<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="soiffer-etal-2025-semantic">
    <titleInfo>
        <title>Semantic Agreement Enables Efficient Open-Ended LLM Cascades</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Duncan</namePart>
        <namePart type="family">Soiffer</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Steven</namePart>
        <namePart type="family">Kolawole</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Virginia</namePart>
        <namePart type="family">Smith</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Saloni</namePart>
            <namePart type="family">Potdar</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Lina</namePart>
            <namePart type="family">Rojas-Barahona</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sebastien</namePart>
            <namePart type="family">Montella</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Suzhou (China)</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-333-3</identifier>
    </relatedItem>
    <abstract>Cascade systems for open-ended text generation face a fundamental challenge: determining output reliability when generation quality lies on a continuous spectrum, often with multiple valid responses. To address this, we propose _semantic agreement_—meaning-level consensus between ensemble outputs—as a training-free signal for reliable deferral. We show that when diverse model outputs agree semantically, their consensus is a stronger reliability signal than token-level confidence. Evaluated from 500M to 70B-parameter models, semantic cascades improve deferral accuracy, match or surpass target-model quality at 40% of the cost, and reduce latency by up to 60%. Our method requires no model internals, works across black-box APIs, and remains robust to model updates, making it a practical baseline for real-world LLM deployment.</abstract>
    <identifier type="citekey">soiffer-etal-2025-semantic</identifier>
    <location>
        <url>https://aclanthology.org/2025.emnlp-industry.171/</url>
    </location>
    <part>
        <date>2025-11</date>
        <extent unit="page">
            <start>2499</start>
            <end>2537</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Semantic Agreement Enables Efficient Open-Ended LLM Cascades
%A Soiffer, Duncan
%A Kolawole, Steven
%A Smith, Virginia
%Y Potdar, Saloni
%Y Rojas-Barahona, Lina
%Y Montella, Sebastien
%S Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou (China)
%@ 979-8-89176-333-3
%F soiffer-etal-2025-semantic
%X Cascade systems for open-ended text generation face a fundamental challenge: determining output reliability when generation quality lies on a continuous spectrum, often with multiple valid responses. To address this, we propose _semantic agreement_—meaning-level consensus between ensemble outputs—as a training-free signal for reliable deferral. We show that when diverse model outputs agree semantically, their consensus is a stronger reliability signal than token-level confidence. Evaluated from 500M to 70B-parameter models, semantic cascades improve deferral accuracy, match or surpass target-model quality at 40% of the cost, and reduce latency by up to 60%. Our method requires no model internals, works across black-box APIs, and remains robust to model updates, making it a practical baseline for real-world LLM deployment.
%U https://aclanthology.org/2025.emnlp-industry.171/
%P 2499-2537

Markdown (Informal)

[Semantic Agreement Enables Efficient Open-Ended LLM Cascades](https://aclanthology.org/2025.emnlp-industry.171/) (Soiffer et al., EMNLP 2025)

• Semantic Agreement Enables Efficient Open-Ended LLM Cascades (Soiffer et al., EMNLP 2025)

ACL

• Duncan Soiffer, Steven Kolawole, and Virginia Smith. 2025. Semantic Agreement Enables Efficient Open-Ended LLM Cascades. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track, pages 2499–2537, Suzhou (China). Association for Computational Linguistics.