<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ren-etal-2023-context">
    <titleInfo>
        <title>Context Compression for Auto-regressive Transformers with Sentinel Tokens</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Siyu</namePart>
        <namePart type="family">Ren</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Qi</namePart>
        <namePart type="family">Jia</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Kenny</namePart>
        <namePart type="family">Zhu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2023-12</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Houda</namePart>
            <namePart type="family">Bouamor</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Juan</namePart>
            <namePart type="family">Pino</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Kalika</namePart>
            <namePart type="family">Bali</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Singapore</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>The quadratic complexity of the attention module makes it gradually become the bulk of compute in Transformer-based LLMs during generation. Moreover, the excessive key-value cache that arises when dealing with long inputs also brings severe issues on memory footprint and inference latency. In this work, we propose a plug-and-play approach that is able to incrementally compress the intermediate activation of a specified span of tokens into compact ones, thereby reducing both memory and computational cost when processing subsequent context. Experiments on both in-domain language modeling and zero-shot open-ended document generation demonstrate the advantage of our approach over sparse attention baselines in terms of fluency, n-gram matching, and semantic similarity. At last, we comprehensively profile the benefit of context compression on improving the system throughout. Code is available at https://github.com/DRSY/KV_Compression.</abstract>
    <identifier type="citekey">ren-etal-2023-context</identifier>
    <identifier type="doi">10.18653/v1/2023.emnlp-main.794</identifier>
    <location>
        <url>https://aclanthology.org/2023.emnlp-main.794</url>
    </location>
    <part>
        <date>2023-12</date>
        <extent unit="page">
            <start>12860</start>
            <end>12867</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Context Compression for Auto-regressive Transformers with Sentinel Tokens
%A Ren, Siyu
%A Jia, Qi
%A Zhu, Kenny
%Y Bouamor, Houda
%Y Pino, Juan
%Y Bali, Kalika
%S Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
%D 2023
%8 December
%I Association for Computational Linguistics
%C Singapore
%F ren-etal-2023-context
%X The quadratic complexity of the attention module makes it gradually become the bulk of compute in Transformer-based LLMs during generation. Moreover, the excessive key-value cache that arises when dealing with long inputs also brings severe issues on memory footprint and inference latency. In this work, we propose a plug-and-play approach that is able to incrementally compress the intermediate activation of a specified span of tokens into compact ones, thereby reducing both memory and computational cost when processing subsequent context. Experiments on both in-domain language modeling and zero-shot open-ended document generation demonstrate the advantage of our approach over sparse attention baselines in terms of fluency, n-gram matching, and semantic similarity. At last, we comprehensively profile the benefit of context compression on improving the system throughout. Code is available at https://github.com/DRSY/KV_Compression.
%R 10.18653/v1/2023.emnlp-main.794
%U https://aclanthology.org/2023.emnlp-main.794
%U https://doi.org/10.18653/v1/2023.emnlp-main.794
%P 12860-12867

Markdown (Informal)

[Context Compression for Auto-regressive Transformers with Sentinel Tokens](https://aclanthology.org/2023.emnlp-main.794) (Ren et al., EMNLP 2023)

• Context Compression for Auto-regressive Transformers with Sentinel Tokens (Ren et al., EMNLP 2023)

ACL

• Siyu Ren, Qi Jia, and Kenny Zhu. 2023. Context Compression for Auto-regressive Transformers with Sentinel Tokens. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 12860–12867, Singapore. Association for Computational Linguistics.