<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="movva-etal-2024-annotation">
    <titleInfo>
        <title>Annotation alignment: Comparing LLM and human annotations of conversational safety</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Rajiv</namePart>
        <namePart type="family">Movva</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Pang</namePart>
        <namePart type="given">Wei</namePart>
        <namePart type="family">Koh</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Emma</namePart>
        <namePart type="family">Pierson</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2024-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Yaser</namePart>
            <namePart type="family">Al-Onaizan</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Mohit</namePart>
            <namePart type="family">Bansal</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Yun-Nung</namePart>
            <namePart type="family">Chen</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Miami, Florida, USA</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>Do LLMs align with human perceptions of safety? We study this question via *annotation alignment*, the extent to which LLMs and humans agree when annotating the safety of user-chatbot conversations. We leverage the recent DICES dataset (Aroyo et al. 2023), in which 350 conversations are each rated for safety by 112 annotators spanning 10 race-gender groups. GPT-4 achieves a Pearson correlation of r=0.59 with the average annotator rating, higher than the median annotator‘s correlation with the average (r=0.51). We show that larger datasets are needed to resolve whether GPT-4 exhibits disparities in how well it correlates with different demographic groups. Also, there is substantial idiosyncratic variation in correlation within groups, suggesting that race &amp; gender do not fully capture differences in alignment. Finally, we find that GPT-4 cannot predict when one demographic group finds a conversation more unsafe than another.</abstract>
    <identifier type="citekey">movva-etal-2024-annotation</identifier>
    <identifier type="doi">10.18653/v1/2024.emnlp-main.511</identifier>
    <location>
        <url>https://aclanthology.org/2024.emnlp-main.511/</url>
    </location>
    <part>
        <date>2024-11</date>
        <extent unit="page">
            <start>9048</start>
            <end>9062</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Annotation alignment: Comparing LLM and human annotations of conversational safety
%A Movva, Rajiv
%A Koh, Pang Wei
%A Pierson, Emma
%Y Al-Onaizan, Yaser
%Y Bansal, Mohit
%Y Chen, Yun-Nung
%S Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F movva-etal-2024-annotation
%X Do LLMs align with human perceptions of safety? We study this question via *annotation alignment*, the extent to which LLMs and humans agree when annotating the safety of user-chatbot conversations. We leverage the recent DICES dataset (Aroyo et al. 2023), in which 350 conversations are each rated for safety by 112 annotators spanning 10 race-gender groups. GPT-4 achieves a Pearson correlation of r=0.59 with the average annotator rating, higher than the median annotator‘s correlation with the average (r=0.51). We show that larger datasets are needed to resolve whether GPT-4 exhibits disparities in how well it correlates with different demographic groups. Also, there is substantial idiosyncratic variation in correlation within groups, suggesting that race & gender do not fully capture differences in alignment. Finally, we find that GPT-4 cannot predict when one demographic group finds a conversation more unsafe than another.
%R 10.18653/v1/2024.emnlp-main.511
%U https://aclanthology.org/2024.emnlp-main.511/
%U https://doi.org/10.18653/v1/2024.emnlp-main.511
%P 9048-9062

Markdown (Informal)

[Annotation alignment: Comparing LLM and human annotations of conversational safety](https://aclanthology.org/2024.emnlp-main.511/) (Movva et al., EMNLP 2024)

• Annotation alignment: Comparing LLM and human annotations of conversational safety (Movva et al., EMNLP 2024)

ACL

• Rajiv Movva, Pang Wei Koh, and Emma Pierson. 2024. Annotation alignment: Comparing LLM and human annotations of conversational safety. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 9048–9062, Miami, Florida, USA. Association for Computational Linguistics.