@inproceedings{zhang-wan-2023-mil,
title = "{MIL}-Decoding: Detoxifying Language Models at Token-Level via Multiple Instance Learning",
author = "Zhang, Xu and
Wan, Xiaojun",
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.11/",
doi = "10.18653/v1/2023.acl-long.11",
pages = "190--202",
abstract = "Despite advances in large pre-trained neural language models, they are prone to generating toxic language, which brings security risks to their applications. We introduce MIL-Decoding, which detoxifies language models at token-level by interpolating it with a trained multiple instance learning (MIL) network.MIL model is trained on a corpus with a toxicity label for each text to predict the overall toxicity and the toxicity of each token in its context. Intuitively, the MIL network computes a toxicity distribution over next tokens according to the generated context which supplements the original language model to avoid toxicity. We evaluate MIL-Decoding with automatic metrics and human evaluation, where MIL-Decoding outperforms other baselines in detoxification while it only hurts generation fluency a little bit."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhang-wan-2023-mil">
<titleInfo>
<title>MIL-Decoding: Detoxifying Language Models at Token-Level via Multiple Instance Learning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xu</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaojun</namePart>
<namePart type="family">Wan</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>Despite advances in large pre-trained neural language models, they are prone to generating toxic language, which brings security risks to their applications. We introduce MIL-Decoding, which detoxifies language models at token-level by interpolating it with a trained multiple instance learning (MIL) network.MIL model is trained on a corpus with a toxicity label for each text to predict the overall toxicity and the toxicity of each token in its context. Intuitively, the MIL network computes a toxicity distribution over next tokens according to the generated context which supplements the original language model to avoid toxicity. We evaluate MIL-Decoding with automatic metrics and human evaluation, where MIL-Decoding outperforms other baselines in detoxification while it only hurts generation fluency a little bit.</abstract>
<identifier type="citekey">zhang-wan-2023-mil</identifier>
<identifier type="doi">10.18653/v1/2023.acl-long.11</identifier>
<location>
<url>https://aclanthology.org/2023.acl-long.11/</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>190</start>
<end>202</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T MIL-Decoding: Detoxifying Language Models at Token-Level via Multiple Instance Learning
%A Zhang, Xu
%A Wan, Xiaojun
%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 zhang-wan-2023-mil
%X Despite advances in large pre-trained neural language models, they are prone to generating toxic language, which brings security risks to their applications. We introduce MIL-Decoding, which detoxifies language models at token-level by interpolating it with a trained multiple instance learning (MIL) network.MIL model is trained on a corpus with a toxicity label for each text to predict the overall toxicity and the toxicity of each token in its context. Intuitively, the MIL network computes a toxicity distribution over next tokens according to the generated context which supplements the original language model to avoid toxicity. We evaluate MIL-Decoding with automatic metrics and human evaluation, where MIL-Decoding outperforms other baselines in detoxification while it only hurts generation fluency a little bit.
%R 10.18653/v1/2023.acl-long.11
%U https://aclanthology.org/2023.acl-long.11/
%U https://doi.org/10.18653/v1/2023.acl-long.11
%P 190-202
Markdown (Informal)
[MIL-Decoding: Detoxifying Language Models at Token-Level via Multiple Instance Learning](https://aclanthology.org/2023.acl-long.11/) (Zhang & Wan, ACL 2023)
ACL