Term Mem [PITH_FULL_IMAGE:figures/full_fig_p015_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: FIG. 10. The composition [PITH_FULL_IMAGE:figures/full_fig_p017_10.png] view at source ↗

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

82 extracted references · 6 linked inside Pith

  1. [1]

    The cor- responding measurement operators are Ω0(𝜃)=cos (𝜃/2)Π0+sin(𝜃/2)Π1, Ω1(𝜃)=sin (𝜃/2)Π0+cos(𝜃/2)Π1,(10) 4 FIG

    Dephasing measurement map A tunable-strength measurement scheme, whose aver- aged back-action is equivalent to a dephasing channel, can be constructed by setting the coupling unitary as [14] 𝑈(deph) 𝑃𝑆 =CNOT𝑆,𝑃(RY(𝜃)𝑃⊗I𝑆), 𝜃∈ [0,𝜋/2),(9) where the first and second indices of the controlled gate denote the control and target qubits, respectively. The cor- ...

  2. [2]

    Thecorresponding map on the system, after projective measurement on the probequbit(Fig

    Amplitude Damping measurement map An unraveling for the Amplitude Damping (AD) chan- nel can be constructed by coupling the system qubit to an auxiliary qubit through [37] 𝑈(AD) 𝑃𝑆 =CNOT 𝑃,𝑆CRY(𝜃)𝑆,𝑃 , 𝜃∈(0,𝜋],(12) where𝜃controlsthedampingstrength. Thecorresponding map on the system, after projective measurement on the probequbit(Fig. 2b),yieldssystemmeas...

  3. [3]

    Thesystemispartitionedinto twosubsets: thequbitsthataremeasured(𝑃)andthosethat are not (𝑀)

    Partial measurement map and reset Instead of the controllable-strength monitoring via tun- able coupling to auxiliary units, as in the two previous examples, one can assign to some of the units of a large system (e.g., a part of a quantum circuit or a many-body system)theroleoftheprobe. Thesystemispartitionedinto twosubsets: thequbitsthataremeasured(𝑃)and...

  4. [4]

    X13, 031012 (2023)

    D.Yang,S.F.Huelga,andM.B.Plenio,Efficientinformation retrievalforsensingviacontinuousmeasurement,Phys.Rev. X13, 031012 (2023)

    2023

  5. [5]

    C.L.Degen,F.Reinhard,andP.Cappellaro,Quantumsens- ing, Rev. Mod. Phys.89, 035002 (2017)

    2017

  6. [6]

    Mujal, R

    P. Mujal, R. Martínez-Peña, G. L. Giorgi, M. C. Soriano, and R. Zambrini, Time-series quantum reservoir computing with weak and projective measurements, npj Quantum Inf. 9, 10.1038/s41534-023-00682-z (2023)

    doi:10.1038/s41534-023-00682-z 2023

  7. [7]

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

    2009

  8. [8]

    Jacobs,Quantum Measurement Theory and its Applica- tions(Cambridge University Press, 2014)

    K. Jacobs,Quantum Measurement Theory and its Applica- tions(Cambridge University Press, 2014)

    2014

  9. [9]

    Ryan-Anderson, J

    C. Ryan-Anderson, J. G. Bohnet, K. Lee, D. Gresh, A. Han- kin,J.P.Gaebler,D.Francois,A.Chernoguzov,D.Lucchetti, N.C.Brown,T.M.Gatterman,S.K.Halit,K.Gilmore,J.A. Gerber,B.Neyenhuis,D.Hayes,andR.P.Stutz,Realization of real-time fault-tolerant quantum error correction, Phys. Rev. X11, 041058 (2021)

    2021

  10. [10]

    Bayat, B

    A. Bayat, B. Alkurtass, P. Sodano, H. Johannesson, and S. Bose, Measurement quench in many-body systems, Phys. Rev. Lett.121, 030601 (2018)

    2018

  11. [11]

    J. M. Koh, S.-N. Sun, M. Motta, and A. J. Minnich, Measurement-induced entanglement phase transition on a superconducting quantum processor with mid-circuit read- out, Nature Physics19, 1314–1319 (2023)

    2023

  12. [12]

    I. M. Gelfand and M. A. Naimark, On the imbedding of normed rings into the ring of operators in hilbert space, Matematiceskij sbornik54, 197 (1943)

    1943

  13. [13]

    K.W.Murch,S.J.Weber,C.Macklin,andI.Siddiqi,Observ- ing single quantum trajectories of a superconducting quan- tum bit, Nature502, 211–214 (2013)

    2013

  14. [14]

    S. Roy, J. T. Chalker, I. V. Gornyi, and Y. Gefen, Measurement-induced steering of quantum systems, Phys. Rev. Res.2, 033347 (2020)

    2020

  15. [15]

    Skinner, J

    B. Skinner, J. Ruhman, and A. Nahum, Measurement- induced phase transitions in the dynamics of entanglement, Phys. Rev. X9, 031009 (2019)

    2019

  16. [16]

    F. Hua, Y. Jin, Y. Chen, S. Vittal, K. Krsulich, L. S. Bishop, J. Lapeyre, A. Javadi-Abhari, and E. Z. Zhang, Exploit- ing qubit reuse through mid-circuit measurement and reset (2023), arXiv:2211.01925 [quant-ph]

    Pith/arXiv arXiv 2023

  17. [17]

    Fujii and K

    K. Fujii and K. Nakajima, Harnessing disordered-ensemble quantumdynamicsformachinelearning,Phys.Rev.Appl.8, 024030 (2017)

    2017

  18. [18]

    Mujal, R

    P. Mujal, R. Martínez-Peña, J. Nokkala, J. García-Beni, G. L. Giorgi, M. C. Soriano, and R. Zambrini, Op- portunities in quantum reservoir computing and extreme learning machines, Advanced Quantum Technologies4, 10.1002/qute.202100027 (2021)

    doi:10.1002/qute.202100027 2021

  19. [19]

    Dressel, T

    J. Dressel, T. A. Brun, and A. N. Korotkov, Implement- ing generalized measurements with superconducting qubits, Phys. Rev. A90, 032302 (2014)

    2014

  20. [20]

    T. A. Brun, L. Diósi, and W. T. Strunz, Test of weak mea- surement on a two- or three-qubit computer, Phys. Rev. A 77, 032101 (2008)

    2008

  21. [21]

    ××× ✓✓✓ [22, 59, 62] AD Yes No ✓✓✓

  22. [22]

    Magnard, P

    P. Magnard, P. Kurpiers, B. Royer, T. Walter, J.-C. Besse, S. Gasparinetti, M. Pechal, J. Heinsoo, S. Storz, A. Blais, and A. Wallraff, Fast and unconditional all-microwave reset of a superconducting qubit, Phys. Rev. Lett.121, 060502 (2018)

    2018

  23. [23]

    Yes"and"No

    ✓✓✓ ✓✓✓ ✓✓✓ Unmonitored – – ××× ✓✓✓ [17] ✓✓✓ [21, 53] ××× ✓✓✓ [48–50] TABLE I. Viability of the online QRC maps with different measurement schemesMand dynamicsΛ. In the table,✓✓✓and×××indicate whetherthecorrespondingmapisorisnotaviableQRCscheme,respectively,while"Yes"and"No"denotewhetheragivenmeasurement schemesatisfiesthepropertiesofstrictcontractivityan...

    2025

  24. [24]

    J.García-Beni,G.L.Giorgi,M.C.Soriano,andR.Zambrini, Scalable photonic platform for real-time quantum reservoir computing, Phys. Rev. Appl.20, 014051 (2023)

    2023

  25. [25]

    Soriano, and R

    J.Nokkala,R.Martínez-Peña,G.L.Giorgi, V.Parigi, M.C. Soriano, and R. Zambrini, Gaussian states of continuous- variable quantum systems provide universal and versatile reservoir computing, Commun. Phys.4, 10.1038/s42005- 021-00556-w (2021)

    doi:10.1038/s42005- 2021

  26. [26]

    Carles, J

    B. Carles, J. Dudas, L. Balembois, J. Grollier, and D. Marković, Experimental quantum reservoir computing with a circuit-quantum-electrodynamics system, Phys. Rev. Appl.25, 054005 (2026)

    2026

  27. [27]

    J. Chen, H. I. Nurdin, and N. Yamamoto, Temporal infor- mation processing on noisy quantum computers, Phys. Rev. Appl.14, 024065 (2020)

    2020

  28. [28]

    Yasuda, Y

    T. Yasuda, Y. Suzuki, T. Kubota, K. Nakajima, Q. Gao, W. Zhang, S. Shimono, H. I. Nurdin, and N. Yamamoto, Quantum reservoir computing with repeated measurements on superconducting devices (2023), arXiv:2310.06706 [quant-ph]

    Pith/arXiv arXiv 2023

  29. [29]

    F. Hu, S. A. Khan, N. T. Bronn, G. Angelatos, G. E. Row- lands,G.J.Ribeill,andH.E.Türeci,Overcomingthecoher- ence time barrier in quantum machine learning on temporal data, Nat. Commun. , 7491 (2024)

    2024

  30. [30]

    Ciccarello, S

    F. Ciccarello, S. Lorenzo, V. Giovannetti, and G. M. Palma, Quantum collision models: Open system dynamics from repeated interactions, Physics Reports954, 1–70 (2022)

    2022

  31. [31]

    Kobayashi, K

    K. Kobayashi, K. Fujii, and N. Yamamoto, Feedback-driven quantum reservoir computing for time-series analysis, PRX Quantum5, 040325 (2024)

    2024

  32. [32]

    Paparelle, J

    I. Paparelle, J. Henaff, J. García-Beni, É. Gillet, D. Mon- tesinos,G.L.Giorgi,M.C.Soriano,R.Zambrini,andV.Pa- rigi, Experimental memory control in continuous-variable opticalquantumreservoircomputing,Nat.Photonics20,413 (2026)

    2026

  33. [33]

    Murauer, R

    J. Murauer, R. Krishnakumar, S. Tornow, and M. Geier- hos, Feedback connections in quantum reservoir computing with mid-circuit measurements, in2025 IEEE International ConferenceonQuantumComputingandEngineering(QCE) (IEEE, 2025) p. 1646–1652

    2025

  34. [34]

    Monomi, W

    T. Monomi, W. Setoyama, and Y. Hasegawa, Feedback- enhancedquantumreservoircomputingwithweakmeasure- ments (2025), arXiv:2503.17939 [quant-ph]

    Pith/arXiv arXiv 2025

  35. [35]

    V. B. Braginsky, F. Y. Khalili, and K. S. Thorne,Quantum Measurement(Cambridge University Press, 1992)

    1992

  36. [36]

    Dressel and A

    J. Dressel and A. N. Jordan, Contextual-value approach to the generalized measurement of observables, Phys. Rev. A 85, 022123 (2012)

    2012

  37. [37]

    T. A. Brun, A simple model of quantum trajectories, Amer- ican Journal of Physics70, 719 (2002)

    2002

  38. [38]

    Manzano, J

    G. Manzano, J. M. Horowitz, and J. M. R. Parrondo, Quan- tum fluctuation theorems for arbitrary environments: Adia- 19 batic and nonadiabatic entropy production, Phys. Rev. X8, 031037 (2018)

    2018

  39. [39]

    C. A. Fuchs and K. Jacobs, Information-tradeoff relations for finite-strength quantum measurements, Phys. Rev. A63, 062305 (2001)

    2001

  40. [40]

    V.B.Braginsky,Y.I.Vorontsov,andK.S.Thorne,Quantum nondemolition measurements, Science209, 547 (1980)

    1980

  41. [41]

    Dressel, S

    J. Dressel, S. Agarwal, and A. N. Jordan, Contextual values of observables in quantum measurements, Phys. Rev. Lett. 104, 240401 (2010)

    2010

  42. [42]

    C.Wringe,M.Trefzer,andS.Stepney,Reservoircomputing benchmarks: a tutorial review and critique, International JournalofParallel,EmergentandDistributedSystems,1–39 (2025)

    2025

  43. [43]

    M. A. Nielsen and I. L. Chuang,Quantum Computation andQuantumInformation: 10thAnniversaryEdition(Cam- bridge University Press, 2010)

    2010

  44. [44]

    Oreshkov and T

    O. Oreshkov and T. A. Brun, Weak measurements are uni- versal, Phys. Rev. Lett.95, 110409 (2005)

    2005

  45. [45]

    Z. K. Minev, S. O. Mundhada, S. Shankar, P. Reinhold, R.Gutiérrez-Jáuregui,R.J.Schoelkopf,M.Mirrahimi,H.J. Carmichael,andM.H.Devoret,Tocatchandreverseaquan- tum jump mid-flight, Nature570, 200 (2019)

    2019

  46. [46]

    W. Chen, M. Abbasi, Y. N. Joglekar, and K. W. Murch, Quantum jumps in the non-hermitian dynamics of a super- conducting qubit, Phys. Rev. Lett.127, 140504 (2021)

    2021

  47. [47]

    Gyurik, F

    C. Gyurik, F. Wudarski, E. J. Philip, A. Sannia, H. Sadeghi, O. Kyriienko, D. Venturelli, and A. A. Gentile, From quan- tum feature maps to quantum reservoir computing: an applicative perspective, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences384, 20250085 (2026)

    2026

  48. [48]

    Suzuki, Q

    Y. Suzuki, Q. Gao, K. C. Pradel, K. Yasuoka, and N. Ya- mamoto,Naturalquantumreservoircomputingfortemporal information processing, Scientific Reports12, 1353 (2022)

    2022

  49. [49]

    Gonon, R

    L. Gonon, R. Martínez-Peña, and J.-P. Ortega, Feedback- driven recurrent quantum neural network universality (2026), arXiv:2506.16332 [quant-ph]

    arXiv 2026

  50. [50]

    Sannia, R

    A. Sannia, R. Martínez-Peña, M. C. Soriano, G. L. Giorgi, and R. Zambrini, Dissipation as a resource for Quantum Reservoir Computing, Quantum8, 1291 (2024)

    2024

  51. [51]

    S.BoydandL.Chua,Fadingmemoryandtheproblemofap- proximating nonlinear operators with Volterra series, IEEE Transactions on Circuits and Systems32, 1150 (1985)

    1985

  52. [52]

    Martínez-Peña and J.-P

    R. Martínez-Peña and J.-P. Ortega, Quantum reservoir com- putinginfinitedimensions,Phys.Rev.E107,035306(2023)

    2023

  53. [53]

    Raginsky, Strictly contractive quantum channels and physically realizable quantum computers, Phys

    M. Raginsky, Strictly contractive quantum channels and physically realizable quantum computers, Phys. Rev. A65, 032306 (2002)

    2002

  54. [54]

    Martínez-Peña and J.-P

    R. Martínez-Peña and J.-P. Ortega, Input-dependence in quantum reservoir computing, Phys. Rev. E111, 065306 (2025)

    2025

  55. [55]

    Kubota, Y

    T. Kubota, Y. Suzuki, S. Kobayashi, Q. H. Tran, N. Ya- mamoto,andK.Nakajima,Temporalinformationprocessing inducedbyquantumnoise,Phys.Rev.Res.5,023057(2023)

    2023

  56. [56]

    Domingo, G

    L. Domingo, G. Carlo, and F. Borondo, Taking advantage of noise in quantum reservoir computing, Sci Rep13, 8790 (2023)

    2023

  57. [57]

    Palacios, R

    A. Palacios, R. Martínez-Peña, M. C. Soriano, G. L. Giorgi, and R. Zambrini, Role of coherence in many-body quantum reservoir computing, Commun. Phys.7(2024)

    2024

  58. [58]

    Liew, Quantum reservoir processing, npj Quantum Inf.5 (2019)

    S.Ghosh,A.Opala,M.Matuszewski,T.Paterek,andT.C.H. Liew, Quantum reservoir processing, npj Quantum Inf.5 (2019)

    2019

  59. [59]

    D. Fry, A. Deshmukh, S. Y.-C. Chen, V. Rastunkov, and V. Markov, Optimizing quantum noise-induced reservoir computing for nonlinear and chaotic time series prediction, Scientific Reports13, 19326 (2023)

    2023

  60. [60]

    Čindrak, L

    S. Čindrak, L. Giebeler, N. Götting, C. Gies, and K. Lüdge, Memory-nonlinearity trade-off across quantum reservoir computing frameworks (2026), arXiv:2603.21371 [quant- ph]

    arXiv 2026

  61. [61]

    Čindrak, B

    S. Čindrak, B. Donvil, K. Lüdge, and L. Jaurigue, Enhanc- ing the performance of quantum reservoir computing and solving the time-complexity problem by artificial memory restriction, Phys. Rev. Res.6, 013051 (2024)

    2024

  62. [62]

    Y.Hou,J.Hua,Z.Wu,W.Xia,Y.Chen,X.Li,Z.Li,X.Peng, andJ.Du,High-accuracytemporalpredictionviaexperimen- tal quantum reservoir computing in correlated spins, Phys. Rev. Lett.136, 120602 (2026)

    2026

  63. [63]

    ✓✓✓ ✓✓✓ ✓✓✓ ✓✓✓ [44, 64] Partial projective No Yes ××× ✓✓✓ ✓✓✓ ××× ✓✓✓ [62] Partial with reset No No ✓✓✓ [23, 58] ✓✓✓

  64. [64]

    García-Beni, I

    J. García-Beni, I. Paparelle, V. Parigi, G. L. Giorgi, M. C. Soriano, and R. Zambrini, Quantum machine learning via continuous-variable cluster states and teleportation, EPJ Quantum Technol.12(2025)

    2025

  65. [65]

    E.L.Connerty,E.N.Evans,G.Angelatos,andV.Narayanan, Predicting chaotic dynamics on NISQ hardware with quan- tum reservoir networks, Commun. Phys. (2026)

    2026

  66. [66]

    E.Ricci,F.Monzani,L.Nigro,andE.Prati,Quantumreser- voir computing induced by controllable damping (2025), arXiv:2508.14621 [quant-ph]

    Pith/arXiv arXiv 2025

  67. [67]

    Franceschetto, M

    G. Franceschetto, M. Płodzień, M. Lewenstein, A. Acín, andP.Mujal,Harnessingquantumbackactionfortime-series processing, Phys. Rev. X16, 021002 (2026)

    2026

  68. [68]

    F.Monzani,E.Ricci,L.Nigro,andE.Prati,Non-unitalnoise in a superconducting quantum computer as a computational resource for reservoir computing (2025), arXiv:2409.07886 [quant-ph]

    Pith/arXiv arXiv 2025

  69. [69]

    E.L.ConnertyandE.N.Evans,Controllablequantummem- ory capacity in quantum reservoir networks with tunable partial-swaps (2026), arXiv:2605.12713 [quant-ph]

    Pith/arXiv arXiv 2026

  70. [70]

    S. Das, G. L. Giorgi, and R. Zambrini, Quantum reservoir computing in jaynes-cummings models: Nonlinear memory andtime-seriesprediction,Phys.Rev.Res.8,023148(2026)

    2026

  71. [71]

    A.Sannia,G.L.Giorgi,S.Longhi,andR.Zambrini,Liouvil- lianskineffectinquantumneuralnetworks,OpticaQuantum 3, 189 (2025)

    2025

  72. [72]

    For any allowed value of𝑝, thecompositionΛ 2◦Λ 1 isunital,Λ 2◦Λ 1(I/𝑑)=I/𝑑,while neithersub-mapis:Λ 1(I/𝑑)=|0⟩⟨0|andΛ 2(I/𝑑)≠I/𝑑for 𝑝 <1

    An example of this is the concatenation of an amplitude damping channel with full damping, i.e.Λ1𝜌=|0⟩⟨0|, and a channel of the formΛ2𝜌=(1−𝑝)𝜌+𝑝𝜎 𝑝 with(𝑑− 1)/𝑑≤𝑝 <1, where𝜎 𝑝 =I/(𝑝𝑑)+ (1−1/𝑝 )|0⟩⟨0|is a valid quantum state in this parameter range and𝑑is the dimension of the Hilbert space. For any allowed value of𝑝, thecompositionΛ 2◦Λ 1 isunital,Λ 2◦Λ 1(...

  73. [73]

    Nigro, G

    L. Nigro, G. L. Giorgi, E. Prati, and R. Zambrini, Ana- logneutral-atomquantumreservoircomputingsatisfyingthe echo state property by design (2026), In preparation

    2026

  74. [74]

    Hayden and J

    P. Hayden and J. Preskill, Black holes as mirrors: quantum information in random subsystems, Journal of high energy physics2007, 120 (2007)

    2007

  75. [75]

    Martínez-Peña, G

    R. Martínez-Peña, G. L. Giorgi, J. Nokkala, M. C. Soriano, and R. Zambrini, Dynamical phase transitions in quantum 20 reservoir computing, Phys. Rev. Lett.127, 100502 (2021)

    2021

  76. [76]

    Gherardini, G

    S. Gherardini, G. Giachetti, S. Ruffo, and A. Trombettoni, Thermalization processes induced by quantum monitoring in multilevel systems, Phys. Rev. E104, 034114 (2021)

    2021

  77. [77]

    G.Llodrà,P.Mujal,R.Zambrini,andG.L.Giorgi,Quantum reservoir computing in atomic lattices, Chaos, Solitons & Fractals195, 116289 (2025)

    2025

  78. [78]

    Llodrà, C

    G. Llodrà, C. Charalambous, G. L. Giorgi, and R. Zam- brini, Benchmarking the role of particle statistics in quan- tum reservoir computing, Advanced Quantum Technologies 6, 2200100 (2023)

    2023

  79. [79]

    Kobayashi and Y

    K. Kobayashi and Y. Motome, Edge of many-body quantum chaosinquantumreservoircomputing,Phys.Rev.Lett.136, 040602 (2026)

    2026

  80. [80]

    Massar, J

    L.Appeltant,M.C.Soriano,G.VanderSande,J.Danckaert, S. Massar, J. Dambre, B. Schrauwen, C. R. Mirasso, and I. Fischer, Information processing using a single dynamical node as complex system, Nature Communications2, 468 (2011)

    2011

Showing first 80 references.