SuperSpike: Supervised Learning in Multilayer Spiking Neural Networks

A vast majority of computation in the brain is performed by spiking neural networks. Despite the ubiquity of such spiking, we currently lack an understanding of how biological spiking neural circuits learn and compute in vivo, as well as how we can instantiate such capabilities in artificial spiking circuits in silico. Here we revisit the problem of supervised learning in temporally coding multilayer spiking neural networks. First, by using a surrogate gradient approach, we derive SuperSpike, a nonlinear voltage-based three-factor learning rule capable of training multilayer networks of deterministic integrate-and-fire neurons to perform nonlinear computations on spatiotemporal spike patterns. Second, inspired by recent results on feedback alignment, we compare the performance of our learning rule under different credit assignment strategies for propagating output errors to hidden units. Specifically, we test uniform, symmetric, and random feedback, finding that simpler tasks can be solved with any type of feedback, while more complex tasks require symmetric feedback. In summary, our results open the door to obtaining a better scientific understanding of learning and computation in spiking neural networks by advancing our ability to train them to solve nonlinear problems involving transformations between different spatiotemporal spike time patterns.

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

[2]  H. Sompolinsky,et al.  The tempotron: a neuron that learns spike timing–based decisions , 2006, Nature Neuroscience.

[3]  Gustavo Carneiro,et al.  On the importance of normalisation layers in deep learning with piecewise linear activation units , 2015, 2016 IEEE Winter Conference on Applications of Computer Vision (WACV).

[4]  Linda Bushnell,et al.  Fast Modifications of the SpikeProp Algorithm , 2006, The 2006 IEEE International Joint Conference on Neural Network Proceedings.

[5]  Jean-Pascal Pfister,et al.  Spike-timing Dependent Plasticity and Mutual Information Maximization for a Spiking Neuron Model , 2004, NIPS.

[6]  Haim Sompolinsky,et al.  Learning Precisely Timed Spikes , 2014, Neuron.

[7]  Nikolaus Kriegeskorte,et al.  Representational Distance Learning for Deep Neural Networks , 2015, Front. Comput. Neurosci..

[8]  Robert Gütig,et al.  Spiking neurons can discover predictive features by aggregate-label learning , 2016, Science.

[9]  Henning Sprekeler,et al.  Functional Requirements for Reward-Modulated Spike-Timing-Dependent Plasticity , 2010, The Journal of Neuroscience.

[10]  W. Senn,et al.  Matching Recall and Storage in Sequence Learning with Spiking Neural Networks , 2013, The Journal of Neuroscience.

[11]  Yoshua Bengio,et al.  Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation , 2013, ArXiv.

[12]  Wieland Brendel,et al.  Learning to represent signals spike by spike , 2017, PLoS Comput. Biol..

[13]  Konrad P. Körding,et al.  Toward an Integration of Deep Learning and Neuroscience , 2016, bioRxiv.

[14]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[15]  F T Zenke.  Memory formation and recall in recurrent spiking neural networks , 2014 .

[16]  Arunava Banerjee,et al.  Learning Precise Spike Train–to–Spike Train Transformations in Multilayer Feedforward Neuronal Networks , 2014, Neural Computation.

[17]  Sander M. Bohte,et al.  Error-Backpropagation in Networks of Fractionally Predictive Spiking Neurons , 2011, ICANN.

[18]  Razvan V. Florian,et al.  The Chronotron: A Neuron That Learns to Fire Temporally Precise Spike Patterns , 2012, PloS one.

[19]  Sander M. Bohte,et al.  Reducing Spike Train Variability: A Computational Theory Of Spike-Timing Dependent Plasticity , 2004, BNAIC.

[20]  Matthew Cook,et al.  Unsupervised learning of digit recognition using spike-timing-dependent plasticity , 2015, Front. Comput. Neurosci..

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

[22]  Jordan Guergiuev,et al.  Biologically feasible deep learning with segregated dendrites , 2016 .

[23]  Peter Auer,et al.  A learning rule for very simple universal approximators consisting of a single layer of perceptrons , 2008, Neural Networks.

[24]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[25]  Jonathan D. Victor,et al.  Metric-space analysis of spike trains: theory, algorithms and application , 1998, q-bio/0309031.

[26]  K. Pawelzik,et al.  Learning of Precise Spike Times with Homeostatic Membrane Potential Dependent Synaptic Plasticity , 2016, PloS one.

[27]  Francis Crick,et al.  The recent excitement about neural networks , 1989, Nature.

[28]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[29]  W. Gerstner,et al.  Neuromodulated Spike-Timing-Dependent Plasticity, and Theory of Three-Factor Learning Rules , 2016, Front. Neural Circuits.

[30]  J. Wickens,et al.  Frontiers in Synaptic Neuroscience Synaptic Neuroscience , 2022 .

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

[32]  Colin J. Akerman,et al.  Random synaptic feedback weights support error backpropagation for deep learning , 2016, Nature Communications.

[33]  Robert A. Legenstein,et al.  Pattern representation and recognition with accelerated analog neuromorphic systems , 2017, 2017 IEEE International Symposium on Circuits and Systems (ISCAS).

[34]  Mark C. W. van Rossum,et al.  A Novel Spike Distance , 2001, Neural Computation.

[35]  Hieu Tat Nguyen,et al.  A gradient descent rule for spiking neurons emitting multiple spikes , 2005, Inf. Process. Lett..

[36]  Wulfram Gerstner,et al.  Towards deep learning with spiking neurons in energy based models with contrastive Hebbian plasticity , 2016, ArXiv.

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

[38]  André Grüning,et al.  Learning Spatiotemporally Encoded Pattern Transformations in Structured Spiking Neural Networks , 2015, Neural Computation.

[39]  Sander M. Bohte,et al.  Reducing the Variability of Neural Responses: A Computational Theory of Spike-Timing-Dependent Plasticity , 2007, Neural Computation.

[40]  Qing Song,et al.  Robust learning in SpikeProp , 2017, Neural Networks.

[41]  Ioana Sporea,et al.  Supervised Learning in Multilayer Spiking Neural Networks , 2012, Neural Computation.

[42]  Trevor Bekolay,et al.  Supplementary Materials for A Large-Scale Model of the Functioning Brain , 2012 .

[43]  T. Toyoizumi,et al.  Learning with three factors: modulating Hebbian plasticity with errors , 2017, Current Opinion in Neurobiology.

[44]  Surya Ganguli,et al.  Deep Learning Models of the Retinal Response to Natural Scenes , 2017, NIPS.

[45]  Denis Fize,et al.  Speed of processing in the human visual system , 1996, Nature.

[46]  Qing Song,et al.  Adaptive learning rate of SpikeProp based on weight convergence analysis , 2015, Neural Networks.

[47]  Stephen Grossberg,et al.  Competitive Learning: From Interactive Activation to Adaptive Resonance , 1987, Cogn. Sci..

[48]  Timothy P Lillicrap,et al.  Towards deep learning with segregated dendrites , 2016, eLife.

[49]  Henry Markram,et al.  Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations , 2002, Neural Computation.

[50]  Andrzej J. Kasinski,et al.  Supervised Learning in Spiking Neural Networks with ReSuMe: Sequence Learning, Classification, and Spike Shifting , 2010, Neural Computation.

[51]  W. Gerstner,et al.  Predicting non-linear dynamics: a stable local learning scheme for recurrent spiking neural networks. , 2017 .

[52]  E. Oja.  Simplified neuron model as a principal component analyzer , 1982, Journal of mathematical biology.

[53]  Jean-Pascal Pfister,et al.  Optimal Spike-Timing-Dependent Plasticity for Precise Action Potential Firing in Supervised Learning , 2005, Neural Computation.

[54]  Wulfram Gerstner,et al.  Neuronal Dynamics: From Single Neurons To Networks And Models Of Cognition , 2014 .

[55]  Everton J. Agnes,et al.  Diverse synaptic plasticity mechanisms orchestrated to form and retrieve memories in spiking neural networks , 2015, Nature Communications.

[56]  Ronald J. Williams,et al.  A Learning Algorithm for Continually Running Fully Recurrent Neural Networks , 1989, Neural Computation.

[57]  Stefan Schliebs,et al.  Span: Spike Pattern Association Neuron for Learning Spatio-Temporal Spike Patterns , 2012, Int. J. Neural Syst..

[58]  Simon de Montigny,et al.  On the Analytical Solution of Firing Time for SpikeProp , 2016, Neural Computation.

[59]  André Grüning,et al.  Supervised Learning in Spiking Neural Networks for Precise Temporal Encoding , 2016, PloS one.

[60]  Parul Parashar,et al.  Neural Networks in Machine Learning , 2014 .

[61]  D. Feldman.  The Spike-Timing Dependence of Plasticity , 2012, Neuron.

[62]  Walter Senn,et al.  A Gradient Learning Rule for the Tempotron , 2009, Neural Computation.

[63]  Wulfram Gerstner,et al.  Stochastic variational learning in recurrent spiking networks , 2014, Front. Comput. Neurosci..

[64]  W. Singer,et al.  Different voltage-dependent thresholds for inducing long-term depression and long-term potentiation in slices of rat visual cortex , 1990, Nature.

[65]  Tim Gollisch,et al.  Rapid Neural Coding in the Retina with Relative Spike Latencies , 2008, Science.

[66]  Sander M. Bohte,et al.  Error-backpropagation in temporally encoded networks of spiking neurons , 2000, Neurocomputing.

[67]  Pierre Baldi,et al.  Learning in the Machine: Random Backpropagation and the Learning Channel , 2016, ArXiv.

[68]  Ha Hong,et al.  Performance-optimized hierarchical models predict neural responses in higher visual cortex , 2014, Proceedings of the National Academy of Sciences.

[69]  Wulfram Gerstner,et al.  Predicting non-linear dynamics by stable local learning in a recurrent spiking neural network , 2017, eLife.

[70]  Wulfram Gerstner,et al.  Limits to high-speed simulations of spiking neural networks using general-purpose computers , 2014, Front. Neuroinform..

[71]  E. Izhikevich.  Solving the distal reward problem through linkage of STDP and dopamine signaling , 2007, BMC Neuroscience.