<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="qian-yan-2022-language">
    <titleInfo>
        <title>Language Model Detoxification in Dialogue with Contextualized Stance Control</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Jing</namePart>
        <namePart type="family">Qian</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xifeng</namePart>
        <namePart type="family">Yan</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2022-12</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: EMNLP 2022</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Yoav</namePart>
            <namePart type="family">Goldberg</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Zornitsa</namePart>
            <namePart type="family">Kozareva</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Yue</namePart>
            <namePart type="family">Zhang</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Abu Dhabi, United Arab Emirates</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>To reduce the toxic degeneration in a pretrained Language Model (LM), previous work on Language Model detoxification has focused on reducing the toxicity of the generation itself (self-toxicity) without consideration of the context. As a result, a type of implicit offensive language where the generations support the offensive language in the context is ignored. Different from the LM controlling tasks in previous work, where the desired attributes are fixed for generation, the desired stance of the generation depends on the offensiveness of the context. Therefore, we propose a novel control method to do context-dependent detoxification with the stance taken into consideration. We introduce meta prefixes to learn the contextualized stance control strategy and to generate the stance control prefix according to the input context. The generated stance prefix is then combined with the toxicity control prefix to guide the response generation. Experimental results show that our proposed method can effectively learn the context-dependent stance control strategies while keeping a low self-toxicity of the underlying LM.</abstract>
    <identifier type="citekey">qian-yan-2022-language</identifier>
    <identifier type="doi">10.18653/v1/2022.findings-emnlp.406</identifier>
    <location>
        <url>https://aclanthology.org/2022.findings-emnlp.406/</url>
    </location>
    <part>
        <date>2022-12</date>
        <extent unit="page">
            <start>5548</start>
            <end>5558</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Language Model Detoxification in Dialogue with Contextualized Stance Control
%A Qian, Jing
%A Yan, Xifeng
%Y Goldberg, Yoav
%Y Kozareva, Zornitsa
%Y Zhang, Yue
%S Findings of the Association for Computational Linguistics: EMNLP 2022
%D 2022
%8 December
%I Association for Computational Linguistics
%C Abu Dhabi, United Arab Emirates
%F qian-yan-2022-language
%X To reduce the toxic degeneration in a pretrained Language Model (LM), previous work on Language Model detoxification has focused on reducing the toxicity of the generation itself (self-toxicity) without consideration of the context. As a result, a type of implicit offensive language where the generations support the offensive language in the context is ignored. Different from the LM controlling tasks in previous work, where the desired attributes are fixed for generation, the desired stance of the generation depends on the offensiveness of the context. Therefore, we propose a novel control method to do context-dependent detoxification with the stance taken into consideration. We introduce meta prefixes to learn the contextualized stance control strategy and to generate the stance control prefix according to the input context. The generated stance prefix is then combined with the toxicity control prefix to guide the response generation. Experimental results show that our proposed method can effectively learn the context-dependent stance control strategies while keeping a low self-toxicity of the underlying LM.
%R 10.18653/v1/2022.findings-emnlp.406
%U https://aclanthology.org/2022.findings-emnlp.406/
%U https://doi.org/10.18653/v1/2022.findings-emnlp.406
%P 5548-5558

Markdown (Informal)

[Language Model Detoxification in Dialogue with Contextualized Stance Control](https://aclanthology.org/2022.findings-emnlp.406/) (Qian & Yan, Findings 2022)

• Language Model Detoxification in Dialogue with Contextualized Stance Control (Qian & Yan, Findings 2022)

ACL

• Jing Qian and Xifeng Yan. 2022. Language Model Detoxification in Dialogue with Contextualized Stance Control. In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 5548–5558, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.