A Time Encoding Approach to Training Spiking Neural Networks

Karen Adam
IEEE, 2022
Conference paper

Abstract

While Spiking Neural Networks (SNNs) have been gaining in popularity, it seems that the algorithms used to train them are not powerful enough to solve the same tasks as those tackled by classical Artificial Neural Networks (ANNs).In this paper, we provide an extra tool to help us understand and train SNNs by using theory from the field of time encoding. Time encoding machines (TEMs) can be used to model integrate-and-fire neurons and have well-understood reconstruction properties.We will see how one can take inspiration from the field of TEMs to interpret the spike times of SNNs as constraints on the SNNs' weight matrices. More specifically, we study how to train one-layer SNNs by solving a set of linear constraints, and how to train two-layer SNNs by leveraging the all-or-none and asynchronous properties of the spikes emitted by SNNs. These properties of spikes result in an alternative to backpropagation which is not possible in the case of simultaneous and graded activations as in classical ANNs.

Official source

https://infoscience.epfl.ch/record/296073?ln=en

About this result

This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.

Related concepts

Related publications

Related publications (21)

How single neuron properties shape chaotic dynamics and signal transmission in random neural networks

Wulfram Gerstner, Samuel Pavio Muscinelli, Tilo Schwalger

While most models of randomly connected neural networks assume single-neuron models with simple dynamics, neurons in the brain exhibit complex intrinsic dynamics over multiple timescales. We analyze how the dynamical properties of single neurons and recurrent connections interact to shape the effective dynamics in large randomly connected networks. A novel dynamical mean-field theory for strongly connected networks of multi-dimensional rate neurons shows that the power spectrum of the network activity in the chaotic phase emerges from a nonlinear sharpening of the frequency response function of single neurons. For the case of two-dimensional rate neurons with strong adaptation, we find that the network exhibits a state of resonant chaos, characterized by robust, narrow-band stochastic oscillations. The coherence of stochastic oscillations is maximal at the onset of chaos and their correlation time scales with the adaptation timescale of single units. Surprisingly, the resonance frequency can be predicted from the properties of isolated neurons, even in the presence of heterogeneity in the adaptation parameters. In the presence of these internally-generated chaotic fluctuations, the transmission of weak, low-frequency signals is strongly enhanced by adaptation, whereas signal transmission is not influenced by adaptation in the non-chaotic regime. Our theoretical framework can be applied to other mechanisms at the level of single neurons, such as synaptic filtering, refractoriness or spike synchronization. These results advance our understanding of the interaction between the dynamics of single units and recurrent connectivity, which is a fundamental step toward the description of biologically realistic neural networks. Author summary Biological neural networks are formed by a large number of neurons whose interactions can be extremely complex. Such systems have been successfully studied using random network models, in which the interactions among neurons are assumed to be random. However, the dynamics of single units are usually described using over-simplified models, which might not capture several salient features of real neurons. Here, we show how accounting for richer single-neuron dynamics results in shaping the network dynamics and determines which signals are better transmitted. We focus on adaptation, an important mechanism present in biological neurons that consists in the decrease of their firing rate in response to a sustained stimulus. Our mean-field approach reveals that the presence of adaptation shifts the network into a previously unreported dynamical regime, that we term resonant chaos, in which chaotic activity has a strong oscillatory component. Moreover, we show that this regime is advantageous for the transmission of low-frequency signals. Our work bridges the microscopic dynamics (single neurons) to the macroscopic dynamics (network), and shows how the global signal-transmission properties of the network can be controlled by acting on the single-neuron dynamics. These results paves the way for further developments that include more complex neural mechanisms, and considerably advance our understanding of realistic neural networks.

2019

The way our brain learns to disentangle complex signals into unambiguous concepts is fascinating but remains largely unknown. There is evidence, however, that hierarchical neural representations play a key role in the cortex. This thesis investigates biologically plausible models of unsupervised learning of hierarchical representations as found in the brain and modern computer vision models. We use computational modeling to address three main questions at the intersection of artificial intelligence (AI) and computational neuroscience.The first question is: What are useful neural representations and when are deep hierarchical representations needed? We approach this point with a systematic study of biologically plausible unsupervised feature learning in a shallow 2-layer networks on digit (MNIST) and object (CIFAR10) classification. Surprisingly, random features support high performance, especially for large hidden layers. When combined with localized receptive fields, random feature networks approach the performance of supervised backpropagation on MNIST, but not on CIFAR10. We suggest that future models of biologically plausible learning should outperform such random feature benchmarks on MNIST, or that such models should be evaluated in different ways.The second question is: How can hierarchical representations be learned with mechanisms supported by neuroscientific evidence? We cover this question by proposing a unifying Hebbian model, inspired by common models of V1 simple and complex cells based on unsupervised sparse coding and temporal invariance learning. In shallow 2-layer networks, our model reproduces learning of simple and complex cell receptive fields, as found in V1. In deeper networks, we stack multiple layers of Hebbian learning but find that it does not yield hierarchical representations of increasing usefulness. From this, we hypothesise that standard Hebbian rules are too constrained to build increasingly useful representations, as observed in higher areas of the visual cortex or deep artificial neural networks.The third question is: Can AI inspire learning models that build deep representations and are still biologically plausible? We address this question by proposing a learning rule that takes inspiration from neuroscience and recent advances in self-supervised deep learning. The proposed rule is Hebbian, i.e. only depends on pre- and post-synaptic neuronal activity, but includes additional local factors, namely predictive dendritic input and widely broadcasted modulation factors. Algorithmically, this rule applies self-supervised contrastive predictive learning to a causal, biological setting using saccades. We find that networks trained with this generalised Hebbian rule build deep hierarchical representations of images, speech and video.We see our modeling as a potential starting point for both, new hypotheses, that can be tested experimentally, and novel AI models that could benefit from added biological realism.

EPFL2021

Timing is Everything

Karen Adam

In this thesis, timing is everything. In the first part, we mean this literally, as we tackle systems that encode information using timing alone. In the second part, we adopt the standard, metaphoric interpretation of this saying and show the importance of choosing the right time to sample a system, when efficiency is a key consideration.Time encoding machines, or, alternately, integrate-and-fire neurons, encode their inputs using spikes, the timings of which depend on the input and therefore hold information about it. These devices can be made more power efficient than their clocked counterparts and have thus been studied in the fields of signal processing, event-based vision, computational neuroscience and machine learning. However, their timing-based spiking output has so far often been considered a nuisance that one must make do with, rather than a potential advantage. In this thesis, we show that this timing-based output equips spiking devices with capabilities that are out of reach for classical encoding and processing systems.We first discover the benefits of time encoding on multi-channel encoding and recovery of a signal: with time encoding, clock alignment is easy to solve, although it poses problems in the classical sampling scenario. Then, we study the time encoding of low-dimensional signals and see that the asynchrony of spikes allows for a lower sample complexity in comparison with synchronous sampling. Thanks to this same asynchrony, time encoding of video results in an entanglement between spatial sampling density and temporal resolution---a relationship which is not present in frame-based video. Finally, we show that the all-or-none nature of the spikes allows training spiking neural networks in a layer-by-layer fashion---a feat that is impossible with clocked, artificial neural networks, due to the credit assignment problem.The second part of this thesis shows that choosing the right timing of samples can be crucial to ensure efficiency when performing nanoscale magnetic sensing. We are given a stochastic process, where each sample at time t follows a Bernoulli distribution, and which is characterized by oscillation frequencies that we are interested in recovering. We search for an optimal approach to sample this process, such that the variance of the frequencies' estimates is minimized, given constraints on the measurement time. The models we assume stem from the field of nanoscale magnetic sensing, where the number of parameters to be estimated varies with the number of spins one is trying to sense. We present an adaptive approach to choosing samples in both the single-spin and two-spin cases and compare the adaptive algorithm's performance to classical approaches to sampling.In both parts of the thesis, we move away from classical amplitude sampling and consider cases where timing takes the forefront and amplitude information is merely binary, to shows that timing can carry information and that it can control the amount of information gain.

EPFL2022