@inproceedings{wang-etal-2023-aprompt,
title = "{AP}rompt: Attention Prompt Tuning for Efficient Adaptation of Pre-trained Language Models",
author = "Wang, Qifan and
Mao, Yuning and
Wang, Jingang and
Yu, Hanchao and
Nie, Shaoliang and
Wang, Sinong and
Feng, Fuli and
Huang, Lifu and
Quan, Xiaojun and
Xu, Zenglin and
Liu, Dongfang",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.567",
doi = "10.18653/v1/2023.emnlp-main.567",
pages = "9147--9160",
abstract = "With the continuous growth of large language models, the process of fine-tuning these models for new tasks has become increasingly parameter-intensive. Prompt tuning, a method that involves tuning a small set of soft prompts, has emerged as an effective and efficient approach for adapting large pre-trained language models. However, most existing prompt tuning approaches only introduce prompts at the input layer, limiting their performance and leaving large rooms for improvement. In this work, we propose a novel Attention Prompt tuning method, namely APrompt, for efficient adaptation of pre-trained language models. We first demonstrate that existing prompt tuning can be considered as a special case of attention prompt tuning. We then formally introduce APrompt, which incorporates query, key, and value prompts into the attention layer to guide the attention computation during fine-tuning. Experimental results on the SuperGLUE benchmark consistently demonstrate that our proposed approach outperforms state-of-the-art baselines and full fine-tuning method with pre-trained models at different scales. In addition, a comprehensive set of ablation studies validate the effectiveness of the prompt design, as well as the efficiency of our approach.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2023-aprompt">
<titleInfo>
<title>APrompt: Attention Prompt Tuning for Efficient Adaptation of Pre-trained Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Qifan</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuning</namePart>
<namePart type="family">Mao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jingang</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hanchao</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shaoliang</namePart>
<namePart type="family">Nie</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sinong</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fuli</namePart>
<namePart type="family">Feng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lifu</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaojun</namePart>
<namePart type="family">Quan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zenglin</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dongfang</namePart>
<namePart type="family">Liu</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>With the continuous growth of large language models, the process of fine-tuning these models for new tasks has become increasingly parameter-intensive. Prompt tuning, a method that involves tuning a small set of soft prompts, has emerged as an effective and efficient approach for adapting large pre-trained language models. However, most existing prompt tuning approaches only introduce prompts at the input layer, limiting their performance and leaving large rooms for improvement. In this work, we propose a novel Attention Prompt tuning method, namely APrompt, for efficient adaptation of pre-trained language models. We first demonstrate that existing prompt tuning can be considered as a special case of attention prompt tuning. We then formally introduce APrompt, which incorporates query, key, and value prompts into the attention layer to guide the attention computation during fine-tuning. Experimental results on the SuperGLUE benchmark consistently demonstrate that our proposed approach outperforms state-of-the-art baselines and full fine-tuning method with pre-trained models at different scales. In addition, a comprehensive set of ablation studies validate the effectiveness of the prompt design, as well as the efficiency of our approach.</abstract>
<identifier type="citekey">wang-etal-2023-aprompt</identifier>
<identifier type="doi">10.18653/v1/2023.emnlp-main.567</identifier>
<location>
<url>https://aclanthology.org/2023.emnlp-main.567</url>
</location>
<part>
<date>2023-12</date>
<extent unit="page">
<start>9147</start>
<end>9160</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T APrompt: Attention Prompt Tuning for Efficient Adaptation of Pre-trained Language Models
%A Wang, Qifan
%A Mao, Yuning
%A Wang, Jingang
%A Yu, Hanchao
%A Nie, Shaoliang
%A Wang, Sinong
%A Feng, Fuli
%A Huang, Lifu
%A Quan, Xiaojun
%A Xu, Zenglin
%A Liu, Dongfang
%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 wang-etal-2023-aprompt
%X With the continuous growth of large language models, the process of fine-tuning these models for new tasks has become increasingly parameter-intensive. Prompt tuning, a method that involves tuning a small set of soft prompts, has emerged as an effective and efficient approach for adapting large pre-trained language models. However, most existing prompt tuning approaches only introduce prompts at the input layer, limiting their performance and leaving large rooms for improvement. In this work, we propose a novel Attention Prompt tuning method, namely APrompt, for efficient adaptation of pre-trained language models. We first demonstrate that existing prompt tuning can be considered as a special case of attention prompt tuning. We then formally introduce APrompt, which incorporates query, key, and value prompts into the attention layer to guide the attention computation during fine-tuning. Experimental results on the SuperGLUE benchmark consistently demonstrate that our proposed approach outperforms state-of-the-art baselines and full fine-tuning method with pre-trained models at different scales. In addition, a comprehensive set of ablation studies validate the effectiveness of the prompt design, as well as the efficiency of our approach.
%R 10.18653/v1/2023.emnlp-main.567
%U https://aclanthology.org/2023.emnlp-main.567
%U https://doi.org/10.18653/v1/2023.emnlp-main.567
%P 9147-9160
Markdown (Informal)
[APrompt: Attention Prompt Tuning for Efficient Adaptation of Pre-trained Language Models](https://aclanthology.org/2023.emnlp-main.567) (Wang et al., EMNLP 2023)
ACL
- Qifan Wang, Yuning Mao, Jingang Wang, Hanchao Yu, Shaoliang Nie, Sinong Wang, Fuli Feng, Lifu Huang, Xiaojun Quan, Zenglin Xu, and Dongfang Liu. 2023. APrompt: Attention Prompt Tuning for Efficient Adaptation of Pre-trained Language Models. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 9147–9160, Singapore. Association for Computational Linguistics.