Saved in:
Bibliographic Details
Main Authors: Wang, Jingya, Deng, Xin, Wei, Wenjie, Zhang, Dehao, Wang, Shuai, Sun, Qian, Zhang, Jieyuan, Liu, Hanwen, Xie, Ning, Zhang, Malu
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.07710
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908707829317632
author Wang, Jingya
Deng, Xin
Wei, Wenjie
Zhang, Dehao
Wang, Shuai
Sun, Qian
Zhang, Jieyuan
Liu, Hanwen
Xie, Ning
Zhang, Malu
author_facet Wang, Jingya
Deng, Xin
Wei, Wenjie
Zhang, Dehao
Wang, Shuai
Sun, Qian
Zhang, Jieyuan
Liu, Hanwen
Xie, Ning
Zhang, Malu
contents Leveraging the event-driven paradigm, Spiking Neural Networks (SNNs) offer a promising approach for energy-efficient Transformer architectures.While ANN-to-SNN conversion avoids the high training cost of directly trained Spiking Transformers, existing approaches still struggle to handle the nonlinear operations within Transformer blocks, and often require additional fine-tuning of pretrained ANNs.To address these limitations, we propose a training-free and high-performance ANN-to-SNN conversion framework tailored for Transformer architectures. Specifically, we introduce a Multi-basis Exponential Decay (MBE) neuron that combines exponential decay with a multi-basis encoding strategy to effectively approximate nonlinear operations, eliminating the need for weight modifications in pretrained ANNs.Extensive experiments across diverse tasks (CV, NLU, NLG) and mainstream Transformer architectures (ViT, RoBERTa, GPT-2) demonstrate that our method achieves near-lossless conversion accuracy with significantly lower latency. This provides a promising pathway for the efficient and scalable deployment of Spiking Transformers in real-world applications.
format Preprint
id arxiv_https___arxiv_org_abs_2508_07710
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Training-Free ANN-to-SNN Conversion for High-Performance Spiking Transformer
Wang, Jingya
Deng, Xin
Wei, Wenjie
Zhang, Dehao
Wang, Shuai
Sun, Qian
Zhang, Jieyuan
Liu, Hanwen
Xie, Ning
Zhang, Malu
Machine Learning
Artificial Intelligence
Leveraging the event-driven paradigm, Spiking Neural Networks (SNNs) offer a promising approach for energy-efficient Transformer architectures.While ANN-to-SNN conversion avoids the high training cost of directly trained Spiking Transformers, existing approaches still struggle to handle the nonlinear operations within Transformer blocks, and often require additional fine-tuning of pretrained ANNs.To address these limitations, we propose a training-free and high-performance ANN-to-SNN conversion framework tailored for Transformer architectures. Specifically, we introduce a Multi-basis Exponential Decay (MBE) neuron that combines exponential decay with a multi-basis encoding strategy to effectively approximate nonlinear operations, eliminating the need for weight modifications in pretrained ANNs.Extensive experiments across diverse tasks (CV, NLU, NLG) and mainstream Transformer architectures (ViT, RoBERTa, GPT-2) demonstrate that our method achieves near-lossless conversion accuracy with significantly lower latency. This provides a promising pathway for the efficient and scalable deployment of Spiking Transformers in real-world applications.
title Training-Free ANN-to-SNN Conversion for High-Performance Spiking Transformer
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2508.07710