<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="cornelius-etal-2026-cane">
    <titleInfo>
        <title>CANE is a Format: Rethinking Output Formats for Few-Shot LLM-Based Named Entity Recognition</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Joseph</namePart>
        <namePart type="family">Cornelius</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Kei</namePart>
        <namePart type="given">K</namePart>
        <namePart type="family">Ito</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Fabio</namePart>
        <namePart type="family">Rinaldi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-462-0</identifier>
    </relatedItem>
    <abstract>Output format selection is a necessary but understudied dimension of few-shot named entity recognition (NER) with large language models. Existing formats carry characteristic failure modes under autoregressive generation, yet output format has not been systematically evaluated as a primary variable in the few-shot setting across diverse domains and entity structural complexity. We evaluate 21 output formats over 17 datasets across two levels of structural entity complexity, using 9 open-weight models without fine-tuning, ranging from 1.7B to 70B parameters. Output format is the main factor affecting performance, with a 41.7 F1 points difference between the best and worst formats. This exceeds the gains from scaling models parameters within-family. Furthermore, output format accounts for approximately 49.1% of F1 variance on simple datasets and 74.9% on complex ones. We describe three mechanistically distinct failure modes that explain these observed differences; each existing format violates at least one of three properties jointly necessary for robust few-shot NER. We propose CANE (Context-Anchored Natural Expression), a format family addressing all three properties. All CANE variants ranked the highest on simple and complex datasets. This includes the GENIA dataset where all other formats had near-zero scores across nine models, indicating the necessity of the three properties. These findings highlight the importance of output format in few-shot named entity recognition, particularly for annotations involving overlapping or discontinuous entities.</abstract>
    <identifier type="citekey">cornelius-etal-2026-cane</identifier>
    <identifier type="doi">10.18653/2026.findings-emnlp.1398</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-emnlp.1398/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>29773</start>
            <end>29800</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T CANE is a Format: Rethinking Output Formats for Few-Shot LLM-Based Named Entity Recognition
%A Cornelius, Joseph
%A Ito, Kei K.
%A Rinaldi, Fabio
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F cornelius-etal-2026-cane
%X Output format selection is a necessary but understudied dimension of few-shot named entity recognition (NER) with large language models. Existing formats carry characteristic failure modes under autoregressive generation, yet output format has not been systematically evaluated as a primary variable in the few-shot setting across diverse domains and entity structural complexity. We evaluate 21 output formats over 17 datasets across two levels of structural entity complexity, using 9 open-weight models without fine-tuning, ranging from 1.7B to 70B parameters. Output format is the main factor affecting performance, with a 41.7 F1 points difference between the best and worst formats. This exceeds the gains from scaling models parameters within-family. Furthermore, output format accounts for approximately 49.1% of F1 variance on simple datasets and 74.9% on complex ones. We describe three mechanistically distinct failure modes that explain these observed differences; each existing format violates at least one of three properties jointly necessary for robust few-shot NER. We propose CANE (Context-Anchored Natural Expression), a format family addressing all three properties. All CANE variants ranked the highest on simple and complex datasets. This includes the GENIA dataset where all other formats had near-zero scores across nine models, indicating the necessity of the three properties. These findings highlight the importance of output format in few-shot named entity recognition, particularly for annotations involving overlapping or discontinuous entities.
%R 10.18653/2026.findings-emnlp.1398
%U https://aclanthology.org/2026.findings-emnlp.1398/
%U https://doi.org/10.18653/2026.findings-emnlp.1398
%P 29773-29800

Markdown (Informal)

[CANE is a Format: Rethinking Output Formats for Few-Shot LLM-Based Named Entity Recognition](https://aclanthology.org/2026.findings-emnlp.1398/) (Cornelius et al., Findings 2026)

• CANE is a Format: Rethinking Output Formats for Few-Shot LLM-Based Named Entity Recognition (Cornelius et al., Findings 2026)

ACL

• Joseph Cornelius, Kei K Ito, and Fabio Rinaldi. 2026. CANE is a Format: Rethinking Output Formats for Few-Shot LLM-Based Named Entity Recognition. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 29773–29800, Budapest, Hungary. Association for Computational Linguistics.