<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="remaki-etal-2025-limics">
    <titleInfo>
        <title>LIMICS at ArchEHR-QA 2025: Prompting LLMs Beats Fine-Tuned Embeddings</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Adam</namePart>
        <namePart type="family">Remaki</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Armand</namePart>
        <namePart type="family">Violle</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Vikram</namePart>
        <namePart type="family">Natraj</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Étienne</namePart>
        <namePart type="family">Guével</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Akram</namePart>
        <namePart type="family">Redjdal</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-08</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 24th Workshop on Biomedical Language Processing (Shared Tasks)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Sarvesh</namePart>
            <namePart type="family">Soni</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Dina</namePart>
            <namePart type="family">Demner-Fushman</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Vienna, Austria</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-276-3</identifier>
    </relatedItem>
    <abstract>In this paper, we investigated two approaches to clinical question-answering based on patient-formulated questions, supported by their narratives and brief medical records. The first approach leverages zero- and few-shot prompt engineering techniques with GPT-based Large Language Models (LLMs), incorporating strategies such as prompt chaining and chain-of-thought reasoning to guide the models in generating answers. The second approach adopts a two-steps structure: first, a text-classification stage uses embedding-based models (e.g., BERT variants) to identify sentences within the medical record that are most relevant to the given question; then, we prompt an LLM to paraphrase them into an answer so that it is generated exclusively from these selected sentences. Our empirical results demonstrate that the first approach outperforms the classification-guided pipeline, achieving the highest score on the development set and the test set using prompt chaining. Code: github.com/armandviolle/BioNLP-2025</abstract>
    <identifier type="citekey">remaki-etal-2025-limics</identifier>
    <identifier type="doi">10.18653/v1/2025.bionlp-share.18</identifier>
    <location>
        <url>https://aclanthology.org/2025.bionlp-share.18/</url>
    </location>
    <part>
        <date>2025-08</date>
        <extent unit="page">
            <start>150</start>
            <end>159</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T LIMICS at ArchEHR-QA 2025: Prompting LLMs Beats Fine-Tuned Embeddings
%A Remaki, Adam
%A Violle, Armand
%A Natraj, Vikram
%A Guével, Étienne
%A Redjdal, Akram
%Y Soni, Sarvesh
%Y Demner-Fushman, Dina
%S Proceedings of the 24th Workshop on Biomedical Language Processing (Shared Tasks)
%D 2025
%8 August
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-276-3
%F remaki-etal-2025-limics
%X In this paper, we investigated two approaches to clinical question-answering based on patient-formulated questions, supported by their narratives and brief medical records. The first approach leverages zero- and few-shot prompt engineering techniques with GPT-based Large Language Models (LLMs), incorporating strategies such as prompt chaining and chain-of-thought reasoning to guide the models in generating answers. The second approach adopts a two-steps structure: first, a text-classification stage uses embedding-based models (e.g., BERT variants) to identify sentences within the medical record that are most relevant to the given question; then, we prompt an LLM to paraphrase them into an answer so that it is generated exclusively from these selected sentences. Our empirical results demonstrate that the first approach outperforms the classification-guided pipeline, achieving the highest score on the development set and the test set using prompt chaining. Code: github.com/armandviolle/BioNLP-2025
%R 10.18653/v1/2025.bionlp-share.18
%U https://aclanthology.org/2025.bionlp-share.18/
%U https://doi.org/10.18653/v1/2025.bionlp-share.18
%P 150-159

Markdown (Informal)

[LIMICS at ArchEHR-QA 2025: Prompting LLMs Beats Fine-Tuned Embeddings](https://aclanthology.org/2025.bionlp-share.18/) (Remaki et al., BioNLP 2025)

• LIMICS at ArchEHR-QA 2025: Prompting LLMs Beats Fine-Tuned Embeddings (Remaki et al., BioNLP 2025)

ACL

• Adam Remaki, Armand Violle, Vikram Natraj, Étienne Guével, and Akram Redjdal. 2025. LIMICS at ArchEHR-QA 2025: Prompting LLMs Beats Fine-Tuned Embeddings. In Proceedings of the 24th Workshop on Biomedical Language Processing (Shared Tasks), pages 150–159, Vienna, Austria. Association for Computational Linguistics.