@inproceedings{fei-etal-2024-extending,
title = "Extending Context Window of Large Language Models via Semantic Compression",
author = "Fei, Weizhi and
Niu, Xueyan and
Zhou, Pingyi and
Hou, Lu and
Bai, Bo and
Deng, Lei and
Han, Wei",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Findings of the Association for Computational Linguistics ACL 2024",
month = aug,
year = "2024",
address = "Bangkok, Thailand and virtual meeting",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-acl.306",
pages = "5169--5181",
abstract = "Transformer based Large Language Models (LLMs) often impose limitations on the length of the text input to ensure the generation of fluent and relevant responses due to the quadratic complexity. These constraints restrict their applicability in long text scenarios. In this paper, we propose a novel semantic compression method that enables generalization to texts that are 6-8 times longer without incurring significant computational costs or requiring fine-tuning. Our proposed framework draws inspiration from source coding in information theory and employs a pre-trained model to reduce the semantic redundancy of long inputs before passing them to the LLMs for downstream tasks. Experimental results demonstrate that our method effectively extends the context window of LLMs across a range of tasks including question answering, summarization, few-shot learning, and information retrieval. Furthermore, the proposed semantic compression method exhibits consistent fluency in text generation while reducing the associated computational overhead.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="fei-etal-2024-extending">
<titleInfo>
<title>Extending Context Window of Large Language Models via Semantic Compression</title>
</titleInfo>
<name type="personal">
<namePart type="given">Weizhi</namePart>
<namePart type="family">Fei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xueyan</namePart>
<namePart type="family">Niu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pingyi</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Hou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bo</namePart>
<namePart type="family">Bai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lei</namePart>
<namePart type="family">Deng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics ACL 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Lun-Wei</namePart>
<namePart type="family">Ku</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Andre</namePart>
<namePart type="family">Martins</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivek</namePart>
<namePart type="family">Srikumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Bangkok, Thailand and virtual meeting</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Transformer based Large Language Models (LLMs) often impose limitations on the length of the text input to ensure the generation of fluent and relevant responses due to the quadratic complexity. These constraints restrict their applicability in long text scenarios. In this paper, we propose a novel semantic compression method that enables generalization to texts that are 6-8 times longer without incurring significant computational costs or requiring fine-tuning. Our proposed framework draws inspiration from source coding in information theory and employs a pre-trained model to reduce the semantic redundancy of long inputs before passing them to the LLMs for downstream tasks. Experimental results demonstrate that our method effectively extends the context window of LLMs across a range of tasks including question answering, summarization, few-shot learning, and information retrieval. Furthermore, the proposed semantic compression method exhibits consistent fluency in text generation while reducing the associated computational overhead.</abstract>
<identifier type="citekey">fei-etal-2024-extending</identifier>
<location>
<url>https://aclanthology.org/2024.findings-acl.306</url>
</location>
<part>
<date>2024-08</date>
<extent unit="page">
<start>5169</start>
<end>5181</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Extending Context Window of Large Language Models via Semantic Compression
%A Fei, Weizhi
%A Niu, Xueyan
%A Zhou, Pingyi
%A Hou, Lu
%A Bai, Bo
%A Deng, Lei
%A Han, Wei
%Y Ku, Lun-Wei
%Y Martins, Andre
%Y Srikumar, Vivek
%S Findings of the Association for Computational Linguistics ACL 2024
%D 2024
%8 August
%I Association for Computational Linguistics
%C Bangkok, Thailand and virtual meeting
%F fei-etal-2024-extending
%X Transformer based Large Language Models (LLMs) often impose limitations on the length of the text input to ensure the generation of fluent and relevant responses due to the quadratic complexity. These constraints restrict their applicability in long text scenarios. In this paper, we propose a novel semantic compression method that enables generalization to texts that are 6-8 times longer without incurring significant computational costs or requiring fine-tuning. Our proposed framework draws inspiration from source coding in information theory and employs a pre-trained model to reduce the semantic redundancy of long inputs before passing them to the LLMs for downstream tasks. Experimental results demonstrate that our method effectively extends the context window of LLMs across a range of tasks including question answering, summarization, few-shot learning, and information retrieval. Furthermore, the proposed semantic compression method exhibits consistent fluency in text generation while reducing the associated computational overhead.
%U https://aclanthology.org/2024.findings-acl.306
%P 5169-5181
Markdown (Informal)
[Extending Context Window of Large Language Models via Semantic Compression](https://aclanthology.org/2024.findings-acl.306) (Fei et al., Findings 2024)
ACL