Parallel Training in Spiking Neural Networks Journal: Neural Networks

Yanbin Huang, Man Yao, Yuqi Pan, Changze Lv, Siyuan Xu, Xiaoqing Zheng, Bo Xu, Guoqi Li · 2026

Read Parallel Training in Spiking Neural Networks Journal: Neural Networks by Yanbin Huang, Man Yao, Yuqi Pan, Changze Lv, Siyuan Xu, Xiaoqing Zheng, Bo Xu, Guoqi Li (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-01133.txt.

Description

Yanbin Huang and co-authors rethink spike nets for parallel training, dropping the reset step. Reset ties time steps in knots that block parallel runs, yet any swap must keep reset's jobs and keep serial runs cheap at test time. Their decay spike cell meets both bars. On 16k-long runs it trained 25.6x faster than the first parallel cell, and nets trained on 2k runs held firm to 30k. It held up across five job kinds, three net forms, and two spike modes, firing less than old cells.

Catalog record

CREATORYanbin Huang, Man Yao, Yuqi Pan, Changze Lv, Siyuan Xu, Xiaoqing Zheng, Bo Xu, Guoqi Li
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE76827
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602.01133

Browse the full library · About the Tower of Babel