AI Paper Atlas

Research map / Architectures and efficiency / Spiking networks

Difference Predictive Coding for Training Spiking Neural Networks

Ville Karlsson, Nicklas Fianda, Joni-Kristian Kämäräinen

ICLR 2026 · Architectures and efficiency › Spiking networks · approach: Spiking transformers

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
increase scheduler cycle length n from 3 to 7 (a=1.0) (ablation)DiffPC · Numerical precision vs PC (128-200-10, random data)mean absolute difference: 0.1506 -> 0.0117Improved
decrease decay value a from 1.0 to 0.1 at n=3 (ablation)DiffPC · Numerical precision vs PC (128-200-10, random data)mean absolute difference: 0.1506 -> 0.0168Improved
decrease decay value a from 1.0 to 0.1 at n=7 (ablation)DiffPC · Numerical precision vs PC (128-200-10, random data)mean absolute difference: 0.0117 -> 0.0072Improved
use spike-based DiffPC instead of backpropagation (long variant)CNN 10C5S2-5C5S2-50FC-30FC-10FC · CIFAR-10accuracy (%): 63.5 -> 65.6Improved
shorten cycle length from n=16 (DiffPC-Long) to n=12 (DiffPC-Efficient) (ablation)DiffPC on CIFAR-10 CNN · CIFAR-10accuracy (%): 65.6 -> 63.3Hurt
shorten cycle length from n=16 to n=12 (ablation)DiffPC on CIFAR-10 CNN · CIFAR-10 (error propagation)bits per neuron: 1.9 -> 0.7Improved
replace floating-point PC messages with ternary spikes (DiffPC-L vs PC-SE)predictive coding on MNIST MLP · MNIST (error propagation)bits per neuron: 960 -> 0.18Improved
use ternary-spike DiffPC instead of standard PC (smaller model)predictive coding on MNIST MLP · MNIST (error propagation)bits per neuron: 960 -> 0.08Improved
DiffPC-L instead of standard PC-SEpredictive coding on MNIST MLP · MNISTaccuracy (%): 98.3 -> 99.3Improved
DiffPC spike-based communication versus standard PC on timestepspredictive coding on MNIST MLP · MNIST (error propagation)timesteps: 15 -> 120Hurt
DiffPC with larger cycle length n on MNIST and Fashion-MNIST (ablation)DiffPC · MNIST and Fashion-MNISTaccuracy (%)Improved
reduce decay a below 1.0 at small n (ablation)DiffPC · MNIST and Fashion-MNISTaccuracy (%)Improved

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