<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kohli-2026-metric">
    <titleInfo>
        <title>Metric-Dependent Annotation Saturation for Learning from Label Distributions</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Guneet</namePart>
        <namePart type="family">Kohli</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>Proceedings of the First Workshop on Pluralistic AI NLP: Diversity-aware, Sociotechnical, Responsible Alignment (PANDORA 2026)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Shanshan</namePart>
            <namePart type="family">Xu</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Santosh</namePart>
            <namePart type="family">T.Y.S.S.</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Kemal</namePart>
            <namePart type="family">Kurniawan</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Siyao</namePart>
            <namePart type="family">Peng</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sarenne</namePart>
            <namePart type="family">Wallbridge</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Fajri</namePart>
            <namePart type="family">Koto</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Nouha</namePart>
            <namePart type="family">Dziri</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Barbara</namePart>
            <namePart type="family">Plank</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">979-8-89176-484-2</identifier>
    </relatedItem>
    <abstract>When annotators disagree on a label, the disagreement itself carries signal—and the number of annotators needed to capture it depends on the evaluation metric. We fine-tune NLI models on label distributions subsampled from ChaosNLI, a dataset providing 100 independent annotator judgments per item, and identify metric-dependent saturation. In our 3-class NLI setting, entropy correlation—whether the model identifies which items elicit disagreement—requires N \approx 20–50 annotators to converge, while distributional match (KL divergence) saturates by N \approx 10 (87–95% of improvement across five model seeds); the specific thresholds are estimates for this setting, and it is the phenomenon, not the constants, that we expect to generalize. This finding rests on a prior observation: soft labels carry item-specific signal that label smoothing cannot replicate. Across five smoothing intensities, entropy correlation clusters at r \approx 0.45–0.49, while soft labels reach r = 0.643 (p &lt; 0.001); per-item analysis traces this gap to smoothing’s inability to distinguish ambiguous items from clear ones. The soft-label advantage replicates across two architectures (DeBERTa, RoBERTa) and a non-NLI-pretrained baseline; differential saturation additionally appears in an exploratory cross-domain evaluation on content safety. These results suggest that annotation budgets should be informed by the target evaluation metric rather than set uniformly.</abstract>
    <identifier type="citekey">kohli-2026-metric</identifier>
    <location>
        <url>https://aclanthology.org/2026.pandora-1.4/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>88</start>
            <end>104</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Metric-Dependent Annotation Saturation for Learning from Label Distributions
%A Kohli, Guneet
%Y Xu, Shanshan
%Y T.Y.S.S., Santosh
%Y Kurniawan, Kemal
%Y Peng, Siyao
%Y Wallbridge, Sarenne
%Y Koto, Fajri
%Y Dziri, Nouha
%Y Plank, Barbara
%S Proceedings of the First Workshop on Pluralistic AI NLP: Diversity-aware, Sociotechnical, Responsible Alignment (PANDORA 2026)
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-484-2
%F kohli-2026-metric
%X When annotators disagree on a label, the disagreement itself carries signal—and the number of annotators needed to capture it depends on the evaluation metric. We fine-tune NLI models on label distributions subsampled from ChaosNLI, a dataset providing 100 independent annotator judgments per item, and identify metric-dependent saturation. In our 3-class NLI setting, entropy correlation—whether the model identifies which items elicit disagreement—requires N \approx 20–50 annotators to converge, while distributional match (KL divergence) saturates by N \approx 10 (87–95% of improvement across five model seeds); the specific thresholds are estimates for this setting, and it is the phenomenon, not the constants, that we expect to generalize. This finding rests on a prior observation: soft labels carry item-specific signal that label smoothing cannot replicate. Across five smoothing intensities, entropy correlation clusters at r \approx 0.45–0.49, while soft labels reach r = 0.643 (p < 0.001); per-item analysis traces this gap to smoothing’s inability to distinguish ambiguous items from clear ones. The soft-label advantage replicates across two architectures (DeBERTa, RoBERTa) and a non-NLI-pretrained baseline; differential saturation additionally appears in an exploratory cross-domain evaluation on content safety. These results suggest that annotation budgets should be informed by the target evaluation metric rather than set uniformly.
%U https://aclanthology.org/2026.pandora-1.4/
%P 88-104

Markdown (Informal)

[Metric-Dependent Annotation Saturation for Learning from Label Distributions](https://aclanthology.org/2026.pandora-1.4/) (Kohli, PANDORA 2026)

• Metric-Dependent Annotation Saturation for Learning from Label Distributions (Kohli, PANDORA 2026)

ACL

• Guneet Kohli. 2026. Metric-Dependent Annotation Saturation for Learning from Label Distributions. In Proceedings of the First Workshop on Pluralistic AI NLP: Diversity-aware, Sociotechnical, Responsible Alignment (PANDORA 2026), pages 88–104, Budapest, Hungary. Association for Computational Linguistics.