@inproceedings{modarressi-etal-2023-guide,
title = "Guide the Learner: Controlling Product of Experts Debiasing Method Based on Token Attribution Similarities",
author = "Modarressi, Ali and
Amirkhani, Hossein and
Pilehvar, Mohammad Taher",
editor = "Vlachos, Andreas and
Augenstein, Isabelle",
booktitle = "Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics",
month = may,
year = "2023",
address = "Dubrovnik, Croatia",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.eacl-main.143",
doi = "10.18653/v1/2023.eacl-main.143",
pages = "1954--1959",
abstract = "Several proposals have been put forward in recent years for improving out-of-distribution (OOD) performance through mitigating dataset biases. A popular workaround is to train a robust model by re-weighting training examples based on a secondary biased model. Here, the underlying assumption is that the biased model resorts to shortcut features. Hence, those training examples that are correctly predicted by the biased model are flagged as being biased and are down-weighted during the training of the main model. However, assessing the importance of an instance merely based on the predictions of the biased model may be too naive. It is possible that the prediction of the main model can be derived from another decision-making process that is distinct from the behavior of the biased model. To circumvent this, we introduce a fine-tuning strategy that incorporates the similarity between the main and biased model attribution scores in a Product of Experts (PoE) loss function to further improve OOD performance. With experiments conducted on natural language inference and fact verification benchmarks, we show that our method improves OOD results while maintaining in-distribution (ID) performance.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="modarressi-etal-2023-guide">
<titleInfo>
<title>Guide the Learner: Controlling Product of Experts Debiasing Method Based on Token Attribution Similarities</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ali</namePart>
<namePart type="family">Modarressi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hossein</namePart>
<namePart type="family">Amirkhani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics</title>
</titleInfo>
<name type="personal">
<namePart type="given">Andreas</namePart>
<namePart type="family">Vlachos</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Isabelle</namePart>
<namePart type="family">Augenstein</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Dubrovnik, Croatia</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Several proposals have been put forward in recent years for improving out-of-distribution (OOD) performance through mitigating dataset biases. A popular workaround is to train a robust model by re-weighting training examples based on a secondary biased model. Here, the underlying assumption is that the biased model resorts to shortcut features. Hence, those training examples that are correctly predicted by the biased model are flagged as being biased and are down-weighted during the training of the main model. However, assessing the importance of an instance merely based on the predictions of the biased model may be too naive. It is possible that the prediction of the main model can be derived from another decision-making process that is distinct from the behavior of the biased model. To circumvent this, we introduce a fine-tuning strategy that incorporates the similarity between the main and biased model attribution scores in a Product of Experts (PoE) loss function to further improve OOD performance. With experiments conducted on natural language inference and fact verification benchmarks, we show that our method improves OOD results while maintaining in-distribution (ID) performance.</abstract>
<identifier type="citekey">modarressi-etal-2023-guide</identifier>
<identifier type="doi">10.18653/v1/2023.eacl-main.143</identifier>
<location>
<url>https://aclanthology.org/2023.eacl-main.143</url>
</location>
<part>
<date>2023-05</date>
<extent unit="page">
<start>1954</start>
<end>1959</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Guide the Learner: Controlling Product of Experts Debiasing Method Based on Token Attribution Similarities
%A Modarressi, Ali
%A Amirkhani, Hossein
%A Pilehvar, Mohammad Taher
%Y Vlachos, Andreas
%Y Augenstein, Isabelle
%S Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics
%D 2023
%8 May
%I Association for Computational Linguistics
%C Dubrovnik, Croatia
%F modarressi-etal-2023-guide
%X Several proposals have been put forward in recent years for improving out-of-distribution (OOD) performance through mitigating dataset biases. A popular workaround is to train a robust model by re-weighting training examples based on a secondary biased model. Here, the underlying assumption is that the biased model resorts to shortcut features. Hence, those training examples that are correctly predicted by the biased model are flagged as being biased and are down-weighted during the training of the main model. However, assessing the importance of an instance merely based on the predictions of the biased model may be too naive. It is possible that the prediction of the main model can be derived from another decision-making process that is distinct from the behavior of the biased model. To circumvent this, we introduce a fine-tuning strategy that incorporates the similarity between the main and biased model attribution scores in a Product of Experts (PoE) loss function to further improve OOD performance. With experiments conducted on natural language inference and fact verification benchmarks, we show that our method improves OOD results while maintaining in-distribution (ID) performance.
%R 10.18653/v1/2023.eacl-main.143
%U https://aclanthology.org/2023.eacl-main.143
%U https://doi.org/10.18653/v1/2023.eacl-main.143
%P 1954-1959
Markdown (Informal)
[Guide the Learner: Controlling Product of Experts Debiasing Method Based on Token Attribution Similarities](https://aclanthology.org/2023.eacl-main.143) (Modarressi et al., EACL 2023)
ACL