arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2404.09516cs.LGcs.AIcs.CLcs.CVcs.MM
收藏

作为Transformer替代方案的新一代网络状态空间模型:综述

State Space Model for New-Generation Network Alternative to Transformers: A Survey

School of Computer Science and Technology, Anhui University(安徽大学计算机科学与技术学院);School of Artificial Intelligence, Anhui University(安徽大学人工智能学院);Peng Cheng Laboratory(鹏城实验室);Harbin Institute of Technology (HITSZ)(哈尔滨工业大学(深圳));School of Computer Science, Peking University(北京大学计算机学院);School of Electronic and Computer Engineering, Shenzhen Graduate School, Peking University(北京大学深圳研究生院电子与计算机工程学院)
Xiao Wang, Shiao Wang, Yuhe Ding, Yuehang Li, Wentao Wu, Yao Rong, Weizhe Kong, Ju Huang, Shihao Li, Haoxiang Yang, Ziwen Wang, Bo Jiang, Chenglong Li, Yaowei W… 展开作者

Xiao Wang, Shiao Wang, Yuhe Ding, Yuehang Li, Wentao Wu, Yao Rong, Weizhe Kong, Ju Huang, Shihao Li, Haoxiang Yang, Ziwen Wang, Bo Jiang, Chenglong Li, Yaowei Wang, Yonghong Tian, Jin Tang

AI总结:

该综述首次全面梳理作为Transformer替代方案的状态空间模型(SSM),阐述其原理、多领域应用,开展实验与统计对比分析,并提出未来研究方向,助力SSM理论与应用发展。

摘要:

在后深度学习时代,Transformer架构已在预训练大模型及各类下游任务中展现出强大性能。然而,该架构的巨大计算需求令许多研究者望而却步。为进一步降低注意力模型的复杂度,学界已开展大量工作设计更高效的方法。其中,状态空间模型(State Space Model, SSM)作为基于自注意力的Transformer模型的潜在替代方案,近年来受到越来越多的关注。本文首次对相关工作进行全面综述,并提供实验对比与分析以更好地展现SSM的特性与优势。具体而言,我们首先详细阐述其原理,帮助读者快速掌握SSM的核心思想;随后深入梳理现有SSM及其各类应用,涵盖自然语言处理、计算机视觉、图、多模态与多媒体、点云/事件流、时序数据等多个领域。此外,我们对这些模型进行了统计对比与分析,希望帮助读者理解不同结构在各类任务上的有效性。最后,我们提出了该方向可能的研究点,以更好地推动SSM的理论模型与应用发展。更多相关工作将持续更新在以下GitHub仓库:https://github.com/Event-AHU/Mamba_State_Space_Model_Paper_List。

英文摘要:

In the post-deep learning era, the Transformer architecture has demonstrated its powerful performance across pre-trained big models and various downstream tasks. However, the enormous computational demands of this architecture have deterred many researchers. To further reduce the complexity of attention models, numerous efforts have been made to design more efficient methods. Among them, the State Space Model (SSM), as a possible replacement for the self-attention based Transformer model, has drawn more and more attention in recent years. In this paper, we give the first comprehensive review of these works and also provide experimental comparisons and analysis to better demonstrate the features and advantages of SSM. Specifically, we first give a detailed description of principles to help the readers quickly capture the key ideas of SSM. After that, we dive into the reviews of existing SSMs and their various applications, including natural language processing, computer vision, graph, multi-modal and multi-media, point cloud/event stream, time series data, and other domains. In addition, we give statistical comparisons and analysis of these models and hope it helps the readers to understand the effectiveness of different structures on various tasks. Then, we propose possible research points in this direction to better promote the development of the theoretical model and application of SSM. More related works will be continuously updated on the following GitHub: https://github.com/Event-AHU/Mamba_State_Space_Model_Paper_List.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑