Dynamic Synaptic Modulation of LMG Qubits populations in a Bio-Inspired Quantum Brain

E. Romera; J. J. Torres

REVIEW 5 major objections 3 minor 57 references

Dynamic Synaptic Modulation of LMG Qubits populations in a Bio-Inspired Quantum Brain

T0 review · 5 major / 3 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read A quantum network with synapse-like feedback settles into stable, brain-like population rhythms near half excitation.

desk verdict A legitimate new combination of LMG collective dynamics with Tsodyks-Markram synaptic feedback, but several headline claims don't survive contact with the equations. read the letter →

arxiv 2602.16003 v2 pith:LXSQHUTK submitted 2026-02-17 quant-ph physics.comp-ph

J. J. Torres , E. Romera This is my paper · ORCID

classification quant-phphysics.comp-ph

keywords quantumbrainLipkin-Meshkov-Glickmodelsynapticplasticityhomeostasiscollectivespindynamicsneuralnetworksentanglementrhythmogenesis

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper attempts to show that attaching a synapse-inspired, activity-dependent coupling to a fully connected qubit network — the Lipkin-Meshkov-Glick (LMG) model — turns a rigid many-body quantum system into a self-regulating one. The coupling g(t)=g0 r(t) weakens when collective activity rises and recovers when it falls, and this feedback drives the qubit population toward a metastable operating point near N/2 excited qubits, generates activity-dependent oscillation periods, and makes larger networks more stable. The authors call this the first quantum neural architecture with short-term synaptic plasticity that stays computationally tractable because the dynamics remain in the symmetric subspace. If the proposal is right, it provides a minimal, scalable building block for quantum-brain-like systems on future hardware, with population homeostasis and rhythmogenesis arising without external tuning.

What carries the argument

The Lipkin-Meshkov-Glick (LMG) Hamiltonian, an all-to-all coupled collective spin model restricted to the symmetric subspace j=N/2, where [J^2, H]=0 keeps the dimensionality polynomial. Onto it the authors graft a feedback loop: the collective excitation operator E(t)=1/2+Jz(t)/N drives a synaptic efficacy r(t) and release probability U(t) via two first-order equations (4)-(5), which in turn rescale the coupling g0 r(t). This closed loop of unitary Schrödinger evolution and classical update is the mechanism that produces the emergent population homeostasis and rhythmogenesis.

What would settle it

Take a system with N=20, γ=0.8, g0=0.5, start it in the fully excited Dicke state, and add a weak continuous measurement of Jz with back-action (e.g., quantum trajectories). If the measured signal replaces the ideal ⟨Jz⟩, the homeostatic convergence to N/2 and the periodic fidelity revivals should degrade or disappear; observing that would falsify the idealized feedback assumption. Conversely, observing the same convergence under monitored dynamics would support it.

Watch

Extended reading notes

Core claim

The central claim is that a collective quantum spin system with an activity-dependent coupling — g(t)=g0 r(t), where r(t) evolves like a short-term synaptic depression variable — spontaneously organizes its qubit population around N/2 excitations and exhibits periodic, size-dependent collective oscillations. Starting from fully silent or fully saturated initial states, the feedback loop re-stabilizes the system near the half-excited set point; starting from exactly half-excited states, it traps the system. The authors show that increasing N reduces fluctuations, that synaptic depression slows and blurs oscillations while biasing energy-level occupation, and that synaptic facilitation lengthe

Load-bearing premise

The whole dynamics assume ⟨Jz⟩ is available as a noiseless classical signal at every instant, without specifying the measurement that extracts it or the back-action that measurement would have on the quantum state.

Editorial extensions

If this is right

A quantum qubit assembly can be stabilized at a chosen population set point purely by activity-dependent coupling, without external control fields.
Oscillation period becomes tunable via τr and τf, offering a clock or rhythm primitive for quantum neural circuits.
Larger networks exhibit smaller fluctuations and longer dwell times near the operating point, implying a size-dependent robustness that favors scalable implementations.
Initial states with exactly half excitation produce no low-frequency collective oscillations and remain trapped, marking a special operating regime distinct from fully silent or saturated starts.
The model is circuit-implementable, providing a route to test these predictions on quantum hardware.

Reading between the lines

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

The feedback loop uses ⟨Jz⟩ as a classical control signal without specifying the measurement that produces it; a physical readout would introduce back-action and decoherence that could alter the symmetric-subspace dynamics. A natural extension is to include monitored quantum trajectories and compare.
The metastable operating point near N/2 is shown for a handful of initial preparations; a systematic scan over initial states and coupling parameters would reveal whether this set point is generic or an artifact of Dicke-like symmetric states.
The observed reduction of bipartite entanglement at fidelity revivals suggests the system could serve as a controllable entanglement reset mechanism, useful for quantum memory or reinitialization protocols.
Because the dynamics live in the symmetric subspace, a semiclassical single-spin picture may reproduce the population oscillations; comparing the full quantum simulation to that approximation would show whether any genuine quantum advantage underlies the reported homeostasis.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Two things before you spend time on this. The model is a legitimate new combination: the two-qubit dynamic synapse from Torres and Manzano (ref 48) extended to the collective LMG symmetric subspace, with an added facilitation equation. That part is worth a look. But the paper's own equations contradict several of its central results, and the most interesting claims — the 'trapped state', the N/2 homeostatic set point — are either false as stated or baked into the chosen observable.

What's actually new: an N-qubit LMG with g(t)=g0 r(t) where r and U evolve via activity-dependent differential equations, giving a numerically cheap way to study collective oscillations with synaptic feedback. The qualitative phenomenology — anticorrelation between activity and synaptic efficacy, slowing of oscillations with stronger depression, asymmetric transitions — is plausible and likely reproducible in spirit. Using the symmetric subspace to make large N tractable is sensible. The paper is honest about its scope and doesn't claim a computational benchmark.

That said, the problems are not minor. The 'exactly symmetric semi-activation' state is never defined as a wavefunction. If it means the Dicke state |N/2,0⟩, then for N≥4 the state is not an eigenstate of Jx²; Jx² couples |0⟩ to |±2⟩, so the system evolves within that manifold. Only the population E = 1/2 + ⟨Jz⟩/N is stationary. The claim of being 'indefinitely trapped in the same quantum state' is wrong as written. This matters because the 'special operating regime' in Section III.B and the conclusions lean on it. Also, Fig. 1 sets τ_r=0 although Eq. 4 divides by τ_r; Fig. 7 freezes U(t)=0.5 while Eq. 5 would evolve it. These are not typos but issues that make it impossible to know what was actually simulated. Initial states are given only as 'around half excited' or 'N/2−1', without concrete wavefunctions, so the numerics aren't reproducible from the text.

There's also a conceptual issue: the feedback loop treats ⟨E⟩_t as a classical control signal with no measurement model. If you have to measure ⟨Jz⟩ to update r and U, you get back-action and decoherence. The paper doesn't address this, so the 'quantum' dynamics may be silently an open, conditioned process. And much of the 'homeostasis' is the observable: E = 1/2 + ⟨Jz⟩/N equals 1/2 whenever ⟨Jz⟩=0, so any parity-symmetric state sits at the set point. The anticorrelation between r and E is literally the −U r ⟨E⟩ term in Eq. 4. The circularity burden is high.

Still, the core idea — activity-dependent coupling in an LMG model — is defensible and could be made rigorous with concrete initial states, corrected equations, and a measurement model. The flaws are correctable.

Who is this for? People working in quantum neuromorphic computing and quantum reservoir computing who want a toy model of collective dynamics with synaptic plasticity. It doesn't demonstrate a task or hardware implementation, so it's a modeling paper. I'd send it to a serious referee; it has a real kernel. But the referee should demand a rewritten version with the specific fixes: define the initial states, correct or explain the τ_r=0 and U-frozen limits, re-examine the trapped-state claim, and ideally release code.

Referee Report

5 major / 3 minor

Summary. The paper proposes a biologically inspired quantum neural network in which N qubits are governed by the Lipkin-Meshkov-Glick (LMG) Hamiltonian with a synaptic-efficacy feedback: the coupling is g(t)=g0 r(t), and r(t) (and optionally a release probability U(t)) evolve according to rate equations driven by the quantum expectation value E=1/2+⟨Jz⟩/N. The authors simulate the resulting nonlinear dynamics for various initial conditions and system sizes, and report homeostatic convergence of the excited fraction toward N/2, activity-dependent oscillation periods, and effects of synaptic depression and facilitation on collective oscillations and bipartite entanglement. The stated goal is to construct a minimal building block for quantum brain-like systems.

Significance. If the results were established, the model would provide a numerically tractable, symmetric-subspace architecture connecting short-term synaptic plasticity with collective quantum dynamics, and the paper would offer a useful starting point for further studies. The paper does make a concrete modeling proposal and presents numerical explorations for N up to 80; the use of the LMG collective-spin reduction is appropriate and keeps the Hilbert-space dimension manageable. However, the central claims suffer from serious gaps: the trapped-state assertion is demonstrably false for the most natural reading of the initial condition, the initial states are not uniquely specified, and the feedback loop's measurement back-action is not addressed. These issues prevent the results from being reproduced or verified from the text.

major comments (5)

[Section III.A, Eq. (2)] The claim that an 'exactly symmetric semi-activation' leaves the system 'indefinitely trapped in the same quantum state, independently of N' is false for N≥4 if the initial state is the Dicke state |j=N/2,m=0⟩. With h=0 and γ=1, Eq. (2) reduces to H=−(2g/N)Jx². Since Jx²|j,0⟩ contains components with m=±2 (from J±²|j,0⟩), |j,0⟩ is not an eigenstate and the full quantum state evolves within the m=0,±2 manifold. Only the population ⟨Jz⟩ (and hence E) is stationary. This invalidates the 'special operating regime' description in Section III.B and the corresponding conclusions. The initial state must be defined and the claim corrected to refer to population stationarity, not full state stationarity.
[Section III (Figs. 2-9)] The numerical initial states are not defined as wavefunctions. Phrases such as 'around half of the neuronal qubits are excited', 'N/2−1', 'fully silent', and 'fully saturated' do not uniquely specify a quantum state. The dynamics are strongly initial-state dependent (as the paper itself stresses), so without a precise definition—e.g., a Dicke state |N/2,M⟩, a product state, or a coherent spin state—none of the figures can be reproduced. This is load-bearing for the paper's main results.
[Section II, Eqs. (3)-(5)] The feedback loop treats the quantum expectation value ⟨E⟩_t as a classical control signal: r(t) and U(t) evolve according to equations driven by ⟨Jz⟩_t while ρ(t) evolves unitarily. This presumes an ensemble interpretation or a measurement model that is never stated. If ⟨Jz⟩ is obtained by measurement, the state collapse and back-action must be included; if not, no physical mechanism connects a classical variable to a quantum expectation value. The model is internally consistent only under an unstated mean-field or ensemble ansatz, which should be made explicit before the reported dynamics can be assessed.
[Fig. 1 caption and Fig. 7 caption] There are parameter inconsistencies between the equations and the simulations. Fig. 1 lists τ_r=0, but Eq. (4) divides by τ_r, making the equation singular. Fig. 7 sets U(t)=U=0.5 constant even though Eq. (5) would evolve U(t); Section III.B states that only synaptic depression is considered, so this may be intentional, but the caption and text should state clearly that facilitation is being turned off rather than presenting U(t) as the solution of Eq. (5). These inconsistencies hinder reproducibility.
[Section III.A (after Fig. 2)] The reported 'strong anticorrelation' between r(t) and E(t) is structurally imposed by the term −U r ⟨E⟩ in Eq. (4): high activity directly decreases r, and low activity lets r recover. Similarly, the 'N/2 operating point' is hard-coded by the choice E = 1/2 + ⟨Jz⟩/N, which equals 1/2 whenever ⟨Jz⟩=0. These features are properties of the construction rather than emergent discoveries. The authors should explicitly acknowledge this and analyze whether the feedback actually drives ⟨Jz⟩ toward zero beyond the trivial definition, to avoid overclaiming the homeostatic mechanism.

minor comments (3)

[Throughout] Typos and grammar errors: 'perceptage' (Fig. 1 caption), 'sypatic' (Figs. 5 and 6 captions), 'apnel' (page 7), 'wherU' (page 5), 'depresion' (Fig. 8 caption), and several missing spaces/commas. These should be corrected.
[Section IV, Fig. 9] The text says 'τ_f ≈0 means that only synaptic depression is present' but the top row in Fig. 9 appears to use τ_f=1. Clarify the relation between the text and the plotted parameters.
[Section II, Eq. (1)] The mapping from Eq. (1) to Eq. (2) is stated without derivation; in particular, the sign of the coupling g relative to γ_x,γ_y is not explained. A brief derivation or reference would help.

Circularity Check

3 steps flagged · score 6.0 of 10

Several headline behaviors are pre-programmed into the model equations: the r–E anticorrelation is literally the −U r⟨E⟩ term of Eq. (4), the activity-dependent slowing is the g(t)=g0 r(t) substitution, and the N/2 'set point' is the midpoint of the chosen observable E=1/2+⟨Jz⟩/N.

self definitional [Section III.A, paragraph after Fig. 2; Eq. (4)]
"The results of our study reveal a strong anticorrelation between r(t) and ⟨E(t)⟩, similar to what is observed in biological systems: when population activity (the number of excited qubits) reaches its maximum, it leads to a reduction in synaptic efficacy, which subsequently reaches a relative minimum."

This is the content of Eq. (4): dr/dt = (1-r)/τ_r − U r ⟨E⟩_t. The −U r⟨E⟩ term makes r decrease whenever ⟨E⟩ is large, while the recovery term (1-r)/τ_r makes r increase when ⟨E⟩ is small. The reported anti-phase between r and E is therefore the integrated consequence of a term placed in the model by hand; it is not an emergent or independently predicted relation. The same applies to the later conclusion that synaptic efficacy 'provides a negative-feedback loop' — that loop is the definition of Eq. (4).
self definitional [Section II, g(t)=g0 r(t); Section III.B, Fig. 7 discussion]
"To account for dynamical processes affecting qubits interactions similar to those reported in [48], we consider g(t) = g0r(t) ... First, the frequency of the oscillations strongly decreases with τr"

H_LMG is proportional to g (Eq. 2), and g is defined as g0 r(t). Therefore any feedback-induced decrease of r — which stronger depression (larger τ_r) produces — directly lowers the Hamiltonian's energy scale and thus the many-body Rabi frequency. The reported 'activity-dependent oscillation periods' and the strong frequency decrease with τ_r are the time-scale modulation inserted by the definition g(t)=g0 r(t); the numerics trace the constructed coupling back out, rather than deriving an independent property.

1 more flagged steps

self definitional [Section III.A, first paragraph after Fig. 1; Eq. (6); Conclusions, Section V]
"when the initial condition corresponds to an exactly symmetric semi-activation (data not shown), the system becomes indefinitely trapped in the same quantum state, independently of the system size N. In neurodynamic terms, population activity remains strictly stationary over time (without fluctuations), such that no transitions or oscillations among microstates emerge."

The observable is defined as E = 1/2 + ⟨J_z⟩/N. An 'exactly symmetric semi-activation' is a state with ⟨J_z⟩=0, so E is exactly the midpoint 1/2 by definition. For h=0, H_LMG (Eq. 2) contains only J_x² and J_y², and ⟨0|[J_x²,J_z]|0⟩=0 for the Dicke state |N/2,0⟩, so d⟨J_z⟩/dt=0 and the population E stays at 1/2 by symmetry. The 'homeostatic set point' and 'special operating regime' are thus conventions of the chosen observable and initial state, not attractors created by the feedback. Moreover, for N≥4 the state is not literally trapped: J_x²|N/2,0⟩ has m=±2 components, so only the population expectation is stationary.

full rationale

The paper is a numerical modeling study, not a fit-to-data or first-principles derivation, and most of its microstructure results (fidelity, von Neumann/linear entropy, size-dependent fluctuation suppression) are genuine outputs of the coupled nonlinear dynamics and are not circular. I found no load-bearing self-citation chain: [48] is used transparently as the source of the two-qubit dynamic-synapse ansatz, and the classical short-term-plasticity equations are explicitly attributed to the neuroscience literature. There is no 'uniqueness theorem' and no fitted parameter renamed as a prediction. However, three headline claims are equivalent to the model's construction. The strong r–E anticorrelation of Sec. III.A is the −U r⟨E⟩ term of Eq. (4). The activity-dependent slowing of collective oscillations is the g(t)=g0 r(t) replacement in the LMG Hamiltonian. The N/2 'set point' / 'trapped semi-activation' behavior is the midpoint of the chosen observable E=1/2+⟨J_z⟩/N plus the J_z-parity symmetry of the h=0 LMG Hamiltonian; for the Dicke state |N/2,0⟩ the quantum state actually evolves while only ⟨J_z⟩ stays zero. These reductions affect the central 'homeostasis' and 'rhythmogenesis' claims, while leaving the entanglement and scaling analyses as independent numerical content. Hence a partial-circularity score of 6 is appropriate.

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

The model rests on six hand-set parameter families (g0, gamma, tau_r, tau_f, U-bar, r0), all chosen per figure, plus six axioms. The costliest assumptions are the symmetric-subspace restriction used as a dynamical shortcut and the disturbance-free classical feedback of <Jz>. No new physical entities are postulated. The observable E itself is a modeling choice that hard-codes the N/2 operating point. The parameter count is typical for an exploratory neuroscience-inspired model, but it means the reported behaviors are tuned, not predicted.

free parameters (6)

g0 (maximum coupling) = 0.05, 0.5, 2, 0.125, 1.43, 30 (per figure)
Chosen by hand; sets the overall time scale. In Fig. 9 it is deliberately rescaled per N (0.125, 1.43, 30 for N = 2, 10, 20) to 'recover the same temporal scale', which the authors admit is the LMG N-rescaling.
gamma (XY anisotropy) = 1, 0.9, 0.8
Hand-chosen; gamma = 1 in most runs, gamma = 0.8 selected 'to better visualize the effect of the synapse feedback'.
tau_r (synaptic recovery time) = 0, 1, 0.1, 10, 20, 100
Hand-chosen control parameter; tau_r = 0 in Fig. 1 is inconsistent with Eq. 4 (division by tau_r).
tau_f (facilitation time constant) = 1, 10, 100, 1000
Hand-chosen; controls how long U(t) remains elevated.
U-bar (baseline release probability) = 0.5, 0.02
Hand-chosen; in Fig. 7 the caption fixes U(t) = U = 0.5, apparently freezing Eq. 5, which would otherwise evolve U(t).
r0 (initial synaptic efficacy) = 1
Hand-chosen initial condition for r(t).

assumptions (6)

domain assumption The LMG Hamiltonian (Eqs. 1-2) is an adequate model for a population of N interacting quantum neurons.
Section II; the identification of LMG collective states with brain-like collective states is asserted, not derived.
domain assumption All dynamics can be treated in the maximally symmetric subspace j = N/2, so the Hilbert space dimension is N+1 rather than 2^N.
Section II states the ground state lies in j = N/2 and 'markedly reduces spectral complexity'; the polynomial-cost claim (Section I) rests on restricting all dynamics to this sector, stated but not proven for the feedback dynamics.
ad hoc to paper Expectation value <E>_t can be fed back as a classical control signal (Eqs. 4-5) without measurement back-action or decoherence.
Section II Eqs. 3-6; the von Neumann evolution (Eq. 3) assumes pure unitary dynamics, incompatible with a physical readout producing <Jz>_t.
domain assumption Tsodyks-Markram rate equations (Eqs. 4-5) capture the relevant biological short-term synaptic plasticity.
Section II; imported from neuroscience references [1,2,29,46,54,57].
domain assumption External field h = 0.
Section II: 'we shall work with ... h = 0'; non-zero h is declared beyond scope, so all results apply only in the zero-field case.
standard math Parity symmetry pins <Jz> = 0 for exactly-half-excited initial states, so E = 1/2 for all time.
Unstated in Section III.A but required for the 'trapped' claim; H with h = 0, gamma = 1 commutes with the parity operator and the symmetric Dicke state is even, so <Jz> = 0 while the state itself continues to evolve.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Dynamic Synaptic Modulation of LMG Qubits populations in a Bio-Inspired Quantum Brain." pith.science (2026). https://pith.science/paper/LXSQHUTK

@misc{pith2026260216003,
  author       = {Pith},
  title        = {Pith review of: Dynamic Synaptic Modulation of LMG Qubits populations in a Bio-Inspired Quantum Brain},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LXSQHUTK}},
  note         = {Machine review of arXiv:2602.16003}
}

read the original abstract

We present a biologically inspired quantum neural network that encodes neuronal populations as fully connected qubits governed by the Lipkin-Meshkov-Glick (LMG) quantum Hamiltonian and modulated by a synaptic-efficacy feedback implementing activity-dependent changes in the collective time scale. The framework links collective quantum many-body modes and collective-state structure to population homeostasis and rhythmogenesis, outlining scalable computational primitives long-lived operating regimes, activity-dependent oscillation periods, and size-dependent robustness that position LMG-based architectures as promising blueprints for bio-inspired quantum brains on future quantum hardware.

Figures

Figures reproduced from arXiv: 2602.16003 by the authors.

**Figure 2.** Time evolution of the number of excited neuronal qubits [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

**Figure 3.** Time evolution of the number of excited neuronal qubits [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Figures from the paper (6 more)

**Figure 4.** Figure 4: Time evolution of the number of excited neuronal qubits [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

**Figure 5.** Figure 5: Time evolution of the fidelity neuronal qubits for an initial state with (left) no neuronal [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

**Figure 6.** Figure 6: Von Neumann entropy SL time evolution of the neuronal qubits for an initial state with (left) no neuronal qubits excited or all the neuronal qubits excited, (right) around half of the neuronal qubits excited (N/2 − 1), with parameters N = 80.γ = 1, initial sypatic effi…

**Figure 7.** Figure 7: Effect of the synaptic plasticity feedback mechanisms on LMG collective dynamics for [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]

**Figure 8.** Figure 8: Temporal evolution of the linear entropy [PITH_FULL_IMAGE:figures/full_fig_p015_8.png]

**Figure 9.** Figure 9: The effect of synaptic facilitation on the behaviour of the LMG brain model. The figure [PITH_FULL_IMAGE:figures/full_fig_p016_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

57 extracted references · 3 linked inside Pith

[1]

L. F. Abbott and W. G. Regehr. Synaptic computation.Nature, 431(7010):796–803, 2004

2004
[2]

L. F. Abbott, J. A. Varela, K. Sen, and S. B. Nelson. Synaptic depression and cortical gain control. Science, 275(5297):221–224, 1997

1997
[3]

Amit.Modeling Brain Function

Daniel J. Amit.Modeling Brain Function. Cambridge University Press, Cambridge, 2012

2012
[4]

Quantum machine learning.Nature, 549(7671):195–202, Sep 2017

Jacob Biamonte, Peter Wittek, Nicola Pancotti, Patrick Rebentrost, Nathan Wiebe, and Seth Lloyd. Quantum machine learning.Nature, 549(7671):195–202, Sep 2017

2017
[5]

Briegel and Gemma De las Cuevas

Hans J. Briegel and Gemma De las Cuevas. Projective simulation for artificial intelligence.Scientific Reports, 2:400, 2012

2012
[6]

Calixto, R

M. Calixto, R. Romera, and R. del Real. Parity-symmetry-adapted coherent states and entanglement in quantum phase transitions of vibron models.Journal of Physics A: Mathematical and Theoretical, 45:365301, 2012

2012
[7]

Entanglement and quantum phase dia- grams of symmetric multi-qubit systems.Journal of Statistical Mechanics: Theory and Experiment, 2017(10):103103, 2017

Manuel Calixto, Octavio Castaños, and Elvira Romera. Entanglement and quantum phase dia- grams of symmetric multi-qubit systems.Journal of Statistical Mechanics: Theory and Experiment, 2017(10):103103, 2017

2017
[8]

Quantum neuron: An elementary building block for machine learning on quantum computers.arXiv preprint arXiv:1711.11240, 2017

Yudong Cao, Gian Giacomo Guerreschi, and Alán Aspuru-Guzik. Quantum neuron: An elementary building block for machine learning on quantum computers.arXiv preprint arXiv:1711.11240, 2017

arXiv 2017

Show all 57 references

[9]

Castaños, M

O. Castaños, M. Calixto, F. Pérez-Bernal, and E. Romera. Identifying the order of a quantum phase transition by means of wehrl entropy in phase space.Physical Review E, 92:052106, 2015

2015
[10]

Castaños, R

O. Castaños, R. López-Peña, J. Hirsch, and E. López-Moreno. Classical and quantum phase transitions in the lipkin–meshkov–glick model.Physical Review B, 74:104118, 2006

2006
[11]

Cerezo, Andrew Arrasmith, Ryan Babbush, Simon C

M. Cerezo, Andrew Arrasmith, Ryan Babbush, Simon C. Benjamin, Suguru Endo, Keisuke Fujii, Jarrod R. McClean, Kosuke Mitarai, Xiao Yuan, Lukasz Cincio, and Patrick J. Coles. Variational quantum algorithms.Nature Reviews Physics, 3(9):625–644, 2021

2021
[12]

An analytical review of quantum neural network models and relevant research

Simantini Chakraborty, Tamal Das, Saurav Sutradhar, Mrinmoy Das, and Suman Deb. An analytical review of quantum neural network models and relevant research. In2020 5th International Conference on Communication and Electronics Systems (ICCES), pages 1395–1400, 2020. 20

2020
[13]

Variational quantum circuits for deep reinforcement learning.IEEE Access, 8:141007–141024, 2020

Samuel Yen-Chi Chen, Chao-Han Huck Yang, Jun Qi, Pin-Yu Chen, Xiaoli Ma, and Hsi-Sheng Goan. Variational quantum circuits for deep reinforcement learning.IEEE Access, 8:141007–141024, 2020

2020
[14]

Quantum reinforcement learning.IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), 38(5):1207–1220, 2008

Daoyi Dong, Chunlin Chen, Hanxiong Li, and Tzyh-Jong Tarn. Quantum reinforcement learning.IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), 38(5):1207–1220, 2008

2008
[15]

Vedran Dunjko and Hans J. Briegel. Machine learning and artificial intelligence in the quantum domain: a review of recent progress.Reports on Progress in Physics, 81:074001, 2018

2018
[16]

Dusuel and J

S. Dusuel and J. Vidal. Finite-size scaling exponents of the lipkin-meshkov-glick model.Physical Review Letters, 93:237204, 2004

2004
[17]

Fischer, Stefan H

Dietmar G. Fischer, Stefan H. Kienle, and Matthias Freyberger. Quantum-state estimation by self- learning measurements.Phys. Rev. A, 61:032306, Feb 2000

2000
[18]

Spin squeezing with arbitrary quadratic collective-spin interactions.Phys

Zhiyao Hu, Qixian Li, Xuanchen Zhang, Long-Gang Huang, He-bin Zhang, and Yong-Chun Liu. Spin squeezing with arbitrary quadratic collective-spin interactions.Phys. Rev. A, 108(2):023722, August 2023

2023
[19]

Themechanismsandfunctionsofsynapticfacilitation.Neuron, 94(3):447–464, 2017

SkylerL.JackmanandWadeG.Regehr. Themechanismsandfunctionsofsynapticfacilitation.Neuron, 94(3):447–464, 2017

2017
[20]

Jurcevic, B

P. Jurcevic, B. P. Lanyon, P. Hauke, C. Hempel, P. Zoller, R. Blatt, and C. F. Roos. Quasiparticle engineering and entanglement propagation in a quantum many-body system.Nature, 511:202, 2014

2014
[21]

Squeezed spin states.Physical Review A, 47:5138, 1993

Masahiro Kitagawa and Masahito Ueda. Squeezed spin states.Physical Review A, 47:5138, 1993

1993
[22]

An artificial spiking quantum neuron.npj Quantum Information, 7:59, 2021

Lars Bojer Kristensen, Matthias Degroote, Peter Wittek, Alán Aspuru-Guzik, and Nikolaj Thomas Zinner. An artificial spiking quantum neuron.npj Quantum Information, 7:59, 2021

2021
[23]

A. J. Leggett. Bose–einstein condensation in the alkali gases: Some fundamental concepts.Reviews of Modern Physics, 73:307–356, 2001

2001
[24]

H. J. Lipkin, N. Meshkov, and A. J. Glick. Validity of many-body approximation methods for a solvable model: I. exact solutions and perturbation theory.Nuclear Physics, 62:188, 1965

1965
[25]

H. J. Lipkin, N. Meshkov, and A. J. Glick. Validity of many-body approximation methods for a solvable model: Ii. linearization procedures.Nuclear Physics, 62:199, 1965

1965
[26]

H. J. Lipkin, N. Meshkov, and A. J. Glick. Validity of many-body approximation methods for a solvable model: Iii. diagram summations.Nuclear Physics, 62:211, 1965

1965
[27]

Reinforcement learning with quantum variational circuit

Owen Lockwood and Mei Si. Reinforcement learning with quantum variational circuit. InProceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, volume 16, pages 94–100, 2020

2020
[28]

The speed of quantum and classical learning for performing thekth root of not.New Journal of Physics, 11:113018, 11 2009

Daniel Manzano, Marcin Pawłowski, and Časlav Brukner. The speed of quantum and classical learning for performing thekth root of not.New Journal of Physics, 11:113018, 11 2009

2009
[29]

Redistribution of synaptic efficacy between neocortical pyramidal neurons.Nature, 382(6594):807–810, 1996

Henry Markram, Joachim Lübke, Michael Frotscher, and Bert Sakmann. Redistribution of synaptic efficacy between neocortical pyramidal neurons.Nature, 382(6594):807–810, 1996

1996
[30]

Joshua Mautner, Adi Makmal, Daniel Manzano, Markus Tiersch, and Hans J. Briegel. Projective simulation for classical learning agents: A comprehensive investigation.New Generation Computing, 21 33(1):60–79, 2015

2015
[31]

Mejías and Joaquín J

Jorge F. Mejías and Joaquín J. Torres. Maximum memory capacity on neural networks with short-term synaptic depression and facilitation.Neural Computation, 21:851–871, 2009

2009
[32]

Mejías and Joaquín J

Jorge F. Mejías and Joaquín J. Torres. Emergence of resonances in neural systems: the interplay between adaptive threshold and short-term synaptic plasticity.PLoS ONE, 6:e17255, 2011

2011
[33]

Millán, and Joaquín J

Gustavo Menesse, Ana P. Millán, and Joaquín J. Torres. Astrocyte-mediated higher-order control of synaptic plasticity.arXiv preprint arXiv:2507.07693 [q-bio.NC], 2025

2025 arXiv
[34]

Gustavo Menesse and Joaquín J. Torres. Information dynamics of in silico EEG brain waves: Insights into oscillations and functions.PLoS Computational Biology, 20(9):e1012369, 2024

2024
[35]

Micheli, D

A. Micheli, D. Jaksch, J. I. Cirac, and P. Zoller. Many-particle entanglement in two-component bose–einstein condensates.Physical Review A, 67:013607, 2003

2003
[36]

G. J. Milburn, J. Corney, E. M. Wright, and D. F. Walls. Quantum dynamics of an atomic bose–einstein condensate in a double-well potential.Physical Review A, 55:4318–4324, 1997

1997
[37]

Synaptic theory of working memory.Science, 319(5869):1543–1546, 2008

Gianluigi Mongillo, Omri Barak, and Misha Tsodyks. Synaptic theory of working memory.Science, 319(5869):1543–1546, 2008

2008
[38]

Nagy and E

A. Nagy and E. Romera. Fisher information, rényi entropy power and quantum phase transition in the dicke model.Physica A, 391:3650, 2012

2012
[39]

Torres, Hilbert J

Lovorka Pantic, Joaquín J. Torres, Hilbert J. Kappen, and Stan C. A. M. Gielen. Associative memory with dynamic synapses.Neural Computation, 14:2903–2923, 2002

2002
[40]

Wilkinson, Gian Salis, Max Werninghaus, Michael J

Miroslav Pechal, Finn Roy, Samuel A. Wilkinson, Gian Salis, Max Werninghaus, Michael J. Hartmann, and Stefan Filipp. Direct implementation of a perceptron in superconducting circuit quantum hardware. arXiv preprint arXiv:2111.12669, 2021

2021 arXiv
[41]

Torres, and Joaquín Marro

Jorge Pretel, Joaquín J. Torres, and Joaquín Marro. EEGs disclose significant brain activity correlated with synaptic fickleness.Biology, 10(7):647, 2021

2021
[42]

Richerme, Z.-X

P. Richerme, Z.-X. Gong, A. Lee, C. Senko, J. Smith, M. Foss-Feig, S. Michalakis, A. V. Gorshkov, and C. Monroe. Non-local propagation of correlations in quantum systems with long-range interactions. Nature, 511:198, 2014

2014
[43]

Springer, New York, 1980

Peter Ring and Peter Schuck.The Nuclear Many-Body Problem. Springer, New York, 1980

1980
[44]

Romera, M

E. Romera, M. Calixto, and A. Nagy. Entropic uncertainty and the quantum phase transition in the dicke model.Europhysics Letters, 97:20011, 2012

2012
[45]

Garrahan, Igor Lesanovsky, and Markus Müller

Pietro Rotondo, Matteo Marcuzzi, Juan P. Garrahan, Igor Lesanovsky, and Markus Müller. Open quan- tum generalisation of hopfield neural networks.Journal of Physics A: Mathematical and Theoretical, 51(11):115301, 2018

2018
[46]

Charles F. Stevens. Facilitation and depression at single central synapses.Neuron, 14(4):795–802, 1995

1995
[47]

An artificial neuron implemented on an actual quantum processor.npj Quantum Information, 5:26, 2019

Francesco Tacchino, Chiara Macchiavello, Dario Gerace, and Daniele Bajoni. An artificial neuron implemented on an actual quantum processor.npj Quantum Information, 5:26, 2019. 22

2019
[48]

A model of interacting quantum neurons with a dynamic synapse.New Journal of Physics, 24:073007, 7 2022

J J Torres and D Manzano. A model of interacting quantum neurons with a dynamic synapse.New Journal of Physics, 24:073007, 7 2022

2022
[49]

Torres, Jesus M

Joaquín J. Torres, Jesus M. Cortes, Joaquín Marro, and Hilbert J. Kappen. Competition between synaptic depression and facilitation in attractor neural networks.Neural Computation, 19:2739–2755, 2007

2007
[50]

Torres and Hilbert J

Joaquín J. Torres and Hilbert J. Kappen. Emerging phenomena in neural networks with dynamic synapses and their computational implications.Frontiers in Computational Neuroscience, 7:30, 2013

2013
[51]

Dissipative quantum hopfield network: a numerical analysis

Joaquín J Torres and Daniel Manzano. Dissipative quantum hopfield network: a numerical analysis. New Journal of Physics, 26:103018, 10 2024

2024
[52]

Torres, Lovorca Pantic, and Hilbert J

Joaquín J. Torres, Lovorca Pantic, and Hilbert J. Kappen. Storage capacity of attractor neural networks with depressing synapses.Physical Review E, 66:061910, 2002

2002
[53]

Torrontegui and J

E. Torrontegui and J. J. García-Ripoll. Unitary quantum perceptron as efficient universal approximator. EPL (Europhysics Letters), 125:30004, 3 2019

2019
[54]

Neural networks with dynamic synapses.Neural Computation, 10(4):821–835, 1998

Misha Tsodyks, Klaus Pawelzik, and Henry Markram. Neural networks with dynamic synapses.Neural Computation, 10(4):821–835, 1998

1998
[55]

Nathan Wiebe, Ashish Kapoor, and Krysta M. Svore. Quantum perceptron models. InProceedings of the 30th Conference on Neural Information Processing Systems (NeurIPS 2016), 2016

2016
[56]

Parameter estimation in quantum sensing based on deep reinforcement learning.npj Quantum Information, 8(1):2, 2022

Tong Xiao, Jie Fan, and Guihua Zeng. Parameter estimation in quantum sensing based on deep reinforcement learning.npj Quantum Information, 8(1):2, 2022

2022
[57]

Zucker and Wade G

Robert S. Zucker and Wade G. Regehr. Short-term synaptic plasticity.Annual Review of Physiology, 64(1):355–405, 2002

2002

Pith tools

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

[1] [1]

L. F. Abbott and W. G. Regehr. Synaptic computation.Nature, 431(7010):796–803, 2004

2004

[2] [2]

L. F. Abbott, J. A. Varela, K. Sen, and S. B. Nelson. Synaptic depression and cortical gain control. Science, 275(5297):221–224, 1997

1997

[3] [3]

Amit.Modeling Brain Function

Daniel J. Amit.Modeling Brain Function. Cambridge University Press, Cambridge, 2012

2012

[4] [4]

Quantum machine learning.Nature, 549(7671):195–202, Sep 2017

Jacob Biamonte, Peter Wittek, Nicola Pancotti, Patrick Rebentrost, Nathan Wiebe, and Seth Lloyd. Quantum machine learning.Nature, 549(7671):195–202, Sep 2017

2017

[5] [5]

Briegel and Gemma De las Cuevas

Hans J. Briegel and Gemma De las Cuevas. Projective simulation for artificial intelligence.Scientific Reports, 2:400, 2012

2012

[6] [6]

Calixto, R

M. Calixto, R. Romera, and R. del Real. Parity-symmetry-adapted coherent states and entanglement in quantum phase transitions of vibron models.Journal of Physics A: Mathematical and Theoretical, 45:365301, 2012

2012

[7] [7]

Entanglement and quantum phase dia- grams of symmetric multi-qubit systems.Journal of Statistical Mechanics: Theory and Experiment, 2017(10):103103, 2017

Manuel Calixto, Octavio Castaños, and Elvira Romera. Entanglement and quantum phase dia- grams of symmetric multi-qubit systems.Journal of Statistical Mechanics: Theory and Experiment, 2017(10):103103, 2017

2017

[8] [8]

Quantum neuron: An elementary building block for machine learning on quantum computers.arXiv preprint arXiv:1711.11240, 2017

Yudong Cao, Gian Giacomo Guerreschi, and Alán Aspuru-Guzik. Quantum neuron: An elementary building block for machine learning on quantum computers.arXiv preprint arXiv:1711.11240, 2017

arXiv 2017