Predictive Suppression Layers for Communication-Efficient Spiking Neural Networks
In the authors' words
Feedforward Spiking Neural Networks (SNNs) typically propagate every generated spike indiscriminately, disregarding whether the information is redundant from an information-theoretic perspective. This lack of selectivity induces high redundancy in inter-layer communication, creating an expensive overhead, e.g., in scenarios involving many-core neuromorphic hardware or communication-dominated Internet-of-Things (IoT) where features are transmitted wirelessly. To address this challenge, we trade localized processing for leaner network channels by introducing a minimal predictive coding framework for SNNs. We propose two layer variants sharing a predictor block: error units, which transmit signed spiking residuals, and predictive suppression, which uses residual magnitude to dynamically gate and forward only unpredictable, "surprising" activity. Evaluated on the N-MNIST and Spiking Heidelberg Digits (SHD) datasets using diagnostic metrics that decouple local processing from cross-layer communication, our new predictive coding layers achieve significant communication savings. Numerical results reveal a three-fold reduction in communicated activity, while increasing the task accuracy for both datasets. The latter finding is notable, and suggests that predictive coding layers not only minimize communication overhead, but also produce output feature vectors with a higher representation power.
Appeared: Thursday, September 24. arXiv. Preprint with a published version.
DOI: 10.3217/za2v-bn54
Published version: Proceedings of the 2026 International Conference on Embedded Wireless Systems and Networks (EWSN '26), pp. 309-314, 2026
Authors' comment: 6 pages, 5 figures. Accepted at the 1st Neuromorphic Physical Layer Signal Processing for Wireless Systems Workshop (NeuroPHY 2026), co-located with EWSN 2026