AI Paper Atlas

Research map / Architectures and efficiency / Spiking networks

Time to Spike? Understanding the Representational Power of Spiking Neural Networks in Discrete Time

Duc Anh Nguyen, Ernesto Araya, Adalbert Fono, Gitta Kutyniok

ICML 2025 · Architectures and efficiency › Spiking networks · approach: Training deep spiking networks

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Theoretical analysis of discrete-time leaky integrate-and-fire spiking neural networks: universal approximation with explicit neuron counts, a tight bound on the number of linear/constant regions that grows quadratically in latency T, and experiments on bottleneck width and hyperplane shifts.

Problem

Expressivity of spiking neural networks in discrete time (LIF neurons with direct encoding and membrane potential output) is theoretically poorly understood, especially how latency and depth affect the input-space partitioning and approximation.

Approach

Formalizes discrete-time LIF-SNNs as piecewise constant functions on polyhedral regions, shows that each first-layer neuron induces parallel hyperplanes across time (giving at most (T^2+T+2)/2 regions per neuron), and derives universal approximation with explicit width bounds; depth adds no new activation regions.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Increase SNN latencyDiscrete-time LIF-SNN with a 20-neuron first hidden-layer bottleneck · CIFAR10 training settraining accuracyImproved
Widen subsequent hidden layers from 20 to 200 neuronsLow-latency SNN with a 20-neuron first hidden-layer bottleneck · CIFAR10 training settraining accuracy gain: approximately 7%Improved
Widen subsequent hidden layers from 20 to 200 neuronsSNN with T=4 or T=8 and a 20-neuron first hidden-layer bottleneck · CIFAR10 training settraining accuracy gain: almost reaching 20%Improved
Widen subsequent hidden layers from 20 to 200 neuronsANN with a 20-neuron first hidden-layer bottleneck · CIFAR10 training settraining accuracy gain: approximately 21%Improved
Widen subsequent hidden layers from 20 to 200 neuronsSNN with T=16 and a 20-neuron first hidden-layer bottleneck · CIFAR10 training settraining accuracy gain: approximately 30%Improved
Train the SNN on the linearly separable toy datasetDiscrete-time LIF-SNN, T=1, n1=4, layer 1 · Linearly separable toy datasetnumber of regions: 10 -> 6Hurt
Train the SNN on the linearly separable toy datasetDiscrete-time LIF-SNN, T=2, n1=2, layer 1 · Linearly separable toy datasetnumber of regions: 16 -> 4Hurt
Train the SNN on the linearly separable toy datasetDiscrete-time LIF-SNN, T=2, n1=3, layer 1 · Linearly separable toy datasetnumber of regions: 35 -> 10Hurt
Train the SNN on the linearly separable toy datasetDiscrete-time LIF-SNN, T=2, n1=4, layer 1 · Linearly separable toy datasetnumber of regions: 59 -> 12Hurt
Train the SNN on the linearly separable toy datasetDiscrete-time LIF-SNN, T=1, n1=3, layer 2 · Linearly separable toy datasetnumber of regions: 4 -> 3Hurt
Train the SNN on the linearly separable toy datasetDiscrete-time LIF-SNN, T=2, n1=2, layer 2 · Linearly separable toy datasetnumber of regions: 14 -> 3Hurt

Surprising findings

Limitations

Experimental setup

Datasets: CIFAR10 toy linearly separable dataset in [-1,1]^2

Benchmarks: CIFAR10

Metrics: training accuracy number of linear regions

Baselines: ReLU ANN Heaviside ANN

Builds on: Heaviside ANN expressivity results linear region counting for ReLU networks spike response model

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy