Globally Optimal Training of Spiking Neural Networks via Parameter Reconstruction
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911663139061760 |
|---|---|
| author | Udupi, Himanshu Yang, Xiaocong Zhai, ChengXiang |
| author_facet | Udupi, Himanshu Yang, Xiaocong Zhai, ChengXiang |
| contents | Spiking Neural Networks (SNNs) have been proposed as biologically plausible and energy-efficient alternatives to conventional Artificial Neural Networks (ANNs). However, the training of SNN usually relies on surrogate gradients due to the non-differentiability of the spike function, introducing approximation errors that accumulate across layers. To address this challenge, we extend the work on convexification of parallel feedforward threshold networks to parallel recurrent threshold networks, which subsume parallel SNNs as a structured special case. Building on this theoretical framework, we propose a parameter reconstruction algorithm for SNN training that demonstrates consistent and significant advantages across various tasks, both as a standalone method and in combination with surrogate-gradient training. The ablations further demonstrate the data scalability and robustness to model configurations of our training algorithm, pointing toward its potential in large-scale SNN training. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_08022 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Globally Optimal Training of Spiking Neural Networks via Parameter Reconstruction Udupi, Himanshu Yang, Xiaocong Zhai, ChengXiang Neural and Evolutionary Computing Artificial Intelligence Machine Learning Spiking Neural Networks (SNNs) have been proposed as biologically plausible and energy-efficient alternatives to conventional Artificial Neural Networks (ANNs). However, the training of SNN usually relies on surrogate gradients due to the non-differentiability of the spike function, introducing approximation errors that accumulate across layers. To address this challenge, we extend the work on convexification of parallel feedforward threshold networks to parallel recurrent threshold networks, which subsume parallel SNNs as a structured special case. Building on this theoretical framework, we propose a parameter reconstruction algorithm for SNN training that demonstrates consistent and significant advantages across various tasks, both as a standalone method and in combination with surrogate-gradient training. The ablations further demonstrate the data scalability and robustness to model configurations of our training algorithm, pointing toward its potential in large-scale SNN training. |
| title | Globally Optimal Training of Spiking Neural Networks via Parameter Reconstruction |
| topic | Neural and Evolutionary Computing Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2605.08022 |