A Method on Searching Better Activation Functions

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Sun, Haoyuan, Wu, Zihao, Xia, Bo, Chang, Pu, Dong, Zibin, Yuan, Yifu, Chang, Yongzhe, Wang, Xueqian
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916257183301632
author Sun, Haoyuan
Wu, Zihao
Xia, Bo
Chang, Pu
Dong, Zibin
Yuan, Yifu
Chang, Yongzhe
Wang, Xueqian
author_facet Sun, Haoyuan
Wu, Zihao
Xia, Bo
Chang, Pu
Dong, Zibin
Yuan, Yifu
Chang, Yongzhe
Wang, Xueqian
contents The success of artificial neural networks (ANNs) hinges greatly on the judicious selection of an activation function, introducing non-linearity into network and enabling them to model sophisticated relationships in data. However, the search of activation functions has largely relied on empirical knowledge in the past, lacking theoretical guidance, which has hindered the identification of more effective activation functions. In this work, we offer a proper solution to such issue. Firstly, we theoretically demonstrate the existence of the worst activation function with boundary conditions (WAFBC) from the perspective of information entropy. Furthermore, inspired by the Taylor expansion form of information entropy functional, we propose the Entropy-based Activation Function Optimization (EAFO) methodology. EAFO methodology presents a novel perspective for designing static activation functions in deep neural networks and the potential of dynamically optimizing activation during iterative training. Utilizing EAFO methodology, we derive a novel activation function from ReLU, known as Correction Regularized ReLU (CRReLU). Experiments conducted with vision transformer and its variants on CIFAR-10, CIFAR-100 and ImageNet-1K datasets demonstrate the superiority of CRReLU over existing corrections of ReLU. Extensive empirical studies on task of large language model (LLM) fine-tuning, CRReLU exhibits superior performance compared to GELU, suggesting its broader potential for practical applications.
format Preprint
id arxiv_https___arxiv_org_abs_2405_12954
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Method on Searching Better Activation Functions
Sun, Haoyuan
Wu, Zihao
Xia, Bo
Chang, Pu
Dong, Zibin
Yuan, Yifu
Chang, Yongzhe
Wang, Xueqian
Machine Learning
Artificial Intelligence
The success of artificial neural networks (ANNs) hinges greatly on the judicious selection of an activation function, introducing non-linearity into network and enabling them to model sophisticated relationships in data. However, the search of activation functions has largely relied on empirical knowledge in the past, lacking theoretical guidance, which has hindered the identification of more effective activation functions. In this work, we offer a proper solution to such issue. Firstly, we theoretically demonstrate the existence of the worst activation function with boundary conditions (WAFBC) from the perspective of information entropy. Furthermore, inspired by the Taylor expansion form of information entropy functional, we propose the Entropy-based Activation Function Optimization (EAFO) methodology. EAFO methodology presents a novel perspective for designing static activation functions in deep neural networks and the potential of dynamically optimizing activation during iterative training. Utilizing EAFO methodology, we derive a novel activation function from ReLU, known as Correction Regularized ReLU (CRReLU). Experiments conducted with vision transformer and its variants on CIFAR-10, CIFAR-100 and ImageNet-1K datasets demonstrate the superiority of CRReLU over existing corrections of ReLU. Extensive empirical studies on task of large language model (LLM) fine-tuning, CRReLU exhibits superior performance compared to GELU, suggesting its broader potential for practical applications.
title A Method on Searching Better Activation Functions
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2405.12954