pith. sign in

arxiv: 2605.16114 · v1 · pith:3222RQAVnew · submitted 2026-05-15 · 💻 cs.NE · cs.LG

Scalable neuromorphic computing from autonomous spiking dynamics in a clockless reconfigurable chip

Eric Oliveira Gomes , Damien Rontani This is my paper

Pith reviewed 2026-05-19 17:51 UTC · model grok-4.3

classification 💻 cs.NE cs.LG
keywords neuromorphic computingspiking neuronsclockless FPGAasynchronous circuitsaudio classificationlow-power hardwarereconfigurable chipsBoolean spiking networks
0
0 comments X

The pith

Clockless asynchronous circuits on standard FPGAs generate autonomous spiking dynamics that solve machine-learning tasks at competitive accuracy with low power.

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

This paper establishes that spiking neuron networks can emerge directly from the natural time-continuous evolution of clockless digital logic on ordinary commercial FPGAs. The authors configure these dynamics through adjustable excitatory and inhibitory synaptic weights to process spike-encoded inputs and perform classification. On an audio classification benchmark the resulting system reaches accuracy levels comparable to other neuromorphic approaches while drawing substantially less power than conventional digital hardware. The work positions reconfigurable chips as a practical route to energy-efficient, quasi-analog neuromorphic processors without requiring custom analog fabrication.

Core claim

Autonomous spiking dynamics arise from the asynchronous, time-continuous evolution of clockless digital circuits on FPGAs; when these circuits are arranged as networks of Boolean spiking neurons whose excitatory and inhibitory synaptic weights are set appropriately, the resulting system handles spike-encoded data and achieves competitive performance on machine-learning tasks such as audio classification while consuming significantly less power than traditional digital implementations.

What carries the argument

Networks of interacting Boolean spiking neurons whose excitatory and inhibitory synaptic weights are configured on clockless asynchronous digital circuits whose time-continuous evolution produces the spiking activity.

If this is right

  • Spike-based encoding combined with the inherent high-speed evolution of asynchronous circuits yields fast processing for sensory data.
  • Power consumption drops markedly relative to clocked digital designs that perform equivalent neural computations.
  • Reconfigurable commercial hardware can serve as an intermediate platform between conventional digital processors and dedicated analog neuromorphic chips.
  • Clockless digital circuits become a viable substrate for neuromorphic computing without specialized fabrication steps.

Where Pith is reading between the lines

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

  • Existing FPGA development boards could support rapid prototyping and deployment of neuromorphic algorithms in field settings.
  • The same reconfigurability might allow the hardware to switch between different sensory modalities or to incorporate limited on-chip adaptation rules.
  • Further efficiency gains could appear if the approach is combined with other asynchronous or event-driven sensor interfaces.

Load-bearing premise

Synaptic weights can be chosen so that the autonomous spiking dynamics of the clockless circuits remain stable and accurate when scaled to solve machine-learning tasks.

What would settle it

The FPGA implementation fails to reach competitive accuracy on the audio classification task or produces unstable spiking patterns once the network size or input rate is increased.

Figures

Figures reproduced from arXiv: 2605.16114 by Damien Rontani, Eric Oliveira Gomes.

Figure 1
Figure 1. Figure 1: Overview of the spiking Boolean neuron architecture. (a) Internal structure of the Boolean neurons depicting the interconnection between the two functional blocks implementing the soma and synapses (b) Detailed diagram of the Boolean neuron circuit. The soma module is composed of a bidirectional asynchronous counter module (ACM), modeling membrane depolarization and allowing for excitation and inhibition, … view at source ↗
Figure 2
Figure 2. Figure 2: Overview of a typical B-SNN comprising 196 neurons arranged in a 7×7×4 grid, driven by spiking stimuli. (a) Raster plot over 3.2µs of a spiking input sample representing the spoken English digit “one”. The highlighted region illustrates a representative spike with a full width at half maximum (FWHM) of 2.07ns, measured using a high-speed oscilloscope. (b) Raster plot of the corresponding spiking output, ob… view at source ↗
Figure 3
Figure 3. Figure 3: System pipeline and classification performance of a B-SNN for voice recognition. (a) Overview of the complete experimental pipeline, consisting of event-driven input transmission to the FPGA, retrieval of Boolean-SNN spike events by a custom-built multi-channel time-tagger, and subsequent feature encoding for classification. (b) Accuracies obtained using features derived from rate encoding, latency encodin… view at source ↗
read the original abstract

We propose a scalable neuromorphic architecture based on spiking dynamics emerging from the autonomous time-continuous evolution of clockless (asynchronous) digital circuits. Implemented on commercially available field-programmable gate arrays (FPGAs), our system implements networks of interacting Boolean spiking neurons with configurable excitatory and inhibitory synaptic weights. A complete processing pipeline enables efficient handling of spike-encoded data for solving machine-learning tasks. We demonstrate competitive performance for an audio classification task with spike-based encoding and high-speed processing. Power consumption is significantly lower than traditional digital implementations; this makes our approach an efficient alternative that bridges the gap to dedicated analog neuromorphic systems without the need for specialized hardware design. More generally, our approach establishes clockless digital hardware as a viable platform for neuromorphic computing. It paves the way for reconfigurable chips to be turned into energy-efficient quasi-analog neuromorphic processors.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 2 minor

Summary. The manuscript proposes a scalable neuromorphic architecture that exploits autonomous spiking dynamics arising from the time-continuous evolution of clockless (asynchronous) Boolean circuits implemented on commercial FPGAs. Networks of interacting spiking neurons are configured solely through excitatory and inhibitory synaptic weights; a complete spike-encoded processing pipeline is described and applied to an audio classification task, with claims of competitive accuracy, high-speed operation, and substantially lower power than conventional digital implementations.

Significance. If the central claims are substantiated, the work would demonstrate that clockless digital fabrics on off-the-shelf FPGAs can be turned into reconfigurable, energy-efficient neuromorphic processors without custom analog hardware, thereby bridging digital reconfigurable platforms and dedicated neuromorphic systems.

major comments (2)
  1. [Abstract and §4] Abstract and §4 (audio-classification results): the manuscript asserts 'competitive performance' and 'significantly lower' power but supplies no numerical accuracy figures, baseline comparisons, error bars, dataset size, or power measurements in the provided text; without these data the central empirical claim cannot be evaluated.
  2. [§3 and §5] §3 (network implementation) and §5 (scaling discussion): the assumption that propagation delays, routing skew, and hazards in the asynchronous FPGA fabric remain benign when weights are set for a concrete ML objective is load-bearing, yet no quantitative characterization (e.g., measured spike-timing jitter, accuracy versus network size, or comparison against a clocked baseline on the same fabric) is reported for the audio task.
minor comments (2)
  1. [§2] Clarify the precise Boolean neuron model and the mapping from synaptic weights to FPGA LUT/RAM resources.
  2. Add a table or figure that directly compares power and latency against at least one clocked digital baseline on the same FPGA device.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and the opportunity to improve the manuscript. We address each major comment below and will revise the paper accordingly to strengthen the presentation of results and supporting characterizations.

read point-by-point responses

  1. Referee: [Abstract and §4] Abstract and §4 (audio-classification results): the manuscript asserts 'competitive performance' and 'significantly lower' power but supplies no numerical accuracy figures, baseline comparisons, error bars, dataset size, or power measurements in the provided text; without these data the central empirical claim cannot be evaluated.

    Authors: We agree that the abstract and §4 would benefit from explicit numerical reporting to allow readers to evaluate the claims directly from the text. Although the results are supported by figures in §4, we will revise both the abstract and §4 to include specific accuracy values with error bars, dataset size and split details, baseline comparisons, and quantitative power measurements (including the measurement methodology). revision: yes

  2. Referee: [§3 and §5] §3 (network implementation) and §5 (scaling discussion): the assumption that propagation delays, routing skew, and hazards in the asynchronous FPGA fabric remain benign when weights are set for a concrete ML objective is load-bearing, yet no quantitative characterization (e.g., measured spike-timing jitter, accuracy versus network size, or comparison against a clocked baseline on the same fabric) is reported for the audio task.

    Authors: We acknowledge that additional quantitative evidence would strengthen the claims regarding timing stability under ML-derived weights. The current manuscript relies on observed functional stability for the audio task, but we will add measured spike-timing jitter statistics, accuracy as a function of network size, and a direct comparison against a clocked baseline implemented on the same FPGA fabric. These data will be incorporated into a revised §3 and an expanded §5. revision: yes

Circularity Check

0 steps flagged

No circularity in derivation; claims rest on hardware implementation and empirical demonstration

full rationale

The paper describes a neuromorphic architecture implemented on commercial FPGAs using clockless asynchronous digital circuits to generate autonomous spiking dynamics, configured via excitatory/inhibitory weights for an audio classification task. No equations, fitted parameters, or first-principles derivations appear in the abstract or described claims that reduce by construction to self-defined inputs or predictions. Performance results are presented as experimental outcomes from the physical FPGA realization rather than statistical fits or self-referential models. The central premise is supported by direct hardware behavior and task accuracy measurements, which are externally falsifiable and independent of any internal redefinition or self-citation chain. This is a standard non-circular finding for an implementation-focused hardware paper.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption that clockless digital circuits naturally produce usable spiking dynamics and on the free parameter of configurable synaptic weights; no new physical entities are postulated.

free parameters (1)
  • excitatory and inhibitory synaptic weights
    Weights are described as configurable parameters that must be set to realize functional networks for the audio task.
axioms (1)
  • domain assumption Spiking dynamics emerge autonomously from the time-continuous evolution of clockless digital circuits.
    This premise is invoked to justify the entire neuromorphic architecture without a clock.

pith-pipeline@v0.9.0 · 5677 in / 1316 out tokens · 48574 ms · 2026-05-19T17:51:36.469262+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

58 extracted references · 58 canonical work pages · 5 internal anchors

  1. [1]

    & Rückert, U

    Ostrau, C., Klarhorst, C., Thies, M. & Rückert, U. Benchmarking Neuromorphic Hardware and Its Energy Expenditure. Front. Neurosci.16, 873935, DOI: 10.3389/fnins.2022.873935 (2022)

    work page doi:10.3389/fnins.2022.873935 2022

  2. [2]

    Pal, A.et al.An ultra energy-efficient hardware platform for neuromorphic computing enabled by 2D-TMD tunnel-FETs. Nat. Commun.15, 3392, DOI: 10.1038/s41467-024-46397-3 (2024)

    work page doi:10.1038/s41467-024-46397-3 2024

  3. [3]

    & Haas, H

    Jaeger, H. & Haas, H. Harnessing Nonlinearity: Predicting Chaotic Systems and Saving Energy in Wireless Communication. Science304, 78–80, DOI: 10.1126/science.1091277 (2004)

    work page doi:10.1126/science.1091277 2004

  4. [4]

    Real-time computing without stable states: A new framework for neural computation based on perturbations.Neural Computation, 14(11):2531–2560, 2002

    Maass, W., Natschläger, T. & Markram, H. Real-time computing without stable states: A new framework for neural computation based on perturbations.Neural Comput.14, 2531–2560, DOI: 10.1162/089976602760407955 (2002)

    work page doi:10.1162/089976602760407955 2002

  5. [5]

    & Jaeger, H

    Lukoševiˇcius, M. & Jaeger, H. Reservoir computing approaches to recurrent neural network training.Comput. Sci. Rev.3, 127–149, DOI: 10.1016/j.cosrev.2009.03.005 (2009)

    work page doi:10.1016/j.cosrev.2009.03.005 2009

  6. [6]

    Lugnan, A.et al.Photonic neuromorphic information processing and reservoir computing.APL Photonics5, DOI: 10.1063/1.5129762 (2020)

    work page doi:10.1063/1.5129762 2020

  7. [7]

    Commun.2, DOI: 10.1038/ncomms1476 (2011)

    Appeltant, L.et al.Information processing using a single dynamical node as complex system.Nat. Commun.2, DOI: 10.1038/ncomms1476 (2011)

    work page doi:10.1038/ncomms1476 2011

  8. [8]

    Commun.8, DOI: 10.1038/s41467-017-02337-y (2017)

    Du, C.et al.Reservoir computing using dynamic memristors for temporal information processing.Nat. Commun.8, DOI: 10.1038/s41467-017-02337-y (2017)

    work page doi:10.1038/s41467-017-02337-y 2017

  9. [9]

    Neuromorphic computing with nanoscale spintronic oscillators

    Torrejon, J.et al.Neuromorphic computing with nanoscale spintronic oscillators.Nature547, 428–431, DOI: 10.1038/ nature23011 (2017). 1701.07715

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  10. [10]

    Vowel recognition with four coupled spin-torque nano-oscillators

    Romera, M.et al.V owel recognition with four coupled spin-torque nano-oscillators.Nature563, 230–234, DOI: 10.1038/s41586-018-0632-y (2018). 1711.02704

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1038/s41586-018-0632-y 2018

  11. [11]

    & Rontani, D

    Antonik, P., Marsal, N. & Rontani, D. Large-scale spatiotemporal photonic reservoir computer for image classification. IEEE J. Sel. Top. Quantum Electron.26, 1–12, DOI: 10.1109/JSTQE.2019.2924138 (2020)

    work page doi:10.1109/jstqe.2019.2924138 2019

  12. [12]

    & Ricciardi, C

    Milano, G., Agliuzza, M., De Leo, N. & Ricciardi, C. Speech recognition through physical reservoir computing with neuromorphic nanowire networks. In2022 International Joint Conference on Neural Networks (IJCNN), 1–6, DOI: 10.1109/IJCNN55064.2022.9892078 (IEEE, Padua, Italy, 2022). 12/15

    work page doi:10.1109/ijcnn55064.2022.9892078 2022

  13. [13]

    Shahi, S., Fenton, F. H. & Cherry, E. M. Prediction of chaotic time series using recurrent neural networks and reservoir computing techniques: A comparative study.Mach. Learn. with Appl.8, 100300, DOI: 10.1016/j.mlwa.2022.100300 (2022)

    work page doi:10.1016/j.mlwa.2022.100300 2022

  14. [14]

    K., Bulsara, D

    Yamazaki, K., V o-Ho, V . K., Bulsara, D. & Le, N. Spiking Neural Networks and Their Applications: A Review.Brain Sci. 12, DOI: 10.3390/brainsci12070863 (2022)

    work page doi:10.3390/brainsci12070863 2022

  15. [15]

    Izhikevich, E. M. Neural excitability, spiking and bursting.Int. J. Bifurc. Chaos10, 1171–1266, DOI: 10.1142/ S0218127400000840 (2000). https://doi.org/10.1142/S0218127400000840. 16.Moosavi, S. A., Montakhab, A. & Valizadeh, A. Refractory period in network models of excitable nodes: self-sustaining stable dynamics, extended scaling region and oscillatory ...

    work page doi:10.1142/s0218127400000840 2000

  16. [16]

    Chudnovsky and S

    Isaacson, J. & Scanziani, M. How inhibition shapes cortical activity.Neuron72, 231–243, DOI: https://doi.org/10.1016/j. neuron.2011.09.027 (2011)

    work page doi:10.1016/j 2011

  17. [17]

    Song, S., Miller, K. D. & Abbott, L. F. Competitive Hebbian learning through spike-timing-dependent synaptic plasticity. Nat. Neurosci.3, 919–926, DOI: 10.1038/78829 (2000)

    work page doi:10.1038/78829 2000

  18. [18]

    M., Naud, R

    Gerstner, W., Kistler, W. M., Naud, R. & Paninski, L.Neuronal Dynamics: From Single Neurons to Networks and Models of Cognition(Cambridge University Press, 2014)

    work page 2014

  19. [19]

    Sangwan, V . K. & Hersam, M. C. Neuromorphic nanoelectronic materials.Nat. Nanotechnol.15, 517–528, DOI: 10.1038/s41565-020-0647-z (2020)

    work page doi:10.1038/s41565-020-0647-z 2020

  20. [20]

    Commun.15, 3446, DOI: 10.1038/s41467-024-47764-w (2024)

    D’Agostino, S.et al.DenRAM: neuromorphic dendritic architecture with RRAM for efficient temporal processing with delays.Nat. Commun.15, 3446, DOI: 10.1038/s41467-024-47764-w (2024)

    work page doi:10.1038/s41467-024-47764-w 2024

  21. [21]

    & Panda, P

    Roy, K., Jaiswal, A. & Panda, P. Towards spike-based machine intelligence with neuromorphic computing.Nature575, 607–617, DOI: 10.1038/s41586-019-1677-2 (2019)

    work page doi:10.1038/s41586-019-1677-2 2019

  22. [22]

    Davies, M.et al.Loihi: A Neuromorphic Manycore Processor with On-Chip Learning.IEEE Micro38, 82–99, DOI: 10.1109/MM.2018.112130359 (2018)

    work page doi:10.1109/mm.2018.112130359 2018

  23. [23]

    IEEE Transactions on Comput

    Akopyan, F.et al.TrueNorth: Design and Tool Flow of a 65 mW 1 Million Neuron Programmable Neurosynaptic Chip. IEEE Transactions on Comput. Des. Integr. Circuits Syst.34, 1537–1557, DOI: 10.1109/TCAD.2015.2474396 (2015)

    work page doi:10.1109/tcad.2015.2474396 2015

  24. [24]

    B., Galluppi, F., Temple, S

    Furber, S. B., Galluppi, F., Temple, S. & Plana, L. A. The SpiNNaker Project.Proc. IEEE102, 652–665, DOI: 10.1109/JPROC.2014.2304638 (2014)

    work page doi:10.1109/jproc.2014.2304638 2014

  25. [25]

    Solid-State Circuits55, 2228–2246, DOI: 10.1109/JSSC.2020.2970709 (2020)

    Deng, L.et al.Tianjic: A Unified and Scalable Chip Bridging Spike-Based and Continuous Neural Computation.IEEE J. Solid-State Circuits55, 2228–2246, DOI: 10.1109/JSSC.2020.2970709 (2020)

    work page doi:10.1109/jssc.2020.2970709 2020

  26. [26]

    V .et al.Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations.Proc

    Benjamin, B. V .et al.Neurogrid: A Mixed-Analog-Digital Multichip System for Large-Scale Neural Simulations.Proc. IEEE102, 699–716, DOI: 10.1109/JPROC.2014.2313565 (2014)

    work page doi:10.1109/jproc.2014.2313565 2014

  27. [27]

    Neurosci.16, 795876, DOI: 10.3389/fnins.2022.795876 (2022)

    Pehle, C.et al.The BrainScaleS-2 Accelerated Neuromorphic System With Hybrid Plasticity.Front. Neurosci.16, 795876, DOI: 10.3389/fnins.2022.795876 (2022). 29.Markovi ´c, D., Mizrahi, A., Querlioz, D. & Grollier, J. Physics for neuromorphic computing.Nat. Rev. Phys.2, 499–510, DOI: 10.1038/s42254-020-0208-2 (2020)

    work page doi:10.3389/fnins.2022.795876 2022

  28. [28]

    J.et al.Photonics for artificial intelligence and neuromorphic computing.Nat

    Shastri, B. J.et al.Photonics for artificial intelligence and neuromorphic computing.Nat. Photonics15, 102–114, DOI: 10.1038/s41566-020-00754-y (2021)

    work page doi:10.1038/s41566-020-00754-y 2021

  29. [29]

    R., Shastri, B

    Prucnal, P. R., Shastri, B. J., Ferreira De Lima, T., Nahmias, M. A. & Tait, A. N. Recent progress in semiconductor excitable lasers for photonic spike processing.Adv. Opt. Photonics8, 228, DOI: 10.1364/AOP.8.000228 (2016)

    work page doi:10.1364/aop.8.000228 2016

  30. [30]

    & Brunner, D

    Talukder, R., Skalli, A., Porte, X., Thorpe, S. & Brunner, D. A spiking photonic neural network of 40 000 neurons, trained with latency and rank-order coding for leveraging sparsity.Neuromorphic Comput. Eng.5, 034003, DOI: 10.1088/2634-4386/addee7 (2025)

    work page doi:10.1088/2634-4386/addee7 2025

  31. [31]

    & Prodromakis, T

    Indiveri, G., Linares-Barranco, B., Legenstein, R., Deligeorgis, G. & Prodromakis, T. Integration of nanoscale memristor synapses in neuromorphic computing architectures.Nanotechnology24, 384010, DOI: 10.1088/0957-4484/24/38/384010 (2013)

    work page doi:10.1088/0957-4484/24/38/384010 2013

  32. [32]

    Electron.1, 137–145, DOI: 10.1038/s41928-018-0023-2 (2018)

    Wang, Z.et al.Fully memristive neural networks for pattern classification with unsupervised learning.Nat. Electron.1, 137–145, DOI: 10.1038/s41928-018-0023-2 (2018). 13/15

    work page doi:10.1038/s41928-018-0023-2 2018

  33. [33]

    Z., Ahmadi, A

    Farsa, E. Z., Ahmadi, A. & Keszocze, O. Reconfigurable Digital FPGA Implementations for Neuromorphic Computing: A Survey on Recent Advances and Future Directions.IEEE Transactions on Emerg. Top. Comput. Intell.9, 3210–3232, DOI: 10.1109/TETCI.2025.3551934 (2025)

    work page doi:10.1109/tetci.2025.3551934 2025

  34. [34]

    Liu, Y ., Yenamachintala, S. S. & Li, P. Energy-efficient fpga spiking neural accelerators with supervised and unsupervised spike-timing-dependent-plasticity.J. Emerg. Technol. Comput. Syst.15, DOI: 10.1145/3313866 (2019)

    work page doi:10.1145/3313866 2019

  35. [35]

    Large-Scale Field-Programmable Analog Arrays.Proc

    Hasler, J. Large-Scale Field-Programmable Analog Arrays.Proc. IEEE108, 1283–1302, DOI: 10.1109/jproc.2019.2950173 (2020)

    work page doi:10.1109/jproc.2019.2950173 2019

  36. [36]

    & Hasler, J

    Natarajan, A. & Hasler, J. Hodgkin–Huxley Neuron and FPAA Dynamics.IEEE Transactions on Biomed. Circuits Syst. 12, 918–926, DOI: 10.1109/TBCAS.2018.2837055 (2018)

    work page doi:10.1109/tbcas.2018.2837055 2018

  37. [37]

    & Hasler, J

    Bhattacharyya, S. & Hasler, J. O. A Six-Transistor Integrate-and-Fire Neuron Enabling Chaotic Dynamics.IEEE Transactions on Biomed. Circuits Syst.19, 1005–1017, DOI: 10.1109/TBCAS.2025.3526762 (2025)

    work page doi:10.1109/tbcas.2025.3526762 2025

  38. [38]

    D., Soriano, M

    Haynes, N. D., Soriano, M. C., Rosin, D. P., Fischer, I. & Gauthier, D. J. Reservoir computing with a single time-delay autonomous Boolean node.Phys. Rev. E91, 020801, DOI: 10.1103/PhysRevE.91.020801 (2015)

    work page doi:10.1103/physreve.91.020801 2015

  39. [39]

    Shani, I.et al.Dynamics of analog logic-gate networks for machine learning.Chaos: An Interdiscip. J. Nonlinear Sci.29, 123130, DOI: 10.1063/1.5123753 (2019)

    work page doi:10.1063/1.5123753 2019

  40. [40]

    Excitability in autonomous Boolean networks

    Rosin, D. P., Rontani, D., Gauthier, D. J. & Schöll, E. Excitability in autonomous Boolean networks.Epl100, DOI: 10.1209/0295-5075/100/30003 (2012). 1208.6181

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1209/0295-5075/100/30003 2012

  41. [41]

    Zhang, R.et al.Boolean chaos.Phys. Rev. E - Stat. Nonlinear , Soft Matter Phys.80, DOI: 10.1103/PhysRevE.80.045202 (2009). 0906.4124

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1103/physreve.80.045202 2009

  42. [42]

    Ultra-Fast Physical Generation of Random Numbers Using Hybrid Boolean Networks

    Rosin, D. P., Rontani, D. & Gauthier, D. J. Ultrafast physical generation of random numbers using hybrid Boolean networks. Phys. Rev. E - Stat. Nonlinear , Soft Matter Phys.87, 57–79, DOI: 10.1103/PhysRevE.87.040902 (2013). 1209.0142

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.1103/physreve.87.040902 2013

  43. [43]

    P., Rontani, D., Gauthier, D

    Rosin, D. P., Rontani, D., Gauthier, D. J. & Schöll, E. Control of synchronization patterns in neural-like boolean networks. Phys. Rev. Lett.110, DOI: 10.1103/PhysRevLett.110.104102 (2013)

    work page doi:10.1103/physrevlett.110.104102 2013

  44. [44]

    DeLong, Ramon Fernandez Mir, and Jacques D

    Cramer, B., Stradmann, Y ., Schemmel, J. & Zenke, F. The Heidelberg Spiking Data Sets for the Systematic Evaluation of Spiking Neural Networks.IEEE Transactions on Neural Networks Learn. Syst.33, 2744–2757, DOI: 10.1109/TNNLS. 2020.3044364 (2022)

    work page doi:10.1109/tnnls 2020

  45. [45]

    & Harvey, R

    Strata, P. & Harvey, R. Dale’s principle.Brain Res. Bull.50, 349–350, DOI: https://doi.org/10.1016/S0361-9230(99) 00100-8 (1999)

    work page doi:10.1016/s0361-9230(99 1999

  46. [46]

    A., Singhal, R

    Biswas, A., Medhe, S. A., Singhal, R. & Ganguly, U. Temporal and Spatial Reservoir Ensembling Techniques for Liquid State Machines, DOI: 10.48550/arXiv.2411.11414 (2024). 2411.11414

    work page doi:10.48550/arxiv.2411.11414 2024

  47. [47]

    L., Schnupp, J

    Jenison, R. L., Schnupp, J. W. H., Reale, R. A. & Brugge, J. F. Auditory Space-Time Receptive Field Dynamics Revealed by Spherical White-Noise Analysis.The J. Neurosci.21, 4408–4415, DOI: 10.1523/JNEUROSCI.21-12-04408.2001 (2001)

    work page doi:10.1523/jneurosci.21-12-04408.2001 2001

  48. [48]

    Ringach, D. L. Mapping receptive fields in primary visual cortex.The J. Physiol.558, 717–728, DOI: 10.1113/jphysiol. 2004.065771 (2004)

    work page doi:10.1113/jphysiol 2004

  49. [49]

    E., Eltawil, A

    Guo, W., Fouda, M. E., Eltawil, A. M. & Salama, K. N. Neural Coding in Spiking Neural Networks: A Comparative Study for Robust Neuromorphic Systems.Front. Neurosci.15, 638474, DOI: 10.3389/fnins.2021.638474 (2021). 52.Bishop, C. M.Pattern Recognition and Machine Learning(Springer, New York, 2006). 53.Pedregosa, F.et al.Scikit-learn: Machine learning in Py...

    work page doi:10.3389/fnins.2021.638474 2021

  50. [50]

    & Botteldooren, D

    Sun, P., Wu, J., Devos, P. & Botteldooren, D. Towards parameter-free attentional spiking neural networks.Neural Networks 185, 107154, DOI: 10.1016/j.neunet.2025.107154 (2025)

    work page doi:10.1016/j.neunet.2025.107154 2025

  51. [51]

    2404.18508

    Schöne, M.et al.Scalable Event-by-event Processing of Neuromorphic Sensory Signals With Deep State-Space Models, DOI: 10.48550/arXiv.2404.18508 (2024). 2404.18508

    work page doi:10.48550/arxiv.2404.18508 2024

  52. [52]

    & Legenstein, R

    Baronig, M., Ferrand, R., Sabathiel, S. & Legenstein, R. Advancing spatio-temporal processing through adaptation in spiking neural networks.Nat. Commun.16, 5776, DOI: 10.1038/s41467-025-60878-z (2025)

    work page doi:10.1038/s41467-025-60878-z 2025

  53. [53]

    & Masquelier, T

    Hammouamri, I., Khalfaoui-Hassani, I. & Masquelier, T. Learning Delays in Spiking Neural Networks using Dilated Convolutions with Learnable Spacings, DOI: 10.48550/arXiv.2306.17670 (2023). 2306.17670. 14/15

    work page doi:10.48550/arxiv.2306.17670 2023

  54. [54]

    J., Van Leekwijck, W

    Deckers, L., Tsang, I. J., Van Leekwijck, W. & Latré, S. Extended liquid state machines for speech recognition.Front. Neurosci.16, 1023470, DOI: 10.3389/fnins.2022.1023470 (2022)

    work page doi:10.3389/fnins.2022.1023470 2022

  55. [55]

    ArXiv:2404.02248 [cs]

    Matinizadeh, S.et al.A Fully-Configurable Open-Source Software-Defined Digital Quantized Spiking Neural Core Architecture, DOI: 10.48550/arXiv.2404.02248 (2024). ArXiv:2404.02248 [cs]

    work page doi:10.48550/arxiv.2404.02248 2024

  56. [56]

    & Carlo, S

    Carpegna, A., Savino, A. & Carlo, S. D. Spiker+: A Framework for the Generation of Efficient Spiking Neural Networks FPGA Accelerators for Inference at the Edge.IEEE Transactions on Emerg. Top. Comput.13, 784–798, DOI: 10.1109/TETC.2024.3511676 (2025)

    work page doi:10.1109/tetc.2024.3511676 2024

  57. [57]

    Liu, D. C. & Nocedal, J. On the limited memory BFGS method for large scale optimization.Math. Program.45, 503–528, DOI: 10.1007/BF01589116 (1989)

    work page doi:10.1007/bf01589116 1989

  58. [58]

    D.et al.Inexpensive electronics and software for photon statistics and correlation spectroscopy.Am

    Gamari, B. D.et al.Inexpensive electronics and software for photon statistics and correlation spectroscopy.Am. J. Phys. 82, DOI: 10.1119/1.4869188 (2014). Acknowledgements The authors would like to gratefully acknowledge the financial support of the Conseil of Region Grand-Est, the AirForce Office for Scientific Research (AFOSR) and Office for Naval Resea...

    work page doi:10.1119/1.4869188 2014