@inproceedings{zhan-etal-2023-contrastive,
title = "Contrastive Learning with Adversarial Examples for Alleviating Pathology of Language Model",
author = "Zhan, Pengwei and
Yang, Jing and
Huang, Xiao and
Jing, Chunlei and
Li, Jingying and
Wang, Liming",
editor = "Rogers, Anna and
Boyd-Graber, Jordan and
Okazaki, Naoaki",
booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.acl-long.358/",
doi = "10.18653/v1/2023.acl-long.358",
pages = "6493--6508",
abstract = "Neural language models have achieved superior performance. However, these models also suffer from the pathology of overconfidence in the out-of-distribution examples, potentially making the model difficult to interpret and making the interpretation methods fail to provide faithful attributions. In this paper, we explain the model pathology from the view of sentence representation and argue that the counter-intuitive bias degree and direction of the out-of-distribution examples' representation cause the pathology. We propose a Contrastive learning regularization method using Adversarial examples for Alleviating the Pathology (ConAAP), which calibrates the sentence representation of out-of-distribution examples. ConAAP generates positive and negative examples following the attribution results and utilizes adversarial examples to introduce direction information in regularization. Experiments show that ConAAP effectively alleviates the model pathology while slightly impacting the generalization ability on in-distribution examples and thus helps interpretation methods obtain more faithful results."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhan-etal-2023-contrastive">
<titleInfo>
<title>Contrastive Learning with Adversarial Examples for Alleviating Pathology of Language Model</title>
</titleInfo>
<name type="personal">
<namePart type="given">Pengwei</namePart>
<namePart type="family">Zhan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jing</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiao</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chunlei</namePart>
<namePart type="family">Jing</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jingying</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Liming</namePart>
<namePart type="family">Wang</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>Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</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>Neural language models have achieved superior performance. However, these models also suffer from the pathology of overconfidence in the out-of-distribution examples, potentially making the model difficult to interpret and making the interpretation methods fail to provide faithful attributions. In this paper, we explain the model pathology from the view of sentence representation and argue that the counter-intuitive bias degree and direction of the out-of-distribution examples’ representation cause the pathology. We propose a Contrastive learning regularization method using Adversarial examples for Alleviating the Pathology (ConAAP), which calibrates the sentence representation of out-of-distribution examples. ConAAP generates positive and negative examples following the attribution results and utilizes adversarial examples to introduce direction information in regularization. Experiments show that ConAAP effectively alleviates the model pathology while slightly impacting the generalization ability on in-distribution examples and thus helps interpretation methods obtain more faithful results.</abstract>
<identifier type="citekey">zhan-etal-2023-contrastive</identifier>
<identifier type="doi">10.18653/v1/2023.acl-long.358</identifier>
<location>
<url>https://aclanthology.org/2023.acl-long.358/</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>6493</start>
<end>6508</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Contrastive Learning with Adversarial Examples for Alleviating Pathology of Language Model
%A Zhan, Pengwei
%A Yang, Jing
%A Huang, Xiao
%A Jing, Chunlei
%A Li, Jingying
%A Wang, Liming
%Y Rogers, Anna
%Y Boyd-Graber, Jordan
%Y Okazaki, Naoaki
%S Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F zhan-etal-2023-contrastive
%X Neural language models have achieved superior performance. However, these models also suffer from the pathology of overconfidence in the out-of-distribution examples, potentially making the model difficult to interpret and making the interpretation methods fail to provide faithful attributions. In this paper, we explain the model pathology from the view of sentence representation and argue that the counter-intuitive bias degree and direction of the out-of-distribution examples’ representation cause the pathology. We propose a Contrastive learning regularization method using Adversarial examples for Alleviating the Pathology (ConAAP), which calibrates the sentence representation of out-of-distribution examples. ConAAP generates positive and negative examples following the attribution results and utilizes adversarial examples to introduce direction information in regularization. Experiments show that ConAAP effectively alleviates the model pathology while slightly impacting the generalization ability on in-distribution examples and thus helps interpretation methods obtain more faithful results.
%R 10.18653/v1/2023.acl-long.358
%U https://aclanthology.org/2023.acl-long.358/
%U https://doi.org/10.18653/v1/2023.acl-long.358
%P 6493-6508
Markdown (Informal)
[Contrastive Learning with Adversarial Examples for Alleviating Pathology of Language Model](https://aclanthology.org/2023.acl-long.358/) (Zhan et al., ACL 2023)
ACL