Simple framework for constructing functional spiking recurrent neural networks

Cortical microcircuits exhibit complex recurrent architectures that possess dynamically rich properties. The neurons that make up these microcircuits communicate mainly via discrete spikes, and it is not clear how spikes give rise to dynamics that can be used to perform computationally challenging tasks. In contrast, continuous models of rate-coding neurons can be trained to perform complex tasks. Here, we present a simple framework to construct biologically realistic spiking recurrent neural networks (RNNs) capable of learning a wide range of tasks. Our framework involves training a continuous-variable rate RNN with important biophysical constraints and transferring the learned dynamics and constraints to a spiking RNN in a one-to-one manner. We validate our framework on several cognitive task paradigms to replicate previously observed experimental results. We also demonstrate different ways to exploit the biological features of our models to elucidate neural mechanisms underlying cognitive functions.

[1]  Valentin Dragoi,et al.  High order coordination of cortical spiking activity modulates perceptual accuracy , 2019, Nature Neuroscience.

[2]  Terrence J. Sejnowski,et al.  Gradient Descent for Spiking Neural Networks , 2017, NeurIPS.

[3]  A. Riehle,et al.  The Local Field Potential Reflects Surplus Spike Synchrony , 2010, Cerebral cortex.

[4]  Francesca Mastrogiuseppe,et al.  Linking Connectivity, Dynamics, and Computations in Low-Rank Recurrent Neural Networks , 2017, Neuron.

[5]  N. Sigala,et al.  Dynamic Coding for Cognitive Control in Prefrontal Cortex , 2013, Neuron.

[6]  Paul J. Werbos,et al.  Backpropagation Through Time: What It Does and How to Do It , 1990, Proc. IEEE.

[7]  L. Abbott,et al.  From fixed points to chaos: Three models of delayed discrimination , 2013, Progress in Neurobiology.

[8]  Steven W Kennerley,et al.  Reconciling persistent and dynamic hypotheses of working memory coding in prefrontal cortex , 2017, Nature Communications.

[9]  Tiago Branco,et al.  Dendritic nonlinearities are tuned for efficient spike-based computations in cortical circuits , 2015, eLife.

[10]  Terrence J Sejnowski,et al.  Simple framework for constructing functional spiking recurrent neural networks , 2019, Proceedings of the National Academy of Sciences.

[11]  Mark M. Churchland,et al.  Using Firing-Rate Dynamics to Train Recurrent Networks of Spiking Model Neurons , 2016, 1601.07620.

[12]  Angelica Foggetti,et al.  Parvalbumin-positive interneurons of the prefrontal cortex support working memory and cognitive flexibility , 2015, Scientific Reports.

[13]  Matthew Cook,et al.  Fast-classifying, high-accuracy spiking deep networks through weight and threshold balancing , 2015, 2015 International Joint Conference on Neural Networks (IJCNN).

[14]  Nava Levit-Binnun,et al.  A quantitative physical model of the TMS-induced discharge artifacts in EEG , 2018, PLoS Comput. Biol..

[15]  Sommers,et al.  Chaos in random neural networks. , 1988, Physical review letters.

[16]  Joel Z. Leibo,et al.  Prefrontal cortex as a meta-reinforcement learning system , 2018, bioRxiv.

[17]  Thomas Miconi Biologically plausible learning in recurrent neural networks reproduces neural dynamics observed during cognitive tasks , 2017 .

[18]  Thomas Miconi,et al.  Biologically plausible learning in recurrent neural networks reproduces neural dynamics observed during cognitive tasks , 2016, bioRxiv.

[19]  Hilbert J. Kappen,et al.  Learning Universal Computations with Spikes , 2015, PLoS Comput. Biol..

[20]  Christopher Kim,et al.  Learning recurrent dynamics in spiking networks , 2018, bioRxiv.

[21]  Jean-Jacques E. Slotine,et al.  Learning Nonlinear Dynamics in Efficient, Balanced Spiking Networks Using Local Plasticity Rules , 2018, AAAI.

[22]  Iulia-Alexandra Lungu,et al.  Theory and Tools for the Conversion of Analog to Spiking Convolutional Neural Networks , 2016, ArXiv.

[23]  Jessica A. Cardin,et al.  Inhibitory Interneurons Regulate Temporal Precision and Correlations in Cortical Circuits , 2018, Trends in Neurosciences.

[24]  Xiao-Jing Wang,et al.  A dendritic disinhibitory circuit mechanism for pathway-specific gating , 2016, Nature Communications.

[25]  Dean V. Buonomano,et al.  ROBUST TIMING AND MOTOR PATTERNS BY TAMING CHAOS IN RECURRENT NEURAL NETWORKS , 2012, Nature Neuroscience.

[26]  E. Miller,et al.  Intrinsic neuronal dynamics predict distinct functional roles during working memory , 2017, Nature Communications.

[27]  Dhanistha Panyasak,et al.  Circuits , 1995, Annals of the New York Academy of Sciences.

[28]  David J. Freedman,et al.  Dynamic population coding of category information in inferior temporal and prefrontal cortex. , 2008, Journal of neurophysiology.

[29]  Theoden I. Netoff,et al.  Blocking NMDAR Disrupts Spike Timing and Decouples Monkey Prefrontal Circuits: Implications for Activity-Dependent Disconnection in Schizophrenia , 2018, Neuron.

[30]  Deepak Khosla,et al.  Spiking Deep Convolutional Neural Networks for Energy-Efficient Object Recognition , 2014, International Journal of Computer Vision.

[31]  Zhewei Zhang,et al.  A neural network model for the orbitofrontal cortex and task space acquisition during reinforcement learning , 2017, bioRxiv.

[32]  Wenjun Gao,et al.  PV Interneurons: Critical Regulators of E/I Balance for Prefrontal Cortex-Dependent Behavior and Psychiatric Disorders , 2018, Front. Neural Circuits.

[33]  P. Goldman-Rakic Cellular basis of working memory , 1995, Neuron.

[34]  Kaushik Roy,et al.  Going Deeper in Spiking Neural Networks: VGG and Residual Architectures , 2018, Front. Neurosci..

[35]  Ilya Sutskever,et al.  Learning Recurrent Neural Networks with Hessian-Free Optimization , 2011, ICML.

[36]  Chris Eliasmith,et al.  Training Spiking Deep Networks for Neuromorphic Hardware , 2016, ArXiv.

[37]  Peter Ford Dominey,et al.  Reservoir Computing Properties of Neural Dynamics in Prefrontal Cortex , 2016, PLoS Comput. Biol..

[38]  M. Scanziani,et al.  How Inhibition Shapes Cortical Activity , 2011, Neuron.

[39]  Christian K. Machens,et al.  Efficient codes and balanced networks , 2016, Nature Neuroscience.

[40]  W. Newsome,et al.  Context-dependent computation by recurrent dynamics in prefrontal cortex , 2013, Nature.

[41]  Xiao-Jing Wang Decision Making in Recurrent Neuronal Circuits , 2008, Neuron.

[42]  R. Desimone,et al.  Modulation of Oscillatory Neuronal Synchronization by Selective Visual Attention , 2001, Science.

[43]  Wilten Nicola,et al.  Supervised learning in spiking neural networks with FORCE training , 2016, Nature Communications.

[44]  David J. Freedman,et al.  Computing by Robust Transience: How the Fronto-Parietal Network Performs Sequential, Category-Based Decisions , 2017, Neuron.

[45]  L. F. Abbott,et al.  Building functional networks of spiking model neurons , 2016, Nature Neuroscience.

[46]  Christopher D. Harvey,et al.  Recurrent Network Models of Sequence Generation and Memory , 2016, Neuron.

[47]  Tobi Delbrück,et al.  Training Deep Spiking Neural Networks Using Backpropagation , 2016, Front. Neurosci..

[48]  L. F. Abbott,et al.  Generating Coherent Patterns of Activity from Chaotic Neural Networks , 2009, Neuron.

[49]  Razvan Pascanu,et al.  Advances in optimizing recurrent networks , 2012, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[50]  S. Funahashi,et al.  Stable and Dynamic Coding for Working Memory in Primate Prefrontal Cortex , 2017, The Journal of Neuroscience.

[51]  Guangyu R. Yang,et al.  Training Excitatory-Inhibitory Recurrent Neural Networks for Cognitive Tasks: A Simple and Flexible Framework , 2016, PLoS Comput. Biol..

[52]  Yang Dan,et al.  Dynamic Modification of Cortical Orientation Tuning Mediated by Recurrent Connections , 2002, Neuron.

[53]  Xiao-Jing Wang,et al.  Reward-based training of recurrent neural networks for cognitive and value-based tasks , 2016, bioRxiv.

[54]  Razvan Pascanu,et al.  On the difficulty of training recurrent neural networks , 2012, ICML.

[55]  Alan Carleton,et al.  Restoring wild-type-like CA1 network dynamics and behavior during adulthood in a mouse model of schizophrenia , 2018, Nature Neuroscience.

[56]  Nitin Gupta,et al.  Oscillatory integration windows in neurons , 2016, Nature Communications.

[57]  Andrew S. Cassidy,et al.  Conversion of artificial recurrent neural networks to spiking neural networks for low-power neuromorphic hardware , 2016, 2016 IEEE International Conference on Rebooting Computing (ICRC).