# Surrogate Gradient Learning in Spiking Neural Networks

@article{Neftci2019SurrogateGL, title={Surrogate Gradient Learning in Spiking Neural Networks}, author={Emre O. Neftci and Hesham Mostafa and Friedemann Zenke}, journal={ArXiv}, year={2019}, volume={abs/1901.09948} }

Spiking neural networks are nature's versatile solution to fault-tolerant and energy efficient signal processing. To translate these benefits into hardware, a growing number of neuromorphic spiking neural network processors attempt to emulate biological neural networks. These developments have created an imminent need for methods and tools to enable such systems to solve real-world signal processing problems. Like conventional neural networks, spiking neural networks can be trained on real… Expand

#### Supplemental Code

#### 105 Citations

An Introduction to Probabilistic Spiking Neural Networks.

- Computer Science
- 2019

This article adopts discrete-time probabilistic models for networked spiking neurons and derive supervised and unsupervised learning rules from first principles via variational inference that enables the direct derivation of learning rules by leveraging the unique time-encoding capabilities of SNNs. Expand

An Introduction to Spiking Neural Networks: Probabilistic Models, Learning Rules, and Applications.

- Computer Science
- 2018

This paper adopts discrete-time probabilistic models for networked spiking neurons, and it derives supervised and unsupervised learning rules from first principles by using variational inference. Expand

MULTIPLE-TIMESCALE SPIKING RECURRENT NEURAL NETWORKS

- 2020

The emergence of brain-inspired neuromorphic computing as a paradigm for edge AI is motivating the search for high-performance and efficient spiking neural networks to run on this hardware. However,… Expand

Training Deep Spiking Neural Networks for Energy-Efficient Neuromorphic Computing

- Computer Science
- ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
- 2020

This paper presents biologically plausible Spike Timing Dependent Plasticity based deterministic and stochastic algorithms for unsupervised representation learning in SNNs, and proposes conversion methodology to map off-the-shelf trained ANN to SNN for energy-efficient inference. Expand

Explicitly Trained Spiking Sparsity in Spiking Neural Networks with Backpropagation

- Computer Science
- ArXiv
- 2020

This work proposes an explicit inclusion of spike counts in the loss function, along with a traditional error loss, causing the backpropagation learning algorithms to optimize weight parameters for both accuracy and spiking sparsity. Expand

An Introduction to Probabilistic Spiking Neural Networks: Probabilistic Models, Learning Rules, and Applications

- Computer Science, Engineering
- IEEE Signal Processing Magazine
- 2019

This work has shown that the sparsity of the synaptic spiking inputs and the corresponding event-driven nature of neural processing can be leveraged by energy-efficient hardware implementations, which can offer significant energy reductions as compared to conventional artificial neural networks. Expand

A Tandem Learning Rule for Efficient and Rapid Inference on Deep Spiking Neural Networks.

- Computer Science
- 2019

The spike count is considered as the discrete neural representation and design ANN neuronal activation function that can effectively approximate the spike count of the coupled SNN in a tandem learning framework that consists of a SNN and an Artificial Neural Network that share weights. Expand

Enabling Deep Spiking Neural Networks with Hybrid Conversion and Spike Timing Dependent Backpropagation

- Computer Science, Mathematics
- ICLR
- 2020

The proposed training methodology converges in less than 20 epochs of spike-based backpropagation for most standard image classification datasets, thereby greatly reducing the training complexity compared to training SNNs from scratch. Expand

Minibatch Processing in Spiking Neural Networks

- Computer Science
- ArXiv
- 2019

To the knowledge, this is the first general-purpose implementation of mini-batch processing in a spiking neural networks simulator, which works with arbitrary neuron and synapse models and shows the effectiveness of large batch sizes in two SNN application domains. Expand

Stochasticity and Robustness in Spiking Neural Networks

- Computer Science, Biology
- Neurocomputing
- 2021

It is demonstrated that noise can be used to make the behavior of IF neurons more robust to synaptic inaccuracy, and it is shown that a noisy network can tolerate the inaccuracy expected when hafnium-oxide based resistive random-access memory is used to encode synaptic weights. Expand

#### References

SHOWING 1-10 OF 42 REFERENCES

Deep Networks Incorporating Spiking Neural Dynamics

- Computer Science
- ArXiv
- 2018

An alternative perspective on the spiking neuron as a particular ANN construct called Spiking Neural Unit (SNU) is proposed, which provides a systematic methodology for implementing and training deep networks incorporating spiking dynamics that achieve accuracies as high, or better than, state-of-the-art ANNs. Expand

Deep Learning in Spiking Neural Networks

- Computer Science, Medicine
- Neural Networks
- 2019

The emerging picture is that SNNs still lag behind ANNs in terms of accuracy, but the gap is decreasing, and can even vanish on some tasks, while SNN's typically require many fewer operations and are the better candidates to process spatio-temporal data. Expand

Gradient Descent for Spiking Neural Networks

- Computer Science, Mathematics
- NeurIPS
- 2018

A gradient descent method for optimizing spiking network models by introducing a differentiable formulation of spiking networks and deriving the exact gradient calculation offers a general purpose supervised learning algorithm for spiking neural networks, thus advancing further investigations on spike-based computation. Expand

Supervised Learning Based on Temporal Coding in Spiking Neural Networks

- Computer Science, Medicine
- IEEE Transactions on Neural Networks and Learning Systems
- 2018

This work shows that in a feedforward spiking network that uses a temporal coding scheme where information is encoded in spike times instead of spike rates, the network input–output relation is differentiable almost everywhere and this relation is piecewise linear after a transformation of variables. Expand

Training Deep Spiking Neural Networks Using Backpropagation

- Computer Science, Medicine
- Front. Neurosci.
- 2016

A novel technique is introduced, which treats the membrane potentials of spiking neurons as differentiable signals, where discontinuities at spike times are considered as noise, which enables an error backpropagation mechanism for deep SNNs that follows the same principles as in conventional deep networks, but works directly on spike signals and membranes potentials. Expand

Spiking Deep Networks with LIF Neurons

- Computer Science
- ArXiv
- 2015

This work demonstrates that biologically-plausible spiking LIF neurons can be integrated into deep networks can perform as well as other spiking models (e.g. integrate-and-fire), and provides new methods for training deep networks to run on neuromorphic hardware. Expand

SuperSpike: Supervised Learning in Multilayer Spiking Neural Networks

- Computer Science, Biology
- Neural Computation
- 2018

SuperSpike is derived, 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. Expand

Biologically inspired alternatives to backpropagation through time for learning in recurrent neural nets

- Computer Science
- ArXiv
- 2019

It is shown that an online merging of locally available information during a computation with suitable top-down learning signals in real-time provides highly capable approximations to back-propagation through time (BPTT). Expand

Supervised learning in spiking neural networks with FORCE training

- Biology, Computer Science
- Nature Communications
- 2017

The direct applicability of the FORCE method to spiking neural networks is demonstrated and it is demonstrated that these networks can be trained to exhibit different dynamic behaviours. Expand

Synaptic Plasticity Dynamics for Deep Continuous Local Learning (DECOLLE)

- Computer Science, Medicine
- Frontiers in Neuroscience
- 2020

DECOLLE networks provide continuously learning machines that are relevant to biology and supportive of event-based, low-power computer vision architectures matching the accuracies of conventional computers on tasks where temporal precision and speed are essential. Expand