Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations
Open accessThe DOI matches its Crossref record; 28 of the 28 references it cites were found in a registry or an index.
Abstract
A key challenge for neural modeling is to explain how a continuous stream of multimodal input from a rapidly changing environment can be processed by stereotypical recurrent circuits of integrate-and-fire neurons in real time. We propose a new computational model for real-time computing on time-varying input that provides an alternative to paradigms based on Turing machines or attractor neural networks. It does not require a task-dependent construction of neural circuits. Instead, it is based on principles of high-dimensional dynamical systems in combination with statistical learning theory and can be implemented on generic evolved or found recurrent circuitry. It is shown that the inherent transient dynamics of the high-dimensional dynamical system formed by a sufficiently large and heterogeneous neural circuit may serve as universal analog fading memory. Readout neurons can learn to extract in real time from the current state of such recurrent neural circuit information about current and past inputs that may be needed for diverse tasks. Stable internal states are not required for giving a stable output, since transient internal states can be transformed by readout neurons into stable target outputs due to the high dimensionality of the dynamical system. Our approach is based on a rigorous computational model, the liquid state machine, that, unlike Turing machines, does not require sequential transitions between well-defined discrete internal states. It is supported, as the Turing machine is, by rigorous mathematical results that predict universal computational power under idealized conditions, but for the biologically more realistic scenario of real-time processing of time-varying inputs. Our approach provides new perspectives for the interpretation of neural coding, the design of experiments and data analysis in neurophysiology, and the solution of problems in robotics and neurotechnology.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Although work in artificial intelligence and machine-learning research has shown that similar or even more complex outcomes can be achieved with a variety of techniques, including deep neural networks with trained weights (Gibney, 2015), reservoir networks with a trained output mapping (Maass et al., 2002;Jaeger & Haas, 2004), networks that are evolved using a genetic algorithm (Iizuka & Di Paolo,”
From A potential mechanism for Gibsonian resonance: behavioral entrainment emerges from local homeostasis in an unsupervised reservoir network (2024) · cites this paper as Maass et al., 2002
In the citing paper
“Finally, we show that a randomly connected circuit model, akin to a “reservoir” network13-15, can reproduce the observed distribution of timescales, provided that the network operates at the critical point (or “edge of chaos”)16-18.”
From A reservoir of time constants for memory traces in cortical neurons (2011) · cites this paper as [13]
In the citing paper
“Reservoir-type networks have been the subject of active research in computational neuroscience and machine learning13-15, but so far experimental support that such networks are adopted by the brain has been lacking.”
From A reservoir of time constants for memory traces in cortical neurons (2011) · cites this paper as [13]
In the citing paper
“Those models predict that the memory of input signals is stored in a large, recurrent and heterogeneous network (reservoir), in a distributed manner, and that a desired output is obtained by a trainable combination of the response signals in the reservoir.”
From A reservoir of time constants for memory traces in cortical neurons (2011) · cites this paper as [13]
Citations
Cites
Wu, Y. (1999) · Technometrics · doi:10.1080/00401706.1999.10485951 · Cited by 27,221
Introduction to the theory of neural computation
(1994) · Neural Networks · doi:10.1016/0893-6080(94)90014-0 · Cited by 6,401
The Synaptic Organization of the Brain
(2004) · Oxford University Press eBooks · doi:10.1093/acprof:oso/9780195159561.001.1 · Cited by 3,546 · cited 2×
“This informationisprocessed by extremely complexbut surprisingly stereotypic microcircuits that can perform a wide spectrum of tasks (Shep-herd, 1988 ; Douglas & Martin, 1998 ; von Melchner, Pallas, & Sur, 2000).”
Dynamics of ongoing activity: explanation of the large variability in evoked cortical responses
Arieli, A., Sterkin, A., Grinvald, A., & Aertsen, A. (1996) · Science · doi:10.1126/science.273.5283.1868 · Cited by 1,660 · cited 1×
“Furthermore, although the input rate may contain salient information that is constant for a particular readout element, it may not be for another (see, for example, Figure 8), indicating that equivalence classes and dynamic stability exist purely from the perspective of the readout elements.”
BackgroundDifferential signaling via the same axon of neocortical pyramidal neurons
Markram, H., Wang, Y., & Tsodyks, M. (1998) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.95.9.5323 · Cited by 1,102 · cited 2×
“In the case of a synaptic connection from a to b , we modeled the synaptic dynamics according to the model proposed in Markram, Wang, and Tsodyks (1998), with the synaptic parameters U (use), D (time constant for depression), and F (time constant for facilitation) randomly chosen from gaussian…”
BackgroundMethodSelf-citationOrganizing principles for a diversity of GABAergic interneurons and synapses in the neocortex
Gupta, A., Wang, Y., & Markram, H. (2000) · Science · doi:10.1126/science.287.5451.273 · Cited by 1,036
Self-citationFading memory and the problem of approximating nonlinear operators with Volterra series
Boyd, S., & Chua, L. (1985) · IEEE Transactions on Circuits and Systems · doi:10.1109/tcs.1985.1085649 · Cited by 991 · cited 2×
“> The proof of this theorem follows from the Stone-Weierstrass approximation theorem, like the proof of theorem 1 in Boyd and Chua (1985).”
BackgroundGradient calculations for dynamic recurrent neural networks: a survey
Pearlmutter, B. A. (1995) · IEEE Transactions on Neural Networks · doi:10.1109/72.410363 · Cited by 582 · cited 1×
“Methods for controllingthe dynamics of recurrent neural networks through adaptive mechanisms are reviewed in Pearlmutter (1995).”
Temporal Information Transformed into a Spatial Code by a Neural Network with Realistic Properties
Buonomano, D. V., & Merzenich, M. M. (1995) · Science · doi:10.1126/science.7863330 · Cited by 449 · cited 1×
“Special cases of this mechanism have already been reported in Buonomano and Merzenich (1995) and Dominey, Arbib, and Joseph (1995).”
BackgroundLearning and Extracting Finite State Automata with Second-Order Recurrent Neural Networks
Giles, C. L., Miller, C. B., Chen, D., Chen, H. H., Sun, G. Z., & Lee, Y. C. (1992) · Neural Computation · doi:10.1162/neco.1992.4.3.393 · Cited by 392
Cited by
Items 201–300 of 1,014
An overview and comparative analysis of Recurrent Neural Networks for Short Term Load Forecasting
Bianchi, F. M., Maiorino, E., Kampffmeyer, M., Rizzi, A., & Jenssen, R. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1705.04378 · Cited by 182
“ESNs, along with Liquid State Machines [137], belong to the class of computational dynamical systems implemented according to the so-called reservoir computing framework [71].”
BackgroundThe Introduction of Time-Scales in Reservoir Computing, Applied to Isolated Digits Recognition
Schrauwen, B., Defour, J., Verstraeten, D., & Van Campenhout, J. (2007) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-74690-4_48 · Cited by 181
“Originally introduced independently as the Liquid State Machine [4] or Echo State Networks [5], the basic concept is to randomly construct an RNN and to leave the This research is partially funded by FWO Flanders project G.”
BackgroundA soft body as a reservoir: case studies in a dynamic model of octopus-inspired soft robotic arm
Nakajima, K., Hauser, H., Kang, R., Guglielmino, E., Caldwell, D. G., & Pfeifer, R. (2013) · Frontiers in Computational Neuroscience · doi:10.3389/fncom.2013.00091 · Cited by 180
“As explained in section 1, our approach is comparable to a reservoir computing approach, which normally uses randomly coupled non-linear elements as a computational resource (Jaeger, 2002; Maass et al., 2002; Jaeger and Haas, 2004).”
BackgroundMethodInfluentialSpike processing with a graphene excitable laser
Shastri, B. J., Nahmias, M. A., Tait, A. N., Rodriguez, A. W., Wu, B., & Prucnal, P. R. (2016) · Scientific Reports · doi:10.1038/srep19126 · Cited by 180
Bollt, E. (2021) · Chaos: An Interdisciplinary Journal of Nonlinear Science · doi:10.1063/5.0024890 · Cited by 179
“The reservoir computing (RC) [44, 58, 87] and the closely related echo state network (ESN) [43, 57] and liquid state machine (LSM) [60, 35] have emerged as a special variant of an RNN, where only the output layer is trained rather than the entire network of weights.”
BackgroundJournal of Advances in Modeling Earth Systems
(2018) · Journal of Advances in Modeling Earth Systems · doi:10.1002/(issn)1942-2466 · Cited by 179
A learning rule for very simple universal approximators consisting of a single layer of perceptrons
Auer, P., Burgsteiner, H., & Maass, W. (2008) · Neural Networks · doi:10.1016/j.neunet.2007.12.036 · Cited by 179
Self-citationCompact reservoir computing with a photonic integrated circuit
Takano, K., Sugano, C., Inubushi, M., Yoshimura, K., Sunada, S., Kanno, K., & Uchida, A. (2018) · Optics Express · doi:10.1364/oe.26.029424 · Cited by 179
Capacity and Trainability in Recurrent Neural Networks
Collins, J., Sohl‐Dickstein, J., & Sussillo, D. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.09913 · Cited by 74
“Previous work on short term memory traces explores the tradeoffs between memory fidelity and duration, for the case that a new input is presented to the RNN at every time step (Jaeger & Haas, 2004; Maass et al., 2002; White et al., 2004; Ganguli et al., 2008; Charles et al., 2014).”
BackgroundRich-Club Organization in Effective Connectivity among Cortical Neurons
Nigam, S., Shimono, M., Ito, S., Yeh, F.-C., Timme, N., Myroshnychenko, M., Lapish, C. C., Tosi, Z., Hottowy, P., Smith, W. C., et al. (2016) · The Journal of Neuroscience · doi:10.1523/jneurosci.2177-15.2016 · Cited by 179
“All neurons were placed at random locations within a three-dimensional cube, with the probability of their being a connection between two neurons given as follows (Maass et al., 2002):”
BackgroundAvalanches and edge-of-chaos learning in neuromorphic nanowire networks
Hochstetter, J., Zhu, R., Loeffler, A., Diaz-Alvarez, A., Nakayama, T., & Kuncic, Z. (2021) · Nature Communications · doi:10.1038/s41467-021-24260-z · Cited by 178
Long-term prediction of chaotic systems with machine learning
Fan, H., Jiang, J., Zhang, C., Wang, X., & Lai, Y.-C. (2020) · Physical Review Research · doi:10.1103/physrevresearch.2.012080 · Cited by 178
Zhang, Y., Li, P., Jin, Y., & Choe, Y. (2015) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2015.2388544 · Cited by 178
“The internal LSM model parameters may be trained by extracting statistical information from the data without using acoustic and language models that are specific to targeted languages and datasets.”
MethodConnectivity, Dynamics, and Memory in Reservoir Computing with Binary and Analog Neurons
Büsing, L., Schrauwen, B., & Legenstein, R. (2010) · Neural Computation · doi:10.1162/neco.2009.01-09-947 · Cited by 177
“There, the computational performance of networks of randomly connected threshold gates was linked to their separation property (for a formal definition, see Maass et al., 2002).”
BackgroundMethodHuman action recognition with a large-scale brain-inspired photonic computer
Antonik, P., Marsal, N., Brunner, D., & Rontani, D. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-019-0110-8 · Cited by 177
Zha, J., Shi, S., Chaturvedi, A., Huang, H., Yang, P., Yao, Y., Li, S., Xia, Y., Zhang, Z., Wang, W., et al. (2023) · Advanced Materials · doi:10.1002/adma.202211598 · Cited by 175
In-memory computing with emerging memory devices: Status and outlook
Mannocci, P., Farronato, M., Lepri, N., Cattaneo, L., Glukhov, A., Sun, Z., & Ielmini, D. (2023) · APL Machine Learning · doi:10.1063/5.0136403 · Cited by 191
Delay-Based Reservoir Computing: Noise Effects in a Combined Analog and Digital Implementation
Soriano, M. C., Ortin, S., Keuninckx, L., Appeltant, L., Danckaert, J., Pesquera, L., & van der Sande, G. (2015) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2014.2311855 · Cited by 174
“The main inspiration underlying RC is the insight that neural systems can process information by generating patterns of transient activity, which are excited by input sensory signals [3].”
BackgroundBiological Foundations and Origin of Syntax
(2009) · The MIT Press eBooks · doi:10.7551/mitpress/9780262013567.001.0001 · Cited by 174
Single-Neuron Stability during Repeated Reaching in Macaque Premotor Cortex
Chestek, C. A., Batista, A. P., Santhanam, G., Yu, B. M., Afshar, A., Cunningham, J. P., Gilja, V., Ryu, S. I., Churchland, M. M., & Shenoy, K. V. (2007) · The Journal of Neuroscience · doi:10.1523/jneurosci.0959-07.2007 · Cited by 174
“Also, computational models have been developed around the concept of neural circuits using single neurons in different ways over time (Maass et al., 2002) or changing tuning readily (Abbott and Blum, 1996; Rokni et al., 2007).”
MethodThe promise of spintronics for unconventional computing
Finocchio, G., Di Ventra, M., Camsari, K. Y., Everschor-Sitte, K., Khalili Amiri, P., & Zeng, Z. (2021) · Journal of Magnetism and Magnetic Materials · doi:10.1016/j.jmmm.2020.167506 · Cited by 173
“Consequently, the optimal performance of an RC system depends strongly on the properties of the reservoir that can be implemented with any non-linear, complex system with short term memory 69 .”
BackgroundShort-Term Memory in Orthogonal Neural Networks
White, O. L., Lee, D. D., & Sompolinsky, H. (2004) · Physical Review Letters · doi:10.1103/physrevlett.92.148102 · Cited by 172
On-line learning and modulation of periodic movements with nonlinear dynamical systems
Gams, A., Ijspeert, A. J., Schaal, S., & Lenarčič, J. (2009) · Autonomous Robots · doi:10.1007/s10514-009-9118-y · Cited by 172
“When the chosen representation is linear in the parameters, regression can be sufficient (Maass et al. 2002; Ijspeert et al. 2002b), and has the advantage of fast, one-shot learning.”
BackgroundInfluentialGhosh, S., Opala, A., Matuszewski, M., Paterek, T., & Liew, T. C. H. (2019) · npj Quantum Information · doi:10.1038/s41534-019-0149-8 · Cited by 171
From fixed points to chaos: Three models of delayed discrimination
Barak, O., Sussillo, D., Romo, R., Tsodyks, M., & Abbott, L. F. (2013) · Progress in Neurobiology · doi:10.1016/j.pneurobio.2013.02.002 · Cited by 171
“This is a novel application of an echo-state type of network (Jaeger, 2001; Maass et al., 2002) that operates in the chaotic rather than in the transient decaying regime typically used for such networks (Sussillo and Abbott, 2009).”
Physical reservoir computing based on spin torque oscillator with forced synchronization
Tsunegi, S., Taniguchi, T., Nakajima, K., Miwa, S., Yakushiji, K., Fukushima, A., Yuasa, S., & Kubota, H. (2019) · Applied Physics Letters · doi:10.1063/1.5081797 · Cited by 171
Machine learning prediction of critical transition and system collapse
Kong, L.-W., Fan, H., Grebogi, C., & Lai, Y. (2021) · Physical Review Research · doi:10.1103/physrevresearch.3.013090 · Cited by 171
Reconstructing computational system dynamics from neural data with recurrent neural networks
Durstewitz, D., Koppe, G., & Thurm, M. I. (2023) · Nature reviews. Neuroscience · doi:10.1038/s41583-023-00740-7 · Cited by 230
“Reservoir computers and echo state machines 75,174–176 are another ingenious RNN design that is popular for DS reconstruction 132,177,178 and were originally introduced as a computationally highly efficient alternative to classical RNNs (Fig.”
BackgroundChaisangmongkon, W., Swaminathan, S. K., Freedman, D. J., & Wang, X.-J. (2017) · Neuron · doi:10.1016/j.neuron.2017.03.002 · Cited by 166
Steil, J. J. (2007) · Neural Networks · doi:10.1016/j.neunet.2007.04.011 · Cited by 165
Potential implementation of reservoir computing models based on magnetic skyrmions
Bourianoff, G., Pinna, D., Sitte, M., & Everschor-Sitte, K. (2018) · AIP Advances · doi:10.1063/1.5006918 · Cited by 157
A Machine Learning‐Based Global Atmospheric Forecast Model
Arcomano, T., Szunyogh, I., Pathak, J., Wikner, A., Hunt, B. R., & Ott, E. (2020) · Geophysical Research Letters · doi:10.1029/2020gl087776 · Cited by 154
“Compared to all of the aforementioned studies, an important new aspect of our work is that we employ reservoir computing (RC) (Jaeger, 2001; Lukoševičius & Jaeger, 2009; Lukoševičius, 2012; Maass et al., 2002) rather than deep learning (e.”
BackgroundMethodTask-adaptive physical reservoir computing
Lee, O., Wei, T., Stenning, K. D., Gartside, J. C., Prestwood, D., Seki, S., Aqeel, A., Karube, K., Kanazawa, N., Taguchi, Y., et al. (2023) · Nature Materials · doi:10.1038/s41563-023-01698-8 · Cited by 188
Echo State Property of Deep Reservoir Computing Networks
Gallicchio, C., & Micheli, A. (2017) · Cognitive Computation · doi:10.1007/s12559-017-9461-9 · Cited by 164
“Although the principles of RC have been instantiated in different ways in literature (see e.g., [28, 40, 42]), the reference RC model in the neuro-computing area is represented by the Echo State Network (ESN) [18, 21].”
MethodImpact of input mask signals on delay-based photonic reservoir computing with semiconductor lasers
Kuriki, Y., Nakayama, J., Takano, K., & Uchida, A. (2018) · Optics Express · doi:10.1364/oe.26.005777 · Cited by 164
Cascadable excitability in microrings
Vaerenbergh, T. V., Fiers, M., Mechet, P., Spuesens, T., Kumar, R., Morthier, G. I., Schrauwen, B., Dambre, J., & Bienstman, P. (2012) · Optics Express · doi:10.1364/oe.20.020292 · Cited by 164
Denoising Source SeparationPreprint
SäreläJaakko, & ValpolaHarri (2005) · CogPrints (University of Southampton) · doi:10.5555/1046920.1058110 · Cited by 164
Theoretical Neuroscience Rising
Abbott, L. F. (2008) · Neuron · doi:10.1016/j.neuron.2008.10.019 · Cited by 163
“of network models known as liquid state or echostate networks (Maass et al., 2002; Jaeger, 2003).”
BackgroundOpportunities in Quantum Reservoir Computing and Extreme Learning Machines
Mujal, P., Martínez‐Peña, R., Nokkala, J., García‐Beni, J., Giorgi, G. L., Soriano, M. C., & Zambrini, R. (2021) · Advanced Quantum Technologies · doi:10.1002/qute.202100027 · Cited by 163
Hill, S. L., Wang, Y., Riachi, I., Schürmann, F., & Markram, H. (2012) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1202128109 · Cited by 163
“Synaptic plasticity may act on top of this foundation to selectively strengthen or weaken multiple recurrent synaptic loops, thereby enabling a diverse range of transient internal states to emerge and a broad variety of cortical computations to be performed within the cortical circuit (63).”
BackgroundSelf-citationStable Ensemble Performance with Single-Neuron Variability during Reaching Movements in Primates
Carmena, J. M., Lebedev, M. A., Henriquez, C. S., & Nicolelis, M. A. L. (2005) · The Journal of Neuroscience · doi:10.1523/jneurosci.2772-05.2005 · Cited by 163
“This variability in neuronal firing and its correlation with behavioral parameters has recently been explained by a model of neural computation based on perturbations (Mass et al., 2002).”
Temporal Spike Sequence Learning via Backpropagation for Deep Spiking Neural NetworksPreprint
Zhang, W., & Li, P. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.10085 · Cited by 278
“Despite the recent progress in SNNs and neuromorphic processor designs, fully leveraging the theoretical computing advantages of SNNs over traditional artificial neural networks (ANNs) (Maass et al., 2002) to achieve a greater computational performance remains challenging.”
BackgroundLearning grammatical structure with Echo State Networks
Tong, M. H., Bickett, A. D., Christiansen, E. M., & Cottrell, G. W. (2007) · Neural Networks · doi:10.1016/j.neunet.2007.04.013 · Cited by 162
Tutorial: Photonic neural networks in delay systems
Brunner, D., Penkovsky, B., Marquez, B. A., Jacquot, M., Fischer, I., & Larger, L. (2018) · Journal of Applied Physics · doi:10.1063/1.5042342 · Cited by 155
Observing spatio-temporal dynamics of excitable media using reservoir computing
Zimmermann, R. S., & Parlitz, U. (2018) · Chaos: An Interdisciplinary Journal of Nonlinear Science · doi:10.1063/1.5022276 · Cited by 161
Memristor-based reservoir computing
Kulkarni, M. S., & Teuscher, C. (2012) · Proceedings of the 2012 IEEE/ACM International Symposium on Nanoscale Architectures · doi:10.1145/2765491.2765531 · Cited by 156
“We apply a methodology that allows to obtain desired functionality from the intrinsic dynamics of such device networks.”
MethodTanaka, H., Akai-Kasaya, M., TermehYousefi, A., Hong, L., Fu, L., Tamukoh, H., Tanaka, D., Asai, T., & Ogawa, T. (2018) · Nature Communications · doi:10.1038/s41467-018-04886-2 · Cited by 154
Low-dimensional dynamics for working memory and time encoding
Cueva, C. J., Saez, A., Marcos, E., Genovesio, A., Jazayeri, M., Romo, R., Salzman, C. D., Shadlen, M. N., & Fusi, S. (2020) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1915984117 · Cited by 182
Deep learning incorporating biologically inspired neural dynamics and in-memory computing
Woźniak, S., Pantazi, A., Bohnstingl, T., & Eleftheriou, E. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-020-0187-0 · Cited by 160
Mapping Spiking Neural Networks to Neuromorphic Hardware
Balaji, A., Catthoor, F., Das, A., Wu, Y., Huynh, K., Dell'Anna, F. G., Indiveri, G., Krichmar, J. L., Dutt, N. D., & Schaafsma, S. (2020) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2019.2951493 · Cited by 159
“Examples include the liquid state machine [27], zero-shot learning [61], one-shot learning [62], lifelong learning [63], transfer learning [64], and deep reinforcement learning [65] among others.”
BackgroundEvolving spiking neural network—a survey
Schliebs, S., & Kasabov, N. (2013) · Evolving Systems · doi:10.1007/s12530-013-9074-9 · Cited by 159
Perceptual training continuously refines neuronal population codes in primary visual cortex
Yan, Y., Rasch, M. J., Chen, M., Xiang, X., Huang, M., Wu, S., & Li, W. (2014) · Nature Neuroscience · doi:10.1038/nn.3805 · Cited by 175
Kawai, Y., Park, J., & Asada, M. (2019) · Neural Networks · doi:10.1016/j.neunet.2019.01.002 · Cited by 158
On the origins of randomization-based feedforward neural networks
Suganthan, P. N., & Katuwal, R. (2021) · Applied Soft Computing · doi:10.1016/j.asoc.2021.107239 · Cited by 158
Berner, R., Groß, T., Kuehn, C., Kurths, J., & Yanchuk, S. (2023) · Physics Reports · doi:10.1016/j.physrep.2023.08.001 · Cited by 202
Experimental photonic quantum memristor
Spagnolo, M., Morris, J., Piacentini, S., Antesberger, M., Massa, F., Crespi, A., Ceccarelli, F., Osellame, R., & Walther, P. J. (2022) · Nature Photonics · doi:10.1038/s41566-022-00973-5 · Cited by 177
“[44] Wolfgang Maass, Thomas Natschläger, and Henry Markram, “Real-time computing without stable states: A new framework for neural computation based on perturbations,” Neural computation 14, 2531–2560 (2002).”
BackgroundBackpropagation through time and the brain
Lillicrap, T. P., & Santoro, A. (2019) · Current Opinion in Neurobiology · doi:10.1016/j.conb.2019.01.011 · Cited by 142
Consciousness is supported by near-critical slow cortical electrodynamics
Toker, D., Pappas, I., Lendner, J. D., Frohlich, J., Mateos, D. M., Muthukumaraswamy, S., Carhart-Harris, R., Paff, M., Vespa, P. M., Monti, M. M., et al. (2022) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2024455119 · Cited by 138
“A salient feature of chaotic RNNs is their ability to transiently stabilize their dynamics in response to external or feedback signals (65–67, 69, 83), which appears to be important for their trainability (65–67, 69).”
BackgroundInfluentialUniversal Memcomputing Machines
Traversa, F. L., & Di Ventra, M. (2015) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2015.2391182 · Cited by 153
“Very recently, a new computing paradigm, inspired by the operation of our own brain, has been proposed which is not based on the UTM concept, and which puts the whole burden of computation directly into the memory.”
MethodBoosting Computational Power through Spatial Multiplexing in Quantum Reservoir Computing
Nakajima, K., Fujii, K., Negoro, M., Mitarai, K., & Kitagawa, M. (2019) · Physical Review Applied · doi:10.1103/physrevapplied.11.034021 · Cited by 153
“, liquid state machine [3]) in the software program running on a PC.”
MethodHinaut, X., & Dominey, P. F. (2013) · PLoS ONE · doi:10.1371/journal.pone.0052946 · Cited by 140
“Reservoir computing is a machine learning technique in which a recurrent network with fixed connections is used to encode the spatiotemporal structure of an input sequence, and connections to a readout layer are trained using efficient algorithms to produce a desired output in response to input sequences [35,36].”
MethodStructure in neural population recordings: an expected byproduct of simpler phenomena?
Elsayed, G. F., & Cunningham, J. P. (2017) · Nature Neuroscience · doi:10.1038/nn.4617 · Cited by 153
The Eighty Five Percent Rule for optimal learning
Wilson, R. C., Shenhav, A., Straccia, M., & Cohen, J. D. (2019) · Nature Communications · doi:10.1038/s41467-019-12552-4 · Cited by 119
Physical reservoir computing using magnetic skyrmion memristor and spin torque nano-oscillator
Jiang, W., Chen, L., Zhou, K., Li, L., Fu, Q., Du, Y., & Liu, R. (2019) · Applied Physics Letters · doi:10.1063/1.5115183 · Cited by 159
In search of the neural circuits of intrinsic motivation
Kaplan, F. (2007) · Frontiers in Neuroscience · doi:10.3389/neuro.01.1.1.017.2007 · Cited by 140
Phoneme Recognition with Large Hierarchical Reservoirs
Triefenbach, F., Jalalvand, A., Schrauwen, B., & Martens, J. (2010) · Ghent University Academic Bibliography (Ghent University) · Cited by 207
Jiang, J., & Lai, Y. (2019) · Physical Review Research · doi:10.1103/physrevresearch.1.033056 · Cited by 168
Nonlinearity in memristors for neuromorphic dynamic systems
Yang, K., Joshua Yang, J., Huang, R., & Yang, Y. (2022) · Small Science · doi:10.1002/smsc.202100049 · Cited by 151
Electrocardiogram Classification Using Reservoir Computing With Logistic Regression
Escalona-Morán, M., Soriano, M. C., Fischer, I., & Mirasso, C. R. (2014) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2014.2332001 · Cited by 165
Emerging 2D Ferroelectric Devices for In‐Sensor and In‐Memory Computing
Chen, C., Zhou, Y., Tong, L., Pang, Y., & Xu, J. (2024) · Advanced Materials · doi:10.1002/adma.202400332 · Cited by 157
Proton exchange membrane fuel cell ageing forecasting algorithm based on Echo State Network
Morando, S., Jemeï, S., Hissel, D., Gouriveau, R., & Noureddine Zerhouni, A. (2016) · International Journal of Hydrogen Energy · doi:10.1016/j.ijhydene.2016.05.286 · Cited by 150
Synaptic pathways in neural microcircuits
Silberberg, G., Grillner, S. E., LeBeau, F. E. N., Maex, R., & Markram, H. (2005) · Trends in Neurosciences · doi:10.1016/j.tins.2005.08.004 · Cited by 163
Self-citationHoerzer, G. M., Legenstein, R., & Maass, W. (2014) · Cerebral Cortex · doi:10.1093/cercor/bhs348 · Cited by 139
“It has already been shown that networks of spiking neurons with stereotypical connection probabilities can in principle support a large variety of computational tasks (Maass et al. 2002; Haeusler and Maass 2007; Haeusler et al. 2008; Buonomano and Maass 2009).”
InfluentialSelf-citationTo spike or not to spike: A probabilistic spiking neuron model
Kasabov, N. K. (2009) · Neural Networks · doi:10.1016/j.neunet.2009.08.010 · Cited by 174
Constructing optimized binary masks for reservoir computing with delay systems
Appeltant, L., Van der Sande, G., Danckaert, J., & Fischer, I. (2014) · Scientific Reports · doi:10.1038/srep03629 · Cited by 144
“(2) is numerically integrated using Heun’s method with a time step of 0.”
MethodA Neuromorphic Event-Based Neural Recording System for Smart Brain-Machine-Interfaces
Corradi, F., & Indiveri, G. (2015) · IEEE Transactions on Biomedical Circuits and Systems · doi:10.1109/tbcas.2015.2479256 · Cited by 157
“Spiking neural networks have been shown to be able to carry out complex spatio-temporal processing and classification tasks [34]–[37].”
BackgroundIdentification of pre-sliding and sliding friction dynamics: Grey box and black-box models
Worden, K., Wong, C. X., Parlitz, U., Hornstein, A., Engster, D., Tjahjowidodo, T., Al‐Bender, F., Rizos, D. D., & Fassois, S. D. (2005) · Mechanical Systems and Signal Processing · doi:10.1016/j.ymssp.2005.09.004 · Cited by 147
“The local dynamics of the elements in the RNN used here for friction modelling follow x i ðtÞ ¼ a i x i ðt À 1Þ þ sðI i ðtÞÞ (24) with the local input I i ðtÞ ¼ w i0 þ X jai w ij x j ðt À 1Þ þ v i uðtÞ, (25) where x i ðtÞ is the state of the ith element at time step t, w ij is the connection strength of the output fro…”
BackgroundInfluential25 years of criticality in neuroscience — established results, open controversies, novel concepts
Wilting, J., & Priesemann, V. (2019) · Current Opinion in Neurobiology · doi:10.1016/j.conb.2019.08.002 · Cited by 172
“Tuning models towards critical phase transitions has been shown to maximize the number of metastable states [12], the dynamic range [13, 14], information transmission in terms of mutual information [15, 16, 17], active information storage [18], and computational power in terms of input-output mappings [19, 20, 21].”
BackgroundShort-term synaptic plasticity in emerging devices for neuromorphic computing
Li, C., Zhang, X., Chen, P., Zhou, K., Yu, J., Wu, G., Xiang, D., Jiang, H., Wang, M., & Liu, Q. (2023) · iScience · doi:10.1016/j.isci.2023.106315 · Cited by 158
Personal recommendation using deep recurrent neural networks in NetEase
Wu, S., Ren, W., Yu, C., Chen, G., Zhang, D., & Jing-Bo, Z. H. U. (2016) · IEEE International Conference on Data Engineering (ICDE) · doi:10.1109/icde.2016.7498326 · Cited by 145
“Alternatively, a fundamental new training technique, Reservoir Computing, was introduced [16][24][14].”
MethodMagnetic skyrmions for unconventional computing
Li, S., Kang, W., Zhang, X., Nie, T., Zhou, Y., Wang, K. L., & Zhao, W. S. (2020) · Materials Horizons · doi:10.1039/d0mh01603a · Cited by 159
Reservoir-based techniques for speech recognition
Verstraeten, D., Schrauwen, B., & Stroobandt, D. (2006) · The 2006 IEEE International Joint Conference on Neural Network Proceedings · doi:10.1109/ijcnn.2006.1716215 · Cited by 131
“The main topic of that article is an investigation of different possible speech frontends, but one of its conclusions is that the LSM is well suited for this task, and that a proper configuration even yields a recognition accuracy that is competitive with stateof-the-art techniques.”
MethodInfluentialNguimdo, R. M., Verschaffelt, G., Danckaert, J., & der Sande, G. V. (2015) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2015.2404346 · Cited by 149
“achieved [1]–[4].”
BackgroundToyoizumi, T., & Abbott, L. F. (2011) · Physical Review E · doi:10.1103/physreve.84.051908 · Cited by 140
Phase-change memtransistive synapses for mixed-plasticity neural computations
Sarwat, S. G., Kersting, B., Moraitis, T., Jonnalagadda, V. P., & Sebastian, A. (2022) · Nature Nanotechnology · doi:10.1038/s41565-022-01095-3 · Cited by 156
“In biological neurons, the short-term dynamics of synapses have been suggested to perform the task of integration of information 31,32 .”
A brain-inspired spiking neural network model with temporal encoding and learning
Yu, Q., Tang, H., Tan, K. C., & Yu, H. (2014) · Neurocomputing · doi:10.1016/j.neucom.2013.06.052 · Cited by 143
“Another possible future direction is to combine our model with other compatible spiking neural networks such as Liquid State Machine (LSM) [55].”
BackgroundMethodIntrinsic neuronal dynamics predict distinct functional roles during working memory
Wasmuht, D. F., Spaak, E., Buschman, T. J., Miller, E. K., & Stokes, M. G. (2018) · Nature Communications · doi:10.1038/s41467-018-05961-4 · Cited by 128
Chen, P., Liu, R., Aihara, K., & Chen, L. (2020) · Nature Communications · doi:10.1038/s41467-020-18381-0 · Cited by 160
A Long Short-Term Memory for AI Applications in Spike-based Neuromorphic Hardware
Rao, A., Plank, P., Wild, A., & Maass, W. (2022) · Nature Machine Intelligence · doi:10.1038/s42256-022-00480-w · Cited by 141
“VPSC,j[t + Δt] → 0 VAHP,j[t + Δt] → 0 if zj[t + Δt] = 1 (10)”
Self-citationThe challenges of modern computing and new opportunities for optics
Li, C., Zhang, X., Li, J., Fang, T., & Dong, X. (2021) · PhotoniX · doi:10.1186/s43074-021-00042-0 · Cited by 152
“Reservoir computing (RC), which find its roots in the concept of liquid-state machine [66] and echo state networks [67], is a novel computational framework derived from recurrent neural networks (RNNs) [68].”
BackgroundRotating neurons for all-analog implementation of cyclic reservoir computing
Liang, X., Zhong, Y., Tang, J., Liu, Z., Yao, P., Sun, K., Zhang, Q., Gao, B., Heidari, H., Qian, H., et al. (2022) · Nature Communications · doi:10.1038/s41467-022-29260-1 · Cited by 149
“(3), it appears that a dynamic neuron for the proposed RNR should satisfy three important characteristics as shown in Fig.”
The unfolding argument: Why IIT and other causal structure theories cannot explain consciousness
Doerig, A., Schurger, A., Hess, K., & Herzog, M. H. (2019) · Consciousness and Cognition · doi:10.1016/j.concog.2019.04.002 · Cited by 140
Comparison of supervised learning methods for spike time coding in spiking neural networks
Kasiński, A., & Ponulak, F. (2006) · Digital library of Zielona Gora (University of Zielona Góra) · Cited by 140
Optical Reservoir Computing Using Multiple Light Scattering for Chaotic Systems Prediction
Dong, J., Rafayelyan, M. S., Krząkała, F., & Gigan, S. (2019) · IEEE Journal of Selected Topics in Quantum Electronics · doi:10.1109/jstqe.2019.2936281 · Cited by 153
“Two other properties to understand the reservoir dynamics are the separation and approximation properties [9].”
BackgroundLearning function from structure in neuromorphic networks
Suárez, L. E., Richards, B. A., Lajoie, G., & Misic, B. (2021) · Nature Machine Intelligence · doi:10.1038/s42256-021-00376-1 · Cited by 134
Compact hardware liquid state machines on FPGA for real-time speech recognition
Schrauwen, B., D’Haene, M., Verstraeten, D., & Campenhout, J. V. (2007) · Neural Networks · doi:10.1016/j.neunet.2007.12.009 · Cited by 162
Furuta, T., Fujii, K., Nakajima, K., Tsunegi, S., Kubota, H., Suzuki, Y., & Miwa, S. (2018) · Physical Review Applied · doi:10.1103/physrevapplied.10.034063 · Cited by 146
“This type of simple RNN is called reservoir computing (RC) [12-14].”
BackgroundReal-Time Ultra-Low Power ECG Anomaly Detection Using an Event-Driven Neuromorphic Processor
Bauer, F. C., Muir, D. R., & Indiveri, G. (2019) · IEEE Transactions on Biomedical Circuits and Systems · doi:10.1109/tbcas.2019.2953001 · Cited by 150
“5 and has been inspired by the paradigm of reservoir computing [26], [27].”
BackgroundPreparatory activity and the expansive null-space
Churchland, M. M., & Shenoy, K. V. (2024) · Nature reviews. Neuroscience · doi:10.1038/s41583-024-00796-z · Cited by 164
“Before it became possible to reliably train recurrent networks on motor tasks, it was known that they could generate rich activity patterns that might form building blocks of outgoing motor commands 57,62,63,105 .”
Hou, Y., Xia, G.-Q., Yang, W.-Y., Wang, D., Jayaprasath, E., Jiang, Z., Hu, C., & Wu, Z.-M. (2018) · Optics Express · doi:10.1364/oe.26.010211 · Cited by 143
Metadata
- Authors
- Maass, W., Natschläger, T., & Markram, H.
- Affiliations
- Graz University of Technology (AT)
- École Polytechnique Fédérale de Lausanne (CH)
- Corresponding author
- Maass, Wolfgang
- Natschläger, Thomas
- Markram, Henry
- Venue
- Neural Computation
- Publisher
- MIT Press
- Published
- 2002-11-01
- Volume
- 14
- Issue
- 11
- Pages
- 2531-2560
- ISSN
- 0899-7667 (ISSN-L)
- 1530-888X
- OpenAlex
- W2103179919
- Cited by
- 4,025
- References
- 28
- Citations by year
- 2026: 286 · 2025: 351 · 2024: 307 · 2023: 309 · 2022: 276 · 2021: 303 · 2020: 239 · 2019: 244 · 2018: 217 · 2017: 173 · 2016: 155 · 2015: 135 · 2014: 142 · 2013: 139 · 2012: 125
- FWCI
- 18.37
- Type
- journal article
- Language
- en
- Open access
- green
- infoscience.epfl.ch
- Locations
- Neural Computation (publishedVersion)
- PubMed (publishedVersion)
- Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion)
- Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion, cc-by-nc-nd)
- Topics
- Life Sciences › Neuroscience › Cognitive Neuroscience › Neural dynamics and brain function
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Reservoir Computing
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Keywords
- liquid state machines, recurrent neural circuits, transient dynamics, high-dimensional dynamical systems, fading memory, integrate-and-fire neurons, real-time processing, time-varying inputs, neural coding, statistical learning theory, computational power, attractor networks
- MeSH terms
- Action Potentials, Computer Simulation, Computer Systems, Computers, Models, Neurological, Neurons, Neural Networks, Computer
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.