@inproceedings{weng-etal-2023-g,
title = "{G}-Tuning: Improving Generalization of Pre-trained Language Models with Generative Adversarial Network",
author = "Weng, Rongxiang and
Cheng, Wen Sen and
Zhang, Min",
editor = "Rogers, Anna and
Boyd-Graber, Jordan and
Okazaki, Naoaki",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2023",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.findings-acl.291/",
doi = "10.18653/v1/2023.findings-acl.291",
pages = "4747--4755",
abstract = "The generalization ability of pre-trained language models (Plms) in downstream tasks is heavily influenced by fine-tuning. The objective of fine-tuning is to transform the latent representation of Plms from a universal space to a target space, allowing the model to be applied to downstream tasks with the capability of generalizing to unseen samples. However, the effect of Plms will be diminished when the training data coverage is insufficient, in which fine-tuning is inadequate to learn the complete mapping. In this study, we propose a new fine-tuning framework, referred to as G-Tuning, that aims to preserve the generalization ability of Plms in downstream tasks. Specifically, we integrate a generative adversarial network into the fine-tuning process to aid in the transformation of the latent representation in the entire space. Empirical evaluations on the GLUE benchmark, as well as two additional demanding scenarios involving domain and language generalization, demonstrate that G-Tuning can accurately map the universal representation to the target space, thus effectively enhancing the generalization performance of Plms across various downstream tasks."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="weng-etal-2023-g">
<titleInfo>
<title>G-Tuning: Improving Generalization of Pre-trained Language Models with Generative Adversarial Network</title>
</titleInfo>
<name type="personal">
<namePart type="given">Rongxiang</namePart>
<namePart type="family">Weng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wen</namePart>
<namePart type="given">Sen</namePart>
<namePart type="family">Cheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Min</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2023</title>
</titleInfo>
<name type="personal">
<namePart type="given">Anna</namePart>
<namePart type="family">Rogers</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jordan</namePart>
<namePart type="family">Boyd-Graber</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Naoaki</namePart>
<namePart type="family">Okazaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Toronto, Canada</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>The generalization ability of pre-trained language models (Plms) in downstream tasks is heavily influenced by fine-tuning. The objective of fine-tuning is to transform the latent representation of Plms from a universal space to a target space, allowing the model to be applied to downstream tasks with the capability of generalizing to unseen samples. However, the effect of Plms will be diminished when the training data coverage is insufficient, in which fine-tuning is inadequate to learn the complete mapping. In this study, we propose a new fine-tuning framework, referred to as G-Tuning, that aims to preserve the generalization ability of Plms in downstream tasks. Specifically, we integrate a generative adversarial network into the fine-tuning process to aid in the transformation of the latent representation in the entire space. Empirical evaluations on the GLUE benchmark, as well as two additional demanding scenarios involving domain and language generalization, demonstrate that G-Tuning can accurately map the universal representation to the target space, thus effectively enhancing the generalization performance of Plms across various downstream tasks.</abstract>
<identifier type="citekey">weng-etal-2023-g</identifier>
<identifier type="doi">10.18653/v1/2023.findings-acl.291</identifier>
<location>
<url>https://aclanthology.org/2023.findings-acl.291/</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>4747</start>
<end>4755</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T G-Tuning: Improving Generalization of Pre-trained Language Models with Generative Adversarial Network
%A Weng, Rongxiang
%A Cheng, Wen Sen
%A Zhang, Min
%Y Rogers, Anna
%Y Boyd-Graber, Jordan
%Y Okazaki, Naoaki
%S Findings of the Association for Computational Linguistics: ACL 2023
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F weng-etal-2023-g
%X The generalization ability of pre-trained language models (Plms) in downstream tasks is heavily influenced by fine-tuning. The objective of fine-tuning is to transform the latent representation of Plms from a universal space to a target space, allowing the model to be applied to downstream tasks with the capability of generalizing to unseen samples. However, the effect of Plms will be diminished when the training data coverage is insufficient, in which fine-tuning is inadequate to learn the complete mapping. In this study, we propose a new fine-tuning framework, referred to as G-Tuning, that aims to preserve the generalization ability of Plms in downstream tasks. Specifically, we integrate a generative adversarial network into the fine-tuning process to aid in the transformation of the latent representation in the entire space. Empirical evaluations on the GLUE benchmark, as well as two additional demanding scenarios involving domain and language generalization, demonstrate that G-Tuning can accurately map the universal representation to the target space, thus effectively enhancing the generalization performance of Plms across various downstream tasks.
%R 10.18653/v1/2023.findings-acl.291
%U https://aclanthology.org/2023.findings-acl.291/
%U https://doi.org/10.18653/v1/2023.findings-acl.291
%P 4747-4755
Markdown (Informal)
[G-Tuning: Improving Generalization of Pre-trained Language Models with Generative Adversarial Network](https://aclanthology.org/2023.findings-acl.291/) (Weng et al., Findings 2023)
ACL