@inproceedings{luo-etal-2024-taking,
title = "Taking a Deep Breath: Enhancing Language Modeling of Large Language Models with Sentinel Tokens",
author = "Luo, Weiyao and
Zheng, Suncong and
Xia, Heming and
Wang, Weikang and
Lei, Yan and
Liu, Tianyu and
Chen, Shuang and
Sui, Zhifang",
editor = "Al-Onaizan, Yaser and
Bansal, Mohit and
Chen, Yun-Nung",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-emnlp.233/",
doi = "10.18653/v1/2024.findings-emnlp.233",
pages = "4034--4040",
abstract = "Large language models (LLMs) have shown promising efficacy across various tasks, becoming powerful tools in numerous aspects of human life. However, Transformer-based LLMs suffer a performance degradation when modeling long-term contexts due to they discard some information to reduce computational overhead. In this work, we propose a simple yet effective method to enable LLMs to take a deep breath, encouraging them to summarize information contained within discrete text chunks. Specifically, we segment the text into multiple chunks and insert special token {\ensuremath{<}}SR{\ensuremath{>}} at the end of each chunk. We then modify the attention mask to integrate the chunk{'}s information into the corresponding {\ensuremath{<}}SR{\ensuremath{>}} token. This facilitates LLMs to interpret information not only from historical individual tokens but also from the {\ensuremath{<}}SR{\ensuremath{>}} token, aggregating the chunk{'}s semantic information. Experiments on language modeling and out-of-domain downstream tasks validate the superiority of our approach."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="luo-etal-2024-taking">
<titleInfo>
<title>Taking a Deep Breath: Enhancing Language Modeling of Large Language Models with Sentinel Tokens</title>
</titleInfo>
<name type="personal">
<namePart type="given">Weiyao</namePart>
<namePart type="family">Luo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Suncong</namePart>
<namePart type="family">Zheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Heming</namePart>
<namePart type="family">Xia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Weikang</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yan</namePart>
<namePart type="family">Lei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianyu</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shuang</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhifang</namePart>
<namePart type="family">Sui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yaser</namePart>
<namePart type="family">Al-Onaizan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohit</namePart>
<namePart type="family">Bansal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yun-Nung</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Miami, Florida, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Large language models (LLMs) have shown promising efficacy across various tasks, becoming powerful tools in numerous aspects of human life. However, Transformer-based LLMs suffer a performance degradation when modeling long-term contexts due to they discard some information to reduce computational overhead. In this work, we propose a simple yet effective method to enable LLMs to take a deep breath, encouraging them to summarize information contained within discrete text chunks. Specifically, we segment the text into multiple chunks and insert special token \ensuremath<SR\ensuremath> at the end of each chunk. We then modify the attention mask to integrate the chunk’s information into the corresponding \ensuremath<SR\ensuremath> token. This facilitates LLMs to interpret information not only from historical individual tokens but also from the \ensuremath<SR\ensuremath> token, aggregating the chunk’s semantic information. Experiments on language modeling and out-of-domain downstream tasks validate the superiority of our approach.</abstract>
<identifier type="citekey">luo-etal-2024-taking</identifier>
<identifier type="doi">10.18653/v1/2024.findings-emnlp.233</identifier>
<location>
<url>https://aclanthology.org/2024.findings-emnlp.233/</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>4034</start>
<end>4040</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Taking a Deep Breath: Enhancing Language Modeling of Large Language Models with Sentinel Tokens
%A Luo, Weiyao
%A Zheng, Suncong
%A Xia, Heming
%A Wang, Weikang
%A Lei, Yan
%A Liu, Tianyu
%A Chen, Shuang
%A Sui, Zhifang
%Y Al-Onaizan, Yaser
%Y Bansal, Mohit
%Y Chen, Yun-Nung
%S Findings of the Association for Computational Linguistics: EMNLP 2024
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F luo-etal-2024-taking
%X Large language models (LLMs) have shown promising efficacy across various tasks, becoming powerful tools in numerous aspects of human life. However, Transformer-based LLMs suffer a performance degradation when modeling long-term contexts due to they discard some information to reduce computational overhead. In this work, we propose a simple yet effective method to enable LLMs to take a deep breath, encouraging them to summarize information contained within discrete text chunks. Specifically, we segment the text into multiple chunks and insert special token \ensuremath<SR\ensuremath> at the end of each chunk. We then modify the attention mask to integrate the chunk’s information into the corresponding \ensuremath<SR\ensuremath> token. This facilitates LLMs to interpret information not only from historical individual tokens but also from the \ensuremath<SR\ensuremath> token, aggregating the chunk’s semantic information. Experiments on language modeling and out-of-domain downstream tasks validate the superiority of our approach.
%R 10.18653/v1/2024.findings-emnlp.233
%U https://aclanthology.org/2024.findings-emnlp.233/
%U https://doi.org/10.18653/v1/2024.findings-emnlp.233
%P 4034-4040
Markdown (Informal)
[Taking a Deep Breath: Enhancing Language Modeling of Large Language Models with Sentinel Tokens](https://aclanthology.org/2024.findings-emnlp.233/) (Luo et al., Findings 2024)
ACL
- Weiyao Luo, Suncong Zheng, Heming Xia, Weikang Wang, Yan Lei, Tianyu Liu, Shuang Chen, and Zhifang Sui. 2024. Taking a Deep Breath: Enhancing Language Modeling of Large Language Models with Sentinel Tokens. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 4034–4040, Miami, Florida, USA. Association for Computational Linguistics.