On the importance of time constants in spiking neural networks
Brandt, Filippa LU ; Bastani, Saeed ; Hunt, Alexander ; Aminifar, Amir LU
and Behmanesh, Baktash
LU
(2026)
p.193-198
- Abstract
- Time constants in spiking neural networks (SNNs) are crucial for determining performance. While prior work shows that learning time constants can improve accuracy, it typically assumes near-optimal initial values and rarely examines recovery from poor initializations. We systematically study how membrane and synaptic time constants affect SNN performance using multiple training strategies. Our results show that suboptimal values for time constants can reduce accuracy by nearly 10%, but networks can recover through optimization during the training process.
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/f2af06ea-55d6-4316-88d9-12b66d97fcda
- author
- Brandt, Filippa
LU
; Bastani, Saeed
; Hunt, Alexander
; Aminifar, Amir
LU
and Behmanesh, Baktash
LU
- organization
-
- Secure and Networked Systems
- NEXTG2COM – a Vinnova Competence Centre in Advanced Digitalisation
- LTH Profile Area: Water
- LU Profile Area: Natural and Artificial Cognition
- LTH Profile Area: AI and Digitalization
- LTH Profile Area: Engineering Health
- ELLIIT: the Linköping-Lund initiative on IT and mobile communication
- Integrated Electronic Systems
- NanoLund: Centre for Nanoscience
- Embedded Electronics Engineering (M.Sc.)
- Sentio: Integrated Sensors and Adaptive Technology for Sustainable Products and Manufacturing
- publishing date
- 2026-03-24
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- keywords
- Spiking neural networks, Neuromorphic computing
- host publication
- European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
- edition
- 34
- pages
- 193 - 198
- ISBN
- 978-2-87587-096-4
- DOI
- 10.14428/esann/2026.ES2026-325
- language
- English
- LU publication?
- yes
- id
- f2af06ea-55d6-4316-88d9-12b66d97fcda
- date added to LUP
- 2026-04-28 11:48:01
- date last changed
- 2026-06-01 21:38:24
@inproceedings{f2af06ea-55d6-4316-88d9-12b66d97fcda,
abstract = {{Time constants in spiking neural networks (SNNs) are crucial for determining performance. While prior work shows that learning time constants can improve accuracy, it typically assumes near-optimal initial values and rarely examines recovery from poor initializations. We systematically study how membrane and synaptic time constants affect SNN performance using multiple training strategies. Our results show that suboptimal values for time constants can reduce accuracy by nearly 10%, but networks can recover through optimization during the training process.}},
author = {{Brandt, Filippa and Bastani, Saeed and Hunt, Alexander and Aminifar, Amir and Behmanesh, Baktash}},
booktitle = {{European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning}},
isbn = {{978-2-87587-096-4}},
keywords = {{Spiking neural networks; Neuromorphic computing}},
language = {{eng}},
month = {{03}},
pages = {{193--198}},
title = {{On the importance of time constants in spiking neural networks}},
url = {{http://dx.doi.org/10.14428/esann/2026.ES2026-325}},
doi = {{10.14428/esann/2026.ES2026-325}},
year = {{2026}},
}