<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="welivita-etal-2026-generates">
    <titleInfo>
        <title>Who Generates More Empathetic Responses—Humans or LLMs? A Comparative Evaluation with Human and LLM Judges</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Anuradha</namePart>
        <namePart type="family">Welivita</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Fawzia</namePart>
        <namePart type="family">Zeitoun</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Pearl</namePart>
        <namePart type="family">Pu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 30th Conference on Computational Natural Language Learning</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Claire</namePart>
            <namePart type="family">Bonial</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Yevgeni</namePart>
            <namePart type="family">Berzak</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">San Diego, California, USA</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-410-1</identifier>
    </relatedItem>
    <abstract>This paper compares the empathetic quality of responses generated by humans and large language models (LLMs). We evaluate four LLMs that were widely used at the time of study—GPT-4, LLaMA-2-70B-Chat, Gemini-1.0-Pro, and Mixtral-8×7B-Instruct—against a human baseline using a large-scale between-subjects study. A total of 1,000 human participants evaluated the empathetic quality of human- and LLM-generated responses to 2,000 dialogue prompts spanning 32 positive and negative emotions. To complement human judgments, we also employed an LLM-as-judge (GPT-4o-mini) to assess the same responses. Across emotions and evaluators, LLM-generated responses were rated as significantly more empathetic than human-written responses. We also observed that both human judges and the LLM-as-judge tended to rate responses generated by their own group more favorably, indicating self-favoring tendencies. These findings highlight both the strong performance of contemporary LLMs in empathetic responding and the need to interpret human- and LLM-based evaluations with care.</abstract>
    <identifier type="citekey">welivita-etal-2026-generates</identifier>
    <location>
        <url>https://aclanthology.org/2026.conll-main.21/</url>
    </location>
    <part>
        <date>2026-07</date>
        <extent unit="page">
            <start>358</start>
            <end>381</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Who Generates More Empathetic Responses—Humans or LLMs? A Comparative Evaluation with Human and LLM Judges
%A Welivita, Anuradha
%A Zeitoun, Fawzia
%A Pu, Pearl
%Y Bonial, Claire
%Y Berzak, Yevgeni
%S Proceedings of the 30th Conference on Computational Natural Language Learning
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, USA
%@ 979-8-89176-410-1
%F welivita-etal-2026-generates
%X This paper compares the empathetic quality of responses generated by humans and large language models (LLMs). We evaluate four LLMs that were widely used at the time of study—GPT-4, LLaMA-2-70B-Chat, Gemini-1.0-Pro, and Mixtral-8×7B-Instruct—against a human baseline using a large-scale between-subjects study. A total of 1,000 human participants evaluated the empathetic quality of human- and LLM-generated responses to 2,000 dialogue prompts spanning 32 positive and negative emotions. To complement human judgments, we also employed an LLM-as-judge (GPT-4o-mini) to assess the same responses. Across emotions and evaluators, LLM-generated responses were rated as significantly more empathetic than human-written responses. We also observed that both human judges and the LLM-as-judge tended to rate responses generated by their own group more favorably, indicating self-favoring tendencies. These findings highlight both the strong performance of contemporary LLMs in empathetic responding and the need to interpret human- and LLM-based evaluations with care.
%U https://aclanthology.org/2026.conll-main.21/
%P 358-381

Markdown (Informal)

[Who Generates More Empathetic Responses—Humans or LLMs? A Comparative Evaluation with Human and LLM Judges](https://aclanthology.org/2026.conll-main.21/) (Welivita et al., CoNLL 2026)

• Who Generates More Empathetic Responses—Humans or LLMs? A Comparative Evaluation with Human and LLM Judges (Welivita et al., CoNLL 2026)

ACL

• Anuradha Welivita, Fawzia Zeitoun, and Pearl Pu. 2026. Who Generates More Empathetic Responses—Humans or LLMs? A Comparative Evaluation with Human and LLM Judges. In Proceedings of the 30th Conference on Computational Natural Language Learning, pages 358–381, San Diego, California, USA. Association for Computational Linguistics.