Pith. sign in

REVIEW 4 major objections 4 minor 35 references

Minimal Neuron Circuits -- Part I: Resonators

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Type-N negative differential resistance is proposed as the functional essence of a sodium channel, and three minimal resonator circuits built from NNDR blocks plus a MOSFET and RC pair are shown to reproduce the $I_{Na,p}+I_K$ spiking…

desk verdict Three new resonator circuits that plausibly replicate I_Na,p+I_K dynamics using NNDR devices, but the universal 'any NNDR device' claim is unsupported and the efficiency argument lacks measurements. read the letter →

arxiv 2506.02341 v1 pith:XGF2CIZ4 submitted 2025-06-03 cs.NE cs.AR

Amr Nabil , T. Nandha Kumar , Haider Abbas F. Almurib This is my paper
classification cs.NEcs.AR
keywords spikingneuralnetworksminimalneuroncircuitsresonatorneuronstype-NnegativedifferentialresistanceINap+IKmodelbifurcationanalysisneuromorphicunipolarmemristor
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 aims to establish a design recipe for minimal spiking neuron circuits: take any circuit or device whose current–voltage curve has a type-N negative differential resistance (NNDR) branch, bias it into that branch, and let it play the sodium channel, while a single low-threshold MOSFET with an RC delay plays the potassium channel. Using that recipe, the paper presents three resonator-type circuits and argues that they implement the $I_{Na,p}+I_K$ neuron model, not the full Hodgkin–Huxley model. The payoff is a neuron built from only three transistors, two capacitors, and one resistor—or one transistor, one memristor, two capacitors, and one resistor—that still shows subthreshold oscillations, signal gain, sustained spiking, and afterhyperpolarization. The paper's stated purpose is methodological: to show that the NNDR property is the core qualitative feature that makes a sodium channel, so the result is a transferable equivalence rather than a one-off circuit.

What carries the argument

The carrying object is the type-N negative differential resistance (NNDR) property, defined as a segment of the I–V characteristic in which current decreases as voltage increases, giving the curve an N shape. The paper treats this property as the functional signature of a sodium channel and uses it to supply the positive feedback needed for an action potential upstroke. Around this element it places the potassium-channel surrogate: a MOSFET with a low threshold whose gate voltage is delayed by an RC network, providing the slow negative feedback that produces the downstroke and afterhyperpolarization. The argument is carried by comparing each circuit's nullclines and bifurcation diagrams with those of the $I_{Na,p}+I_K$ model, using the Shichman–Hodges transistor equations for the numerical state-space solution.

What would settle it

Simulate or measure the proposed circuits with an NNDR element whose I–V curve has a negative slope but a significantly narrower, wider, or differently positioned negative-resistance region than the circuits used here; if spiking, subthreshold oscillations, or the Andronov–Hopf bifurcation disappears while the sign of the slope is unchanged, the claim that NNDR alone is the sodium-channel essence fails.

Watch

Extended reading notes

Core claim

The central discovery is that the sign of the sodium channel's negative-resistance slope, packaged as an N-shaped I–V curve, is enough to reproduce the spiking mechanism of the $I_{Na,p}+I_K$ model. The paper postulates that a sodium channel is functionally a type-N negative differential resistance element, then builds three circuits in which different NNDR blocks—two complementary MOSFETs, two complementary JFETs, or a unipolar memristor with its RESET voltage above its SET voltage—are biased into the negative branch. In all three, the negative branch gives the positive feedback that creates the spike upstroke, while the low-threshold MOSFET and RC network supply the delayed potassium-like current that creates the downstroke and afterhyperpolarization. Each circuit's $V_{out}$-nullcline is an inverted-N curve, spiking begins and ends through supercritical Andronov–Hopf bifurcations, and ramp inputs produce oscillations whose amplitude grows from zero, matching the companion model. The claim is therefore that NNDR plus delayed potassium recovery is the complete qualitative recipe for a resonator neuron.

Load-bearing premise

The load-bearing premise is that any device with a negative-resistance branch can be substituted for the sodium channel no matter where that branch sits, how wide it is, or how asymmetric it is; only the sign of the slope is assumed to matter.

Editorial extensions

If this is right

  • Any NNDR device or circuit—transistor pairs, JFET pairs, or unipolar memristors with RESET above SET—can serve as the sodium channel, so the neuron design is not tied to one fabrication technology.
  • The three circuits use fewer components than published leaky-integrate-and-fire, Hodgkin–Huxley, and Morris–Lecar implementations while still providing signal gain, subthreshold oscillations, and afterhyperpolarization.
  • Because the onset and offset of spiking are both supercritical Andronov–Hopf bifurcations, the neurons naturally respond to ramp or frequency-matched inputs with oscillations that grow from zero amplitude, a resonator signature useful for selective input gating.
  • The same two-block recipe—NNDR element plus MOSFET-with-RC as potassium channel—is carried into the companion Part II for integrator neurons, suggesting the methodology generalizes beyond the resonator type.

Reading between the lines

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

  • Extending beyond the paper: the paper never varies the shape, width, or position of the negative-resistance branch, so an immediate test is to swap in NNDR elements with very different I–V geometries; if spiking vanishes when only the geometry changes, the NNDR sign alone is not the whole story.
  • Extending beyond the paper: the recipe predicts that other well-known NNDR devices, such as tunnel diodes, should also produce resonator spiking with the same MOSFET-plus-RC potassium channel, which is an inexpensive experimental check.
  • Extending beyond the paper: if the equivalence holds at the device level, the area and energy cost of a spiking neuron could reduce to essentially one active element plus an RC delay, which would alter how dense neuromorphic arrays are designed.
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.

Referee Report

4 major / 4 minor

Summary. This manuscript proposes a design methodology for compact "resonator" spiking neuron circuits based on the I_Na,p+I_K model. The authors postulate that the sodium channel behaves functionally as a type-N negative-differential-resistance (NNDR) element, and they implement the fast sodium-like current with one of three NNDR realizations: two complementary MOSFETs, two complementary JFETs, or a unipolar memristor model. In each case, a MOSFET with an RC delay plays the role of the slow potassium current. For the three circuits in Figs. 5 and 8, the paper presents nullclines, simulated voltage waveforms, and bifurcation diagrams in the injected-current parameter, aiming to reproduce the qualitative behavior of the I_Na,p+I_K model: subthreshold oscillations, supercritical Andronov-Hopf onset and offset of spiking, and afterhyperpolarization. The paper closes with a component-count comparison against prior neuron circuit implementations.

Significance. If the proposed substitution is valid, the design recipe is attractive: it reduces the sodium channel to a two-terminal element and allows neuron circuits with very small component counts, while the paper convincingly demonstrates that three different NNDR choices can all produce the desired qualitative dynamics. The explicit nullcline and bifurcation analysis is a strength, and the match of the simulated circuits to the I_Na,p+I_K benchmark is clearly presented. However, the central biological postulate is validated only against the same model from which it was inferred, and the efficiency motivation is argued from component count alone. The manuscript also does not ship code, netlists, or experimental data. The contribution is therefore best assessed as a promising design recipe needing additional support rather than a fully established universal construction.

major comments (4)
  1. [Section IV and Section V] Section IV elevates the NNDR property to the "core qualitative characteristic" that contributes to spiking, and Sections V and VI then use three specific NNDR realizations as sodium-channel stand-ins. The paper never derives conditions on the NDR branch (slope magnitude, voltage window, current scale, asymmetry, or hysteresis) under which the augmented circuit is topologically equivalent to the I_Na,p+I_K model; the three examples only show that hand-picked parameters in the negative-slope region produce spiking. Since the central claim in Section I is that an NNDR circuit/device can be used with a MOSFET and RC circuit to implement the I_Na,p+I_K model, the manuscript needs either a parameter-robustness study across NDR branch shapes or a derivation of sufficient conditions. Otherwise the claim should be weakened to existence of particular NNDR-based designs.
  2. [Section IV and Section III] The postulate that sodium channels exhibit type-N NDR is inferred from the steady-state sodium current of the same I_Na,p+I_K model shown in Fig. 1(b), and the circuits are subsequently validated against that same model in Figs. 2, 3, 6, and 7. Consequently, the agreement between the circuits and the model is partly in-sample and does not independently validate the biological premise. I request a validation against an external benchmark (e.g., Hodgkin-Huxley sodium channel I-V curves or experimental data) or, at minimum, a demonstration that a non-NNDR sodium-model element with the same topology fails to spike, which would isolate the role of the NDR branch.
  3. [Abstract and Section VII (Table 1)] The abstract and Impact Statement claim that the proposed approach is "more efficient" and "scalable", but the only quantitative evidence is the component count in Table 1. No energy per spike, area, power, speed, or fan-out data are reported, and no comparison is made against the energy or area of the prior circuits listed in the table. These efficiency claims should either be supported with measurements or simulations of those metrics, or be restricted to component-count and design-simplicity claims.
  4. [Sections V and VI] All evidence for the three proposed circuits is simulated, but the manuscript provides no netlists, no simulation tool or version, no initial conditions, and no code. Equations (8)-(9) and the caption parameters are a useful start, but the figures cannot be independently checked. I ask for a reproducible artifact, such as SPICE netlists or simulation scripts for at least one circuit, as supplementary material.
minor comments (4)
  1. [Section II] The sentence about time constants appears to swap the standard assignments: the fast sodium activation time constant is usually called tau_m and the slow potassium activation time constant tau_n, but the text says the opposite. Please correct this.
  2. [Section V and Fig. 5] The NNDR sodium block is first referred to as comprising Q3 and Q4 and later as Q2 and Q3; since the circuit is stated to contain only three MOSFETs, the labels should be made consistent.
  3. [Section VI and Table 1] The "Biologically Plausible" column in Table 1 is binary and appears to be assigned by the authors; please state the criteria used (e.g., qualitative match to a specific neuron model) or provide a reference for each row.
  4. [Section V, Fig. 6] The caption and text use inconsistent notation for the injected current and component parameters (e.g., mA in Fig. 2 and uA in Fig. 6); a short table of parameter values with consistent units would improve readability.

Circularity Check

1 steps flagged · score 4.0 of 10

The universal NNDR-as-sodium-channel claim is a postulate validated against the same model from which the NNDR property was inferred; the circuits are intentionally biased into the NDR branch, so the qualitative match is partly by construction.

  1. self definitional [Section IV, first paragraph (Type-N Negative Differential Resistance Devices)]
    "From the characteristics in Fig. 1 (b), we can infer that the sodium ion channel actually possesses this NNDR property. The NNDR property is the core qualitative characteristic that contributes to spiking. This is because the NNDR property allows the sodium channel to amplify voltage changes through a positive feedback effect, which is necessary to generate the upstroke of the spike. Therefore, to find a device/circuit that can emulate the function of a sodium channel, that device/circuit must show NNDR."

    The NNDR property is inferred from the very INa,p+IK model (Fig. 1(b)) that the circuits are later said to 'implement'. The paper then treats NNDR as sufficient for a sodium-channel role ('could still potentially mimic a Sodium channel if it exhibits the NNDR property'), so the three circuits are validated by checking that their N-shaped v-nullcline and Hopf bifurcation match the same model. Since the circuits are deliberately biased in the NNDR negative-slope region (Section V: 'the NNDR circuit must be biased in the negative differential region'), the N-shaped nullcline and resulting spiking are consequences of the design rule, not an independent confirmation that arbitrary NNDR devices implement the model.

full rationale

The paper is not heavily circular because the INa,p+IK model is an external benchmark from Izhikevich, and the proposed circuits are not fitted to reproduce a predefined waveform; they are distinct physical networks whose simulated spiking, nullclines, and Hopf bifurcations are qualitatively compared with the model. However, the central load-bearing step — that any NNDR device can serve as the sodium channel — is stated as a postulate and derived from the same model used for validation. The circuits are constructed so that the NNDR element is biased in its negative-resistance region, which by construction produces the N-shaped v-nullcline and the associated spiking regime. Thus the demonstration supports the design recipe but does not independently establish the universality claim across arbitrary NDR branch shapes, current scales, or voltage windows. The self-citation [18] disputing Mott-memristor neurons is not load-bearing for the central derivation, and the lack of code or hardware data is an evidence gap rather than circularity. Overall, partial circularity in the central claim warrants a moderate score.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

No new physical entities are introduced; the load is carried by hand-picked device parameters and two domain assumptions: the NDR-for-sodium analogy and the sufficiency of qualitative similarity. The neuron model, transistor equations and memristor model are taken from cited prior work.

free parameters (4)
  • Bias voltage V_dc = 3.5 V (Fig. 5/6); 4 V (Fig. 8c)
    Chosen by hand to keep the NNDR element in its negative differential resistance region at the resting operating point.
  • RC time constants C1, C2, R1 = C1=5 nF, C2=0.6 or 1 nF, R1=1 MOhm
    Chosen to set the spike timescale and the potassium-channel-like delay; no systematic sweep or optimization is reported.
  • MOSFET model parameters = K_n1=100 uA/V^2, V_t01=0 V, K_p2=40 uA/V^2, V_t02=2 V, etc.
    Selected so Q1 is partially conducting at rest and the complementary pair shows the desired N-shaped I-V; not obtained from a real process design kit.
  • Unipolar memristor model parameters = R_on=100 kOhm, R_off=1 MOhm, alpha=5e10, beta=1e10, V_rst=3 V, V_set=1 V
    Chosen so the memristor I-V has a reset voltage above set voltage and an NDR branch at the operating point.
assumptions (5)
  • ad hoc to paper The steady-state sodium current I_Na = g_Na m_inf(V)(V - E_Na) has a type-N negative differential resistance branch, and this NDR property is the core mechanism of spiking.
    Stated as a postulate in Section IV and used to justify replacing the sodium channel with arbitrary NNDR circuits. It is not derived from channel kinetics in the paper.
  • domain assumption The I_Na,p+I_K model is an adequate two-equation reduction of Hodgkin-Huxley, with the h gate omitted and sodium activation instantaneous.
    Taken from Izhikevich [20] and used as the benchmark; reasonable but not established in this paper.
  • standard math Shichman-Hodges equations accurately describe the MOSFETs in the proposed circuits.
    Used to compute i_DS currents and nullclines; ideal model with no parasitics and no process variation.
  • domain assumption The modified unipolar memristor model (Eqs. 6-7) with V_rst > V_set represents real devices such as ZnO or SiO_x resistive switches.
    The model is from Fang et al. [27] with the inequality flipped; no device measurements in the paper validate it.
  • ad hoc to paper Qualitative similarity of nullclines and bifurcation diagrams between the circuit and the I_Na,p+I_K model is sufficient to establish biological plausibility.
    The paper's main evidence is qualitative similarity of topology and bifurcation; it never quantifies mismatch or tests response statistics.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Minimal Neuron Circuits -- Part I: Resonators." pith.science (2026). https://pith.science/paper/XGF2CIZ4

@misc{pith2026250602341,
  author       = {Pith},
  title        = {Pith review of: Minimal Neuron Circuits -- Part I: Resonators},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XGF2CIZ4}},
  note         = {Machine review of arXiv:2506.02341}
}
abstract

Spiking Neural Networks have earned increased recognition in recent years owing to their biological plausibility and event-driven computation. Spiking neurons are the fundamental building components of Spiking Neural Networks. Those neurons act as computational units that determine the decision to fire an action potential. This work presents a methodology to implement biologically plausible yet scalable spiking neurons in hardware. We show that it is more efficient to design neurons that mimic the $I_{Na,p}+I_{K}$ model rather than the more complicated Hodgkin-Huxley model. We demonstrate our methodology by presenting eleven novel minimal spiking neuron circuits in Parts I and II of the paper. We categorize the neuron circuits presented into two types: Resonators and Integrators. We discuss the methodology employed in designing neurons of the resonator type in Part I, while we discuss neurons of the integrator type in Part II. In part I, we postulate that Sodium channels exhibit type-N negative differential resistance. Consequently, we present three novel minimal neuron circuits that use type-N negative differential resistance circuits or devices as the Sodium channel. Nevertheless, the aim of the paper is not to present a set of minimal neuron circuits but rather the methodology utilized to construct those circuits.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

35 extracted references · 20 canonical work pages

  1. [1]

    Networks of spiking neurons: The third generation of neural network models,

    W. Maass, “Networks of spiking neurons: The third generation of neural network models,” Neural Networks, vol. 10, no. 9, pp. 1659 –1671, Dec. 1997, doi: 10.1016/S0893-6080(97)00011-7

    doi:10.1016/s0893-6080(97)00011-7 1997

  2. [2]

    Spiking neural networks, an introduction,

    J. Vreeken, “Spiking neural networks, an introduction,” 2003

    work page 2003

  3. [3]

    A quantitative description of membrane current and its application to conduction and excitation in nerve,

    A. L. Hodgkin and A. F. Huxley, “A quantitative description of membrane current and its application to conduction and excitation in nerve,” J Physiol, vol. 117, no. 4, p. 500, Aug. 1952, doi: 10.1113/JPHYSIOL.1952.SP004764

    doi:10.1113/jphysiol.1952.sp004764 1952

  4. [4]

    Lapicque’s introduction of the integrate- and-fire model neuron (1907),

    L. F. Abbott, “Lapicque’s introduction of the integrate- and-fire model neuron (1907),” Brain Res Bull, vol. 50, no. 5–6, pp. 303–304, Nov. 1999, doi: 10.1016/S0361- 9230(99)00161-6

    doi:10.1016/s0361- 1907

  5. [5]

    A Review of spiking neuromorphic hardware communication systems,

    A. R. Young, M. Dean, J. S. Plank, and G. S. Rose, “A Review of spiking neuromorphic hardware communication systems,” IEEE Access , vol. 7, pp. 135606–135620, 2019, doi: 10.1109/ACCESS.2019.2941772

    arXiv 2019

  6. [6]

    Perspective: A review on memristive hardware for neuromorphic computation,

    C. Sung, H. Hwang, and I. K. Yoo, “Perspective: A review on memristive hardware for neuromorphic computation,” J Appl Phys, vol. 124, no. 15, p. 151903, Oct. 2018, doi: 10.1063/1.5037835/347883

    work page doi:10.1063/1.5037835/347883 2018

  7. [7]

    Design of Time -Encoded Spiking Neural Networks in 7 -nm CMOS Technology,

    S. Widmer et al. , “Design of Time -Encoded Spiking Neural Networks in 7 -nm CMOS Technology,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 70, no. 9, pp. 3639 –3643, Sep. 2023, doi: 10.1109/TCSII.2023.3277784

    arXiv 2023

  8. [8]

    Neurogrid: A mixed -analog- digital multichip system for large -scale neural simulations,

    B. V. Benjamin et al. , “Neurogrid: A mixed -analog- digital multichip system for large -scale neural simulations,” Proceedings of the IEEE, vol. 102, no. 5, pp. 699 –716, 2014, doi: 10.1109/JPROC.2014.2313565

    work page arXiv 2014

Show all 35 references
  1. [9]

    Spiking Neural Networks Hardware Implementations and Challenges,

    M. Bouvier et al., “Spiking Neural Networks Hardware Implementations and Challenges,” ACM Journal on Emerging Technologies in Computing Systems (JETC), vol. 15, no. 2, Apr. 2019, doi: 10.1145/3304103

    2019 doi

  2. [10]

    Neuromorphic silicon neuron circuits,

    G. Indiveri et al. , “Neuromorphic silicon neuron circuits,” Front Neurosci, vol. 5, no. MAY, p. 9202, May 2011, doi: 10.3389/FNINS.2011.00073/BIBTEX

    2011 arXiv

  3. [11]

    Emerging Memristive Artificial Synapses and Neurons for Energy -Efficient Neuromorphic Computing,

    S. Choi, J. Yang, and G. Wang, “Emerging Memristive Artificial Synapses and Neurons for Energy -Efficient Neuromorphic Computing,” Advanced Materials, vol. 32, no. 51, p. 2004659, Dec. 2020, doi: 10.1002/ADMA.202004659

    2020 doi

  4. [12]

    Fully memristive neural networks for pattern classification with unsupervised learning,

    Z. Wang et al., “Fully memristive neural networks for pattern classification with unsupervised learning,” Nature Electronics 2018 1:2, vol. 1, no. 2, pp. 137–145, Feb. 2018, doi: 10.1038/s41928-018-0023-2

    2018 doi

  5. [13]

    Various Threshold Switching Devices for Integrate and Fire Neuron Applications,

    D. Lee et al., “Various Threshold Switching Devices for Integrate and Fire Neuron Applications,” Adv Electron Mater, vol. 5, no. 9, p. 1800866, Sep. 2019, doi: 10.1002/AELM.201800866

    2019 doi

  6. [14]

    An Artificial Neuron Based on a Threshold Switching Memristor,

    X. Zhang et al. , “An Artificial Neuron Based on a Threshold Switching Memristor,” IEEE Electron Device Letters, vol. 39, no. 2, pp. 308 –311, Feb. 2018, doi: 10.1109/LED.2017.2782752

    2018

  7. [15]

    Low-Power Artificial Neurons Based on Ag/TiN/HfAlOx/Pt Threshold Switching Memristor for Neuromorphic Computing,

    Y. F. Lu et al., “Low-Power Artificial Neurons Based on Ag/TiN/HfAlOx/Pt Threshold Switching Memristor for Neuromorphic Computing,” IEEE Electron Device Letters, vol. 41, no. 8, pp. 1245 –1248, Aug. 2020, doi: 10.1109/LED.2020.3006581

    2020

  8. [16]

    A scalable neuristor built with Mott memristors,

    M. D. Pickett, G. Medeiros -Ribeiro, and R. S. Williams, “A scalable neuristor built with Mott memristors,” Nature Materials 2013 12:2, vol. 12, no. 2, pp. 114–117, Dec. 2012, doi: 10.1038/nmat3510

    2013 doi

  9. [17]

    Biological plausibility and stochasticity in scalable VO2 active memristor neurons,

    W. Yi, K. K. Tsang, S. K. Lam, X. Bai, J. A. Crowell, and E. A. Flores, “Biological plausibility and stochasticity in scalable VO2 active memristor neurons,” Nature Communications 2018 9:1, vol. 9, no. 1, pp. 1 –10, Nov. 2018, doi: 10.1038/s41467 -018- 07052-w

    2018 doi

  10. [18]

    Mott Memristors and Neuronal Ion Channels: A Qualitative Analysis,

    A. Nabil, T. N. Kumar, and H. A. F. Almurib, “Mott Memristors and Neuronal Ion Channels: A Qualitative Analysis,” IEEE J Emerg Sel Top Circuits Syst, vol. 12, no. 4, pp. 762 –773, Dec. 2022, doi: 10.1109/JETCAS.2022.3221735

    2022

  11. [19]

    Minimal Neuron Circuits —Part II: Integrators,

    A. Nabil, T. N. Kumar, and H. A. F. Almurib, “Minimal Neuron Circuits —Part II: Integrators,” IEEE Transactions on Artificial Intelligence, pp. 1–12, 2025, doi: 10.1109/TAI.2025.3572844

    2025

  12. [20]

    E. M. Izhikevich, Dynamical systems in neuroscience : the geometry of excitability and bursting . MIT Press, 2010

    2010

  13. [21]

    [Analysis of the equations of excitable membranes. I. Reduction of the Hodgkins-Huxley equations to a 2d order system].,

    V. I. Krinskiĭ and I. M. Kokoz, “[Analysis of the equations of excitable membranes. I. Reduction of the Hodgkins-Huxley equations to a 2d order system].,” Biofizika, vol. 18, no. 3, pp. 506–511, 1973

    1973

  14. [22]

    Bipolar —JFET— MOSFET Negative Resistance Devices,

    L. O. Chua, J. Yu, and Y. Yu, “Bipolar —JFET— MOSFET Negative Resistance Devices,” IEEE Trans Circuits Syst , vol. 32, no. 1, pp. 46 –61, 1985, doi: 10.1109/TCS.1985.1085599

    1985

  15. [23]

    Unipolar resistive switching of ZnO -single-wire memristors,

    Y. Huang, Y. Luo, Z. Shen, G. Yuan, and H. Zeng, “Unipolar resistive switching of ZnO -single-wire memristors,” Nanoscale Res Lett, vol. 9, no. 1, pp. 1–5, Aug. 2014, doi: 10.1186/1556 -276X-9- 381/FIGURES/4

    2014 doi

  16. [24]

    Resistive switching characteristics and mechanisms in silicon oxide memory devices,

    Y. F. Chang et al., “Resistive switching characteristics and mechanisms in silicon oxide memory devices,” Physical Sciences Reviews , vol. 1, no. 5, May 2019, doi: 10.1515/PSR -2016- 0011/MACHINEREADABLECITATION/RIS

    2019 doi

  17. [25]

    In situ imaging of the conducting filament in a silicon oxide 11 resistive switch,

    J. Yao, L. Zhong, D. Natelson, and J. M. Tour, “In situ imaging of the conducting filament in a silicon oxide 11 resistive switch,” Scientific Reports 2012 2:1 , vol. 2, no. 1, pp. 1–5, Jan. 2012, doi: 10.1038/srep00242

    2012 doi

  18. [26]

    Čerenkov phonon radiation and phonon structure in electron emission, electroluminescence, and current -voltage characteristics of electroformed Al-Al2O3-Ag diodes,

    T. W. Hickmott, “Čerenkov phonon radiation and phonon structure in electron emission, electroluminescence, and current -voltage characteristics of electroformed Al-Al2O3-Ag diodes,” J Appl Phys , vol. 127, no. 16, Apr. 2020, doi: 10.1063/1.5130758/153429

    2020 doi

  19. [27]

    A compact SPICE model of unipolar memristive devices,

    X. Fang, X. Yang, J. Wu, and X. Yi, “A compact SPICE model of unipolar memristive devices,” IEEE Trans Nanotechnol, vol. 12, no. 5, pp. 843 –850, 2013, doi: 10.1109/TNANO.2013.2275178

    2013

  20. [28]

    A bio -physically inspired silicon neuron,

    E. Farquhar and P. Hasler, “A bio -physically inspired silicon neuron,” IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 52, no. 3, pp. 477–488, 2005, doi: 10.1109/TCSI.2004.842871

    2005

  21. [29]

    A 4 -fJ/spike artificial neuron in 65 nm CMOS technology,

    I. Sourikopoulos et al., “A 4 -fJ/spike artificial neuron in 65 nm CMOS technology,” Front Neurosci, vol. 11, no. MAR, p. 247370, Mar. 2017, doi: 10.3389/FNINS.2017.00123/BIBTEX

    2017

  22. [30]

    Modeling and Simulation of Insulated -Gate Field -Effect Transistor Switching Circuits,

    H. Shichman and D. A. Hodges, “Modeling and Simulation of Insulated -Gate Field -Effect Transistor Switching Circuits,” IEEE J Solid -State Circuits, vol. 3, no. 3, pp. 285 –289, 1968, doi: 10.1109/JSSC.1968.1049902

    1968

  23. [31]

    Compact silicon neuron circuit with spiking and bursting behaviour,

    J. H. B. Wijekoon and P. Dudek, “Compact silicon neuron circuit with spiking and bursting behaviour,” Neural Networks, vol. 21, no. 2 –3, pp. 524–534, Mar. 2008, doi: 10.1016/J.NEUNET.2007.12.037

    2008 doi

  24. [32]

    Hodgkin -Huxley Neuron and FPAA Dynamics,

    A. Natarajan and J. Hasler, “Hodgkin -Huxley Neuron and FPAA Dynamics,” IEEE Trans Biomed Circuits Syst, vol. 12, no. 4, pp. 918 –926, Aug. 2018, doi: 10.1109/TBCAS.2018.2837055

    2018

  25. [33]

    Hardware spiking neurons design: Analog or digital?,

    A. Joubert, B. Belhadj, O. Temam, and R. Héliot, “Hardware spiking neurons design: Analog or digital?,” Proceedings of the International Joint Conference on Neural Networks , 2012, doi: 10.1109/IJCNN.2012.6252600

    2012

  26. [34]

    A low -power adaptive integrate -and-fire neuron circuit,

    G. Indiveri, “A low -power adaptive integrate -and-fire neuron circuit,” Proceedings - IEEE International Symposium on Circuits and Systems, vol. 4, 2003, doi: 10.1109/ISCAS.2003.1206342

    2003 arXiv

  27. [35]

    A 0.3 -V Conductance-Based Silicon Neuron in 0.18 μm CMOS Process,

    M. Akbari, S. M. Hussein, T. I. Chou, and K. T. Tang, “A 0.3 -V Conductance-Based Silicon Neuron in 0.18 μm CMOS Process,” IEEE Transactions on Circuits and Systems II: Express Briefs , vol. 68, no. 10, pp. 3209–3213, Oct. 2021, doi: 10.1109/TCSII.2021.3073838

    2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.