@inproceedings{he-etal-2025-densessm,
title = "{D}ense{SSM}: State Space Models with Dense Hidden Connection for Efficient Large Language Models",
author = "He, Wei and
Han, Kai and
Tang, Yehui and
Wang, Chengcheng and
Yang, Yujie and
Guo, Tianyu and
Wang, Yunhe",
editor = "Chiruzzo, Luis and
Ritter, Alan and
Wang, Lu",
booktitle = "Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)",
month = apr,
year = "2025",
address = "Albuquerque, New Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.naacl-long.467/",
doi = "10.18653/v1/2025.naacl-long.467",
pages = "9243--9254",
ISBN = "979-8-89176-189-6",
abstract = "Large language models (LLMs) face a significant challenge due to the excessive computational and memory requirements of the commonly used Transformer architecture. While state space model (SSM) is a new type of foundational network architecture offering lower computational complexity, their performance has yet to fully rival that of Transformers. This paper introduces DenseSSM, a novel approach to enhance the flow of hidden information between layers in SSMs. By selectively integrating shallow-layer hidden states into deeper layers, DenseSSM retains fine-grained information crucial for the final output. This incremental improvement maintains the training parallelizability and inference efficiency of SSMs while significantly boosting performance. The proposed method is broadly applicable to various SSM types, including RetNet and Mamba, and DenseSSM achieves significant performance improvements on public benchmarks, demonstrating its effectiveness and versatility."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="he-etal-2025-densessm">
<titleInfo>
<title>DenseSSM: State Space Models with Dense Hidden Connection for Efficient Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">He</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kai</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yehui</namePart>
<namePart type="family">Tang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chengcheng</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yujie</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianyu</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yunhe</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-04</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Luis</namePart>
<namePart type="family">Chiruzzo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="family">Ritter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Albuquerque, New Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-189-6</identifier>
</relatedItem>
<abstract>Large language models (LLMs) face a significant challenge due to the excessive computational and memory requirements of the commonly used Transformer architecture. While state space model (SSM) is a new type of foundational network architecture offering lower computational complexity, their performance has yet to fully rival that of Transformers. This paper introduces DenseSSM, a novel approach to enhance the flow of hidden information between layers in SSMs. By selectively integrating shallow-layer hidden states into deeper layers, DenseSSM retains fine-grained information crucial for the final output. This incremental improvement maintains the training parallelizability and inference efficiency of SSMs while significantly boosting performance. The proposed method is broadly applicable to various SSM types, including RetNet and Mamba, and DenseSSM achieves significant performance improvements on public benchmarks, demonstrating its effectiveness and versatility.</abstract>
<identifier type="citekey">he-etal-2025-densessm</identifier>
<identifier type="doi">10.18653/v1/2025.naacl-long.467</identifier>
<location>
<url>https://aclanthology.org/2025.naacl-long.467/</url>
</location>
<part>
<date>2025-04</date>
<extent unit="page">
<start>9243</start>
<end>9254</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T DenseSSM: State Space Models with Dense Hidden Connection for Efficient Large Language Models
%A He, Wei
%A Han, Kai
%A Tang, Yehui
%A Wang, Chengcheng
%A Yang, Yujie
%A Guo, Tianyu
%A Wang, Yunhe
%Y Chiruzzo, Luis
%Y Ritter, Alan
%Y Wang, Lu
%S Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)
%D 2025
%8 April
%I Association for Computational Linguistics
%C Albuquerque, New Mexico
%@ 979-8-89176-189-6
%F he-etal-2025-densessm
%X Large language models (LLMs) face a significant challenge due to the excessive computational and memory requirements of the commonly used Transformer architecture. While state space model (SSM) is a new type of foundational network architecture offering lower computational complexity, their performance has yet to fully rival that of Transformers. This paper introduces DenseSSM, a novel approach to enhance the flow of hidden information between layers in SSMs. By selectively integrating shallow-layer hidden states into deeper layers, DenseSSM retains fine-grained information crucial for the final output. This incremental improvement maintains the training parallelizability and inference efficiency of SSMs while significantly boosting performance. The proposed method is broadly applicable to various SSM types, including RetNet and Mamba, and DenseSSM achieves significant performance improvements on public benchmarks, demonstrating its effectiveness and versatility.
%R 10.18653/v1/2025.naacl-long.467
%U https://aclanthology.org/2025.naacl-long.467/
%U https://doi.org/10.18653/v1/2025.naacl-long.467
%P 9243-9254
Markdown (Informal)
[DenseSSM: State Space Models with Dense Hidden Connection for Efficient Large Language Models](https://aclanthology.org/2025.naacl-long.467/) (He et al., NAACL 2025)
ACL