<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="cheng-etal-2023-compost">
    <titleInfo>
        <title>CoMPosT: Characterizing and Evaluating Caricature in LLM Simulations</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Myra</namePart>
        <namePart type="family">Cheng</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Tiziano</namePart>
        <namePart type="family">Piccardi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Diyi</namePart>
        <namePart type="family">Yang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2023-12</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Houda</namePart>
            <namePart type="family">Bouamor</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Juan</namePart>
            <namePart type="family">Pino</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Kalika</namePart>
            <namePart type="family">Bali</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Singapore</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>Recent work has aimed to capture nuances of human behavior by using LLMs to simulate responses from particular demographics in settings like social science experiments and public opinion surveys. However, there are currently no established ways to discuss or evaluate the quality of such LLM simulations. Moreover, there is growing concern that these LLM simulations are flattened caricatures of the personas that they aim to simulate, failing to capture the multidimensionality of people and perpetuating stereotypes. To bridge these gaps, we present CoMPosT, a framework to characterize LLM simulations using four dimensions: Context, Model, Persona, and Topic. We use this framework to measure open-ended LLM simulations’ susceptibility to caricature, defined via two criteria: individuation and exaggeration. We evaluate the level of caricature in scenarios from existing work on LLM simulations. We find that for GPT-4, simulations of certain demographics (political and marginalized groups) and topics (general, uncontroversial) are highly susceptible to caricature.</abstract>
    <identifier type="citekey">cheng-etal-2023-compost</identifier>
    <identifier type="doi">10.18653/v1/2023.emnlp-main.669</identifier>
    <location>
        <url>https://aclanthology.org/2023.emnlp-main.669</url>
    </location>
    <part>
        <date>2023-12</date>
        <extent unit="page">
            <start>10853</start>
            <end>10875</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T CoMPosT: Characterizing and Evaluating Caricature in LLM Simulations
%A Cheng, Myra
%A Piccardi, Tiziano
%A Yang, Diyi
%Y Bouamor, Houda
%Y Pino, Juan
%Y Bali, Kalika
%S Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
%D 2023
%8 December
%I Association for Computational Linguistics
%C Singapore
%F cheng-etal-2023-compost
%X Recent work has aimed to capture nuances of human behavior by using LLMs to simulate responses from particular demographics in settings like social science experiments and public opinion surveys. However, there are currently no established ways to discuss or evaluate the quality of such LLM simulations. Moreover, there is growing concern that these LLM simulations are flattened caricatures of the personas that they aim to simulate, failing to capture the multidimensionality of people and perpetuating stereotypes. To bridge these gaps, we present CoMPosT, a framework to characterize LLM simulations using four dimensions: Context, Model, Persona, and Topic. We use this framework to measure open-ended LLM simulations’ susceptibility to caricature, defined via two criteria: individuation and exaggeration. We evaluate the level of caricature in scenarios from existing work on LLM simulations. We find that for GPT-4, simulations of certain demographics (political and marginalized groups) and topics (general, uncontroversial) are highly susceptible to caricature.
%R 10.18653/v1/2023.emnlp-main.669
%U https://aclanthology.org/2023.emnlp-main.669
%U https://doi.org/10.18653/v1/2023.emnlp-main.669
%P 10853-10875

Markdown (Informal)

[CoMPosT: Characterizing and Evaluating Caricature in LLM Simulations](https://aclanthology.org/2023.emnlp-main.669) (Cheng et al., EMNLP 2023)

• CoMPosT: Characterizing and Evaluating Caricature in LLM Simulations (Cheng et al., EMNLP 2023)

ACL

• Myra Cheng, Tiziano Piccardi, and Diyi Yang. 2023. CoMPosT: Characterizing and Evaluating Caricature in LLM Simulations. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 10853–10875, Singapore. Association for Computational Linguistics.