pith. sign in

arxiv: 2604.02552 · v1 · submitted 2026-04-02 · 💻 cs.NE · cs.ET

Computing with Living Neurons: Chaos-Controlled Reservoir Computing with Knowledge Transplant

Seung Hyun Kim , Zhi Dou , Gaurav Upadhyay , Anay Pattanaik , Leo Maslov , Lav Varshney , John Beggs , Howard Gritton
show 1 more author
Mattia Gazzola
This is my paper

Pith reviewed 2026-05-13 20:17 UTC · model grok-4.3

classification 💻 cs.NE cs.ET
keywords reservoir computingliving neuronschaos controlknowledge transplantbiological computationneural culturespattern classification
0
0 comments X

The pith

Chaos-controlled reservoir computing stabilizes living neural cultures for robust pattern classification with 300 percent gains in accuracy and longevity over standard methods, enabling knowledge transplant across cultures.

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

The paper presents chaos-controlled Reservoir Computing (cc-RC) as a way to use living neural cultures for computation despite their inherent variability. It identifies each culture's dynamical signature, applies low-power optical chaos control to stabilize activity into a controllable attractor, and trains the readout layer within this regime. This results in robust learning and classification that improves accuracy and model longevity by roughly 300 percent compared to standard reservoir computing across many samples. The authors also introduce Knowledge Transplant, where a learned reservoir map from one culture is moved to a similar student culture, shortening training to minutes and enhancing results. Overall, the approach supports reusable models that can accumulate knowledge across different neural populations and outlast any single culture's lifespan.

Core claim

By pre-training to identify a living neural culture's dynamical signature and phase-portrait attractor, then using low-power optical chaos control to stabilize its spontaneous and stimulus-evoked activity, cc-RC allows reliable readout training for pattern classification. This yields approximately 300 percent improvements in accuracy and longevity over standard reservoir computing. Knowledge Transplant further allows the learned map from an expert culture to be transplanted to an attractor-equivalent student culture, reducing training time to minutes while improving performance and enabling cross-substrate model reuse.

What carries the argument

Chaos-controlled reservoir computing (cc-RC) with its three components of dynamical signature identification, low-power optical chaos control for stabilization, and readout training; plus Knowledge Transplant (KT) for map transfer between equivalent attractors.

If this is right

  • Robust learning and pattern classification is enabled in living neural cultures.
  • Accuracy and model longevity improve by approximately 300% over standard RC.
  • Training time reduces to minutes with Knowledge Transplant.
  • Learned models become reusable across different neural populations.
  • Knowledge accumulation and sharing is possible beyond biological lifespan limits.

Where Pith is reading between the lines

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

  • This method could support hybrid computing systems combining living tissue with digital readouts for adaptive, low-power tasks.
  • Knowledge transplant might allow maintaining computational performance by switching to fresh cultures as old ones age.
  • Future work could test whether transplanted knowledge preserves performance across more diverse culture types.
  • Applications may extend to real-time biological sensors or bio-inspired AI hardware.

Load-bearing premise

Low-power optical chaos control can reliably stabilize spontaneous and stimulus-evoked activity of living neural cultures into a usable attractor regime without destroying the computational richness.

What would settle it

Demonstration that chaos control either fails to create a stable attractor or reduces the variety of responses so that classification performance does not exceed that of uncontrolled standard reservoir computing.

Figures

Figures reproduced from arXiv: 2604.02552 by Anay Pattanaik, Gaurav Upadhyay, Howard Gritton, John Beggs, Lav Varshney, Leo Maslov, Mattia Gazzola, Seung Hyun Kim, Zhi Dou.

Figure 1
Figure 1. Figure 1: Overview of cc-RC setup. (a) ChR2-transfected mESCs are expanded, formed into embryoid body (EB) in suspension culture, and differentiated into motoneurons. On day 7, differentiated EBs are dissociated and plated on the MEA at a density of 5,000cells per mm2 . Cultures mature by day 14. See SI for details. (b) Living neural networks plated on 128-electrodes MEAs realize integrated neural chips. (c) MiV pla… view at source ↗
Figure 2
Figure 2. Figure 2: Pre-flight characterization of reservoir type and quality. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Performance of (naive) reservoir computing. (a) Ten temporally encoded input patterns. Pattern N contains N light pulses (15ms each) within 900ms, followed by 100ms rest (1s total pattern window). Firing rates are measured at the end of each pattern window. (b) Training protocol. During 1 hour, patterns are presented in random order and multichannel firing rates are recorded. A ridge-classification readout… view at source ↗
Figure 4
Figure 4. Figure 4: Chaos-controlled reservoir computing. (a) Chaos control strategy: a low-amplitude triangular modulation (1 Hz, 100% duty cycle, 10% of input intensity) is superimposed during training and testing. Modulation starts 40 seconds before pattern onset to suppress initial transients. (b) Burst-timing regularization. Raster plots without (top) and with (bottom) modulation; orange dots indicate burst onset. Modula… view at source ↗
Figure 5
Figure 5. Figure 5: Knowledge Transplant. (a) KT workflow. An expert reservoir is fully trained to obtain 𝑊e out. Alignment via transformation T allows to transplant 𝑊e out to a student sample 𝑊s out = T −1𝑊e out (b) Attractors’ alignment. Student’s attractor in latent space (estimated via few input-evoked trajectories) is mapped onto the expert’s attractor (estimated via thousands of trajectories) via ridge-regularized regre… view at source ↗
read the original abstract

We introduce chaos-controlled Reservoir Computing (cc-RC) for living neural cultures: dynamically rich substrates of unique potential for adaptive computation. To account for intrinsic biological variability, cc-RC combines: (i) pre-training identification of each culture's dynamical signature and phase-portrait attractor; (ii) low-power optical chaos control to stabilize spontaneous and stimulus-evoked activity; (iii) readout training within this controlled regime. Across hundreds of neural samples, cc-RC enables robust learning and pattern classification, improving both accuracy and model longevity by approximately 300% over standard RC. We further propose Knowledge Transplant (KT), for which the reservoir map learned by an expert culture is transplanted to an attractor-equivalent student culture, reducing training time to minutes while improving performance. By enabling cross-substrate, reusable learned models, KT paves the way for knowledge accumulation and sharing across neural populations, transcending biological lifespan limits.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 1 minor

Summary. The paper introduces chaos-controlled Reservoir Computing (cc-RC) for living neural cultures. It combines pre-training identification of each culture's dynamical signature and phase-portrait attractor, low-power optical chaos control to stabilize spontaneous and stimulus-evoked activity, and readout training within the controlled regime. The central claims are that cc-RC yields approximately 300% improvement in accuracy and model longevity over standard RC across hundreds of neural samples, and that Knowledge Transplant (KT) allows transfer of a learned reservoir map from an expert culture to an attractor-equivalent student culture, reducing training time to minutes while improving performance.

Significance. If the experimental claims are substantiated with full data and dynamical validation, the work would offer a concrete route to mitigate biological variability in living-neuron substrates while enabling reusable, cross-culture models. This could meaningfully extend the practical scope of reservoir computing with biological hardware beyond single-culture lifetimes.

major comments (3)
  1. [Abstract] Abstract: the claim of 'approximately 300% improvement in both accuracy and model longevity' across 'hundreds of neural samples' is stated without any accompanying performance tables, error bars, statistical tests, sample-selection criteria, or raw data summaries, making the central performance assertion impossible to evaluate.
  2. [Methods] Methods / Results: no quantitative check (Lyapunov spectrum, entropy rate, memory-capacity benchmark, or echo-state property metric) is reported to demonstrate that low-power optical chaos control preserves sufficient dynamical richness relative to the uncontrolled case; without this, the assumption that stabilization does not destroy the separation and echo-state properties required for reservoir computation remains unverified.
  3. [Knowledge Transplant] Knowledge Transplant section: the criteria used to declare two cultures 'attractor-equivalent' and the precise procedure for transplanting the reservoir map (weights, state mapping, or readout transfer) are not specified at a level that would allow independent reproduction or assessment of transfer fidelity.
minor comments (1)
  1. [Abstract] The abstract would be clearer if it named the specific classification tasks and input encoding scheme used to obtain the reported accuracy figures.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive comments, which have helped us identify areas for improvement in the manuscript. We address each major comment point by point below and will make the necessary revisions to enhance clarity and reproducibility.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim of 'approximately 300% improvement in both accuracy and model longevity' across 'hundreds of neural samples' is stated without any accompanying performance tables, error bars, statistical tests, sample-selection criteria, or raw data summaries, making the central performance assertion impossible to evaluate.

    Authors: We agree that the abstract's performance claims would benefit from additional supporting details. In the revised manuscript, we will modify the abstract to include a concise summary of the key results, such as mean accuracy improvements with error bars and references to the statistical tests performed. We will also ensure that the main text includes a dedicated table summarizing performance across samples, along with the sample selection criteria and raw data summaries in the supplementary materials. revision: yes

  2. Referee: [Methods] Methods / Results: no quantitative check (Lyapunov spectrum, entropy rate, memory-capacity benchmark, or echo-state property metric) is reported to demonstrate that low-power optical chaos control preserves sufficient dynamical richness relative to the uncontrolled case; without this, the assumption that stabilization does not destroy the separation and echo-state properties required for reservoir computation remains unverified.

    Authors: This is a valid point. The original manuscript focused on the overall performance gains but did not include explicit dynamical metrics for the controlled regime. We will add a new subsection in the Methods detailing the computation of the Lyapunov spectrum, entropy rate, and memory capacity for both controlled and uncontrolled cases. Corresponding results will be presented in the Results section to confirm that the echo-state property and separation capability are maintained under low-power optical control. revision: yes

  3. Referee: [Knowledge Transplant] Knowledge Transplant section: the criteria used to declare two cultures 'attractor-equivalent' and the precise procedure for transplanting the reservoir map (weights, state mapping, or readout transfer) are not specified at a level that would allow independent reproduction or assessment of transfer fidelity.

    Authors: We will revise the Knowledge Transplant section to provide a detailed description of the attractor-equivalence criteria, including quantitative measures such as attractor similarity via phase space reconstruction and dynamical signature matching using correlation metrics. The transplant procedure will be specified step-by-step: the reservoir weights and internal state mappings from the expert culture are directly transferred, while the readout layer is retrained on the student culture using a small set of samples. We will also include an assessment of transfer fidelity through performance comparisons. revision: yes

Circularity Check

0 steps flagged

No circularity: claims rest on empirical experimental outcomes

full rationale

The paper introduces cc-RC and KT as experimental protocols applied to living neural cultures. All reported improvements (accuracy, longevity, training time) are presented as direct measurements from hundreds of biological samples under controlled conditions. No equations, derivations, parameter fits, or self-referential definitions appear in the provided text; the attractor identification and readout training steps are described procedurally rather than as closed mathematical loops. Any self-citations serve only as background and do not substitute for the experimental validation of the central performance claims. The derivation chain is therefore self-contained against external benchmarks (wet-lab results) and receives the default non-circularity finding.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review yields no explicit free parameters, axioms, or invented physical entities; the approach implicitly assumes standard reservoir computing dynamics can be imposed on biological tissue via optical control.

pith-pipeline@v0.9.0 · 5485 in / 1050 out tokens · 42298 ms · 2026-05-13T20:17:03.886739+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

80 extracted references · 80 canonical work pages

  1. [1]

    E. M. Izhikevich,Dynamical Systems in Neuroscience: The Geometry of Excitability and Bursting (MIT Press, 2007)

    work page 2007

  2. [2]

    Dayan, L

    P. Dayan, L. F. Abbott,Theoretical Neuroscience: Computational and Mathematical Modeling of Neural Systems (MIT Press, 2001)

    work page 2001

  3. [3]

    M.I.Rabinovich,P.Varona,A.I.Selverston,H.D.I.Abarbanel,Dynamical principles in neuroscience.Rev. Mod. Phys. 78, 1213–1265 (4 Nov. 2006)

    work page 2006

  4. [4]

    Science338, 60–65 (2012)

    W.Gerstner,H.Sprekeler,G.Deco,TheoryandSimulationinNeuroscience. Science338, 60–65 (2012)

    work page 2012

  5. [5]

    B. E. STEIN, M. A. MEREDITH, Multisensory Integration.Ann. New York Acad. Sci. 608, 51–70 (1990)

    work page 1990

  6. [6]

    H. Tan, Y. Zhou, Q. Tao, J. Rosen, S. van Dijken, Bioinspired multisensory neural network with crossmodal integration and recognition.Nat. Commun. 12, 1120,issn: 2041-1723 (Feb. 2021)

    work page 2041

  7. [7]

    Bolhasani, S

    E. Bolhasani, S. H. Aboutalebi, Y. Merrikhi, Computational Models of MultisensoryIntegrationwithRecurrentNeuralNetworks:ACriticalReview and Future Directions.Adv. Intell. Syst. 8, 2500147 (2026)

    work page 2026

  8. [8]

    J. M. Beggs, D. Plenz, Neuronal avalanches in neocortical circuits.J. neuroscience23, 11167–11177 (2003)

    work page 2003

  9. [9]

    N.Bertschinger,T.Natschläger,R.Legenstein,presentedattheAdvancesin Neural Information Processing Systems, ed. by L. Saul, Y. Weiss, L. Bottou, vol. 17

    work page

  10. [10]

    Zeraati, V

    R. Zeraati, V. Priesemann, A. Levina, Self-Organization Toward Criticality by Synaptic Plasticity.Front. Phys. Volume 9 - 2021, issn: 2296-424X (2021)

    work page 2021

  11. [11]

    G. J. Pagan-Diazet al., Engineering geometrical 3-dimensional untethered in vitro neural tissue mimic.Proc. National Acad. Sci. 116, 25932–25940 (2019)

    work page 2019

  12. [12]

    Aydin et al., Development of 3D neuromuscular bioactuators.APL bioengineering4 (2020)

    O. Aydin et al., Development of 3D neuromuscular bioactuators.APL bioengineering4 (2020)

    work page 2020

  13. [13]

    Yanet al., 3D bioprinting of human neural tissues with functional connectivity

    Y. Yanet al., 3D bioprinting of human neural tissues with functional connectivity. Cell Stem Cell 31, 260–274 (2024)

    work page 2024

  14. [14]

    J.G.Roth et al.,Advancingmodelsofneuraldevelopmentwithbiomaterials. Nat. Rev. Neurosci. 22, 593–615 (2021)

    work page 2021

  15. [15]

    B. J. Kagan et al., In vitro neurons learn and exhibit sentience when embodied in a simulated game-world.Neuron110, 3952–3969 (2022)

    work page 2022

  16. [16]

    Caiet al., Brain organoid reservoir computing for artificial intelligence

    H. Caiet al., Brain organoid reservoir computing for artificial intelligence. Nat. Electron. 6, 1032–1039 (2023)

    work page 2023

  17. [17]

    4771–4780

    L.Iannello et al.,presentedattheProceedingsoftheIEEE/CVFInternational Conference on Computer Vision, pp. 4771–4780

    work page

  18. [18]

    Sumiet al., Biological neurons act as generalization filters in reservoir computing

    T. Sumiet al., Biological neurons act as generalization filters in reservoir computing. Proc. National Acad. Sci. 120, e2217008120 (2023)

    work page 2023

  19. [19]

    Smirnova et al., Organoid intelligence (OI): the new frontier in bio- computing and intelligence-in-a-dish.Front

    L. Smirnova et al., Organoid intelligence (OI): the new frontier in bio- computing and intelligence-in-a-dish.Front. Sci. Volume 1 - 2023, issn: 2813-6330 (2023)

    work page 2023

  20. [20]

    Upadhyayet al., Living Neurosheets: Engineering readily deployable neural architectures.bioRxiv, 2025–11 (2025)

    G. Upadhyayet al., Living Neurosheets: Engineering readily deployable neural architectures.bioRxiv, 2025–11 (2025)

    work page 2025

  21. [21]

    Marder, J.-M

    E. Marder, J.-M. Goaillard, Variability, compensation and homeostasis in neuron and network function.Nat. Rev. Neurosci. 7, 563–574,issn: 1471-0048 (July 2006)

    work page 2006

  22. [22]

    B. E. Mizusaki, C. O’Donnell, Neural circuit function redundancy in brain disorders. Curr. opinion neurobiology 70, 74–80 (2021)

    work page 2021

  23. [23]

    Marom, G

    S. Marom, G. Shahaf, Development, learning and memory in large random networks of cortical neurons: Lessons beyond anatomy.Q. Rev. Biophys. 35, 63–87,issn: 00335835 (2002)

    work page 2002

  24. [24]

    Yaronet al., Dissociated neuronal cultures as model systems for self- organized prediction.Front

    A. Yaronet al., Dissociated neuronal cultures as model systems for self- organized prediction.Front. Neural Circuits19, 1568652 (2025)

    work page 2025

  25. [25]

    S. O. Sandovalet al., Rigor and reproducibility in human brain organoid research: Where we are and where we need to go.Stem Cell Reports 19, 796–816 (2024)

    work page 2024

  26. [26]

    B. B. Ferreira, A. S. de Paula, M. A. Savi, Chaos control applied to heart rhythm dynamics.Chaos, Solitons Fractals 44, 587–599,issn: 0960-0779 (2011)

    work page 2011

  27. [27]

    Garfinkel, M

    A. Garfinkel, M. L. Spano, W. L. Ditto, J. N. Weiss, Controlling Cardiac Chaos. Science 257, 1230–1235,issn: 0036-8075 (1992)

    work page 1992

  28. [28]

    S. J. Schiff, D. H. Duong, T. Chang, M. L. Spanot, W. L. Ditto, Controlling chaos in the brain.6, 615–620 (1993)

    work page 1993

  29. [29]

    Dimakou, G

    A. Dimakou, G. Pezzulo, A. Zangrossi, M. Corbetta, The predictive nature of spontaneous brain activity across scales and species.Neuron113, 1310– 1332 (2025)

    work page 2025

  30. [30]

    Buzsáki, N

    G. Buzsáki, N. Logothetis, W. Singer, Scaling brain size, keeping timing: evolutionary preservation of brain rhythms.Neuron80, 751–764 (2013)

    work page 2013

  31. [31]

    Grigoryeva, J

    L. Grigoryeva, J. P. Ortega, Echo state networks are universal.Neural Networks108, 495–508,issn: 18792782 (2018)

    work page 2018

  32. [32]

    Gonon, J

    L. Gonon, J. P. Ortega, Reservoir Computing Universality with Stochastic Inputs. IEEE Trans. on Neural Networks Learn. Syst. 31, 100–112,issn: 21622388 (2020)

    work page 2020

  33. [33]

    Maass, Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations.2560, 30,issn: 0899-7667 (2001)

    W. Maass, Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations.2560, 30,issn: 0899-7667 (2001)

    work page 2001

  34. [34]

    Maheswaranathan, A

    N. Maheswaranathan, A. Williams, M. Golub, S. Ganguli, D. Sussillo, presentedattheAdvancesinNeuralInformationProcessingSystems,ed.by H. Wallachet al., vol. 32

    work page

  35. [35]

    Mastrogiuseppe, S

    F. Mastrogiuseppe, S. Ostojic, A Geometrical Analysis of Global Stability in Trained Feedback Networks.Neural Comput. 31, 1139–1182,issn: 0899-7667 (June 2019)

    work page 2019

  36. [36]

    The “echo state

    H. Jaeger, “The “echo state” approach to analysing and training recurrent neural networks – with an Erratum note”, tech. rep. 4, pp. 221–223

    work page

  37. [37]

    L. E. Suárez, B. A. Richards, G. Lajoie, B. Misic, Learning function from structure in neuromorphic networks.Nat. Mach. Intell. 3, 771–786,issn: 25225839 (2021)

    work page 2021

  38. [38]

    Reports 2, issn: 20452322 (2012)

    J.Dambre,D.Verstraeten,B.Schrauwen,S.Massar,Informationprocessing capacity of dynamical systems.Sci. Reports 2, issn: 20452322 (2012)

    work page 2012

  39. [39]

    D. A. Wagenaar, J. Pine, S. M. Potter, Effective parameters for stimulation of dissociated cultures using multi-electrode arrays.J. Neurosci. Methods 138, 27–37,issn: 0165-0270 (2004)

    work page 2004

  40. [40]

    Ronchiet al., Single-Cell Electrical Stimulation Using CMOS-Based High-Density Microelectrode Arrays.Front

    S. Ronchiet al., Single-Cell Electrical Stimulation Using CMOS-Based High-Density Microelectrode Arrays.Front. Neurosci.Volume 13 - 2019, issn: 1662-453X (2019)

    work page 2019

  41. [41]

    C.-Y. Wu, D. Whye, R. W. Mason, W. Wang, Efficient differentiation of mouse embryonic stem cells into motor neurons.J. visualized experiments: JoVE, 3813 (2012)

    work page 2012

  42. [42]

    Robotics 10, eadu5830 (2025)

    H.Min et al.,Optogeneticneuromuscularactuationofaminiatureelectronic biohybrid robot.Sci. Robotics 10, eadu5830 (2025). 8/9

    work page 2025

  43. [43]

    Zhanget al., Mind in vitro platforms: Versatile, scalable, robust, and open solutions to interfacing with living neurons.Adv

    X. Zhanget al., Mind in vitro platforms: Versatile, scalable, robust, and open solutions to interfacing with living neurons.Adv. Sci. 11, 2306826 (2024)

    work page 2024

  44. [44]

    J. P. Cunningham, B. M. Yu, K. V. Shenoy, M. Sahani, Inferring neural firing rates from spike trains using Gaussian Processes.Adv. Neural Inf. Process. Syst. 20 - Proc. 2007 Conf., 1–8 (2008)

    work page 2007

  45. [45]

    Sussillo, M

    D. Sussillo, M. M. Churchland, M. T. Kaufman, K. V. Shenoy, A neural networkthatfindsanaturalisticsolutionfortheproductionofmuscleactivity. Nat. Neurosci.18, 1025–1033,issn: 15461726 (2015)

    work page 2015

  46. [46]

    Gosztolai, R

    A. Gosztolai, R. L. Peach, A. Arnaudon, M. Barahona, P. Vandergheynst, MARBLE: interpretable representations of neural population dynamics using geometric deep learning.Nat. Methods22, 612–620,issn: 1548-7105 (Mar. 2025)

    work page 2025

  47. [47]

    DePasquale, D

    B. DePasquale, D. Sussillo, L. Abbott, M. M. Churchland, The centrality of population-level factors to network computation is demonstrated by a versatile approach for training spiking networks.Neuron111, 631–649.e10, issn: 0896-6273 (2023)

    work page 2023

  48. [48]

    G.DiAntonio,T.Gili,A.Gabrielli,M.Mattia,LinearizingandForecasting: A Reservoir Computing Route to Digital Twins of the Brain.Adv. Sci. (Mar. 2026)

    work page 2026

  49. [49]

    Kenet, D

    T. Kenet, D. Bibitchkov, M. Tsodyks, A. Grinvald, A. Arieli, Spontaneously emerging cortical representations of visual attributes.Nature425, 954–956, issn: 1476-4687 (Oct. 2003)

    work page 2003

  50. [50]

    Heineyet al., Neuronal avalanche dynamics and functional connectivity elucidateinformationpropagationinvitro

    K. Heineyet al., Neuronal avalanche dynamics and functional connectivity elucidateinformationpropagationinvitro. Front. Neural Circuits16(2022)

    work page 2022

  51. [51]

    B. Liu, D. V. Buonomano, Ex vivo cortical circuits learn to predict and spontaneously replay temporal patterns.Nat. Commun. 16, 3179, issn: 2041-1723 (Apr. 2025)

    work page 2041

  52. [52]

    Iannelloet al., Criticality in neural cultures: Insights into memory and connectivity in entorhinal-hippocampal networks.Chaos, Solitons Fractals 194, 116184,issn: 0960-0779 (2025)

    L. Iannelloet al., Criticality in neural cultures: Insights into memory and connectivity in entorhinal-hippocampal networks.Chaos, Solitons Fractals 194, 116184,issn: 0960-0779 (2025)

    work page 2025

  53. [53]

    Legenstein, W

    R. Legenstein, W. Maass, Edge of chaos and prediction of computational performance for neural circuit models.Neural Networks20, 323–334,issn: 08936080 (2007)

    work page 2007

  54. [54]

    D.A.Wagenaar,J.Pine,S.M.Potter,Anextremelyrichrepertoireofbursting patterns during the development of cortical cultures.BMC neuroscience 7, 11 (2006)

    work page 2006

  55. [55]

    J. M. Beggs, N. Timme, Being critical of criticality in the brain.Front. Physiol.3 JUN, 1–14,issn: 1664042X (2012)

    work page 2012

  56. [56]

    R.Calvo,C.Martorell,G.B.Morales,S.DiSanto,M.A.Muñoz,Frequency- DependentCovarianceRevealsCriticalSpatiotemporalPatternsofSynchro- nized Activity in the Human Brain.Phys. Rev. Lett.133, 208401 (20 Nov. 2024)

    work page 2024

  57. [57]

    Calvo, C

    R. Calvo, C. Martorell, A. Roig, M. A. Muñoz, Robust Scaling in Hu- man Brain Dynamics Despite Correlated Inputs and Limited Sampling Distortions. Phys. Rev. Lett.136, 068402 (6 Feb. 2026)

    work page 2026

  58. [58]

    W. L. Shew, H. Yang, T. Petermann, R. Roy, D. Plenz, Neuronal avalanches imply maximum dynamic range in cortical networks at criticality.J. neuro- science29, 15595–15600 (2009)

    work page 2009

  59. [59]

    D. R. Chialvo, Emergent complex neural dynamics.Nat. Phys. 6, 744–750 (2010)

    work page 2010

  60. [60]

    M. Dale, J. F. Miller, S. Stepney, M. A. Trefzer, A substrate-independent framework to characterize reservoir computers.Proc. Royal Soc. A: Math. Phys. Eng. Sci. 475, 20180723,issn: 1364-5021 (June 2019)

    work page 2019

  61. [61]

    M. Dale, J. Dewhirst, S. O’Keefe, S. Stepney, The role of structure and complexity on Reservoir Computing quality.Matthew Dale Jack Dewhirst Simon O’Keefe Angelika Sebald Susan Stepney Martin A. Trefzer (2019)

    work page 2019

  62. [62]

    Jaeger, Short term memory in echo state networks.GMD Rep

    H. Jaeger, Short term memory in echo state networks.GMD Rep. 152, 60 (2002)

    work page 2002

  63. [63]

    Schubert, C

    F. Schubert, C. Gros, Local Homeostatic Regulation of the Spectral Radius ofEcho-StateNetworks. Front. Comput. Neurosci.15,1–17,issn:16625188 (2021)

    work page 2021

  64. [64]

    Jaeger, M

    H. Jaeger, M. Lukoševičius, D. Popovici, U. Siewert, Optimization and applications of echo state networks with leaky- integrator neurons.Neural Networks20, 335–352,issn: 08936080 (2007)

    work page 2007

  65. [65]

    Menesse, A

    G. Menesse, A. M. Houben, J. Soriano, J. J. Torres, Integrated information decomposition unveils major structural traits of in silico and in vitro neuronal networks.Chaos: An Interdiscip. J. Nonlinear Sci. 34, 053139, issn: 1054-1500 (May 2024)

    work page 2024

  66. [66]

    Tauste Campo, Inferring neural information flow from spiking data

    A. Tauste Campo, Inferring neural information flow from spiking data. Comput. Struct. Biotechnol. J. 18, 2699–2708,issn: 20010370 (2020)

    work page 2020

  67. [67]

    M. Dale, S. O’Keefe, A. Sebald, S. Stepney, M. A. Trefzer, Reservoir computing quality: connectivity and topology.Nat. Comput. 20, 205–216, issn: 15729796 (2021)

    work page 2021

  68. [68]

    Y.Kawai,J.Park,M.Asada,Asmall-worldtopologyenhancestheechostate property and signal propagation in reservoir computing.Neural Networks 112, 15–23,issn: 18792782 (2019)

    work page 2019

  69. [69]

    Neurophysiol.102, 614–635, issn: 15221598 (2009)

    B.M.Yu et al.,Gaussian-processfactoranalysisforlow-dimensionalsingle- trial analysis of neural population activity.J. Neurophysiol.102, 614–635, issn: 15221598 (2009)

    work page 2009

  70. [70]

    Dowling, Y

    M. Dowling, Y. Zhao, I. M. Park, Real-time variational method for learning neural trajectory and its dynamics. (2023)

    work page 2023

  71. [71]

    D.Sussillo,O.Barak,OpeningtheBlackBox:Low-DimensionalDynamics in High-Dimensional Recurrent Neural Networks.Neural Comput.25, 626– 649, issn: 0899-7667 (Mar. 2013)

    work page 2013

  72. [72]

    G. E. Hinton, R. R. Salakhutdinov, Reducing the Dimensionality of Data with Neural Networks.Science 313, 504–507 (2006)

    work page 2006

  73. [73]

    D.Salaj et al.,Spikefrequencyadaptationsupportsnetworkcomputationson temporally dispersed information.eLife10, 1–33,issn: 2050084X (2021)

    work page 2021

  74. [74]

    Zeraati, A

    R. Zeraati, A. Levina, J. H. Macke, R. Gao, Neural timescales from a computational perspective. 1–37 (2025)

    work page 2025

  75. [75]

    Caroni, F

    P. Caroni, F. Donato, D. Muller, Structural plasticity upon learning: regula- tion and functions.Nat. Rev. Neurosci. 13, 478–490 (2012)

    work page 2012

  76. [76]

    Lakatos, J

    P. Lakatos, J. Gross, G. Thut, A new unifying account of the roles of neuronal entrainment.Curr. Biol. 29, R890–R905 (2019)

    work page 2019

  77. [77]

    Ziaeemehr, M

    A. Ziaeemehr, M. Zarei, A. Sheshbolouki, Emergence of global synchro- nization in directed excitatory networks of type I neurons.Sci. Reports 10, 1–11,issn: 20452322 (2020)

    work page 2020

  78. [78]

    P. S. Skardal, A. Arenas, Control of coupled oscillator networks with application to microgrid technologies.Sci. Adv. 1, 1–6,issn: 23752548 (2015)

    work page 2015

  79. [79]

    L. N. Driscoll, K. Shenoy, D. Sussillo, Flexible multitask computation in recurrent networks utilizes shared dynamical motifs.Nat. Neurosci. 27, 1349–1363,issn: 1546-1726 (July 2024)

    work page 2024

  80. [80]

    M. Huh, B. Cheung, T. Wang, P. Isola, presented at the Proceedings of the 41stInternationalConferenceonMachineLearning,ed.byR.Salakhutdinov et al., vol. 235, pp. 20617–20642. 9/9

    work page