<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="tadivalasa-2026-retrieve">
    <titleInfo>
        <title>Retrieve, Reason, Verify: A Modular Multi-Model System for DocSem</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Pranavi</namePart>
        <namePart type="family">Tadivalasa</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>The Proceedings of Workshop on Document Intelligence and Understanding</title>
        </titleInfo>
        <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">Amit</namePart>
            <namePart type="family">Agarwal</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Vivek</namePart>
            <namePart type="family">Gupta</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Yova</namePart>
            <namePart type="family">Kementchedjhieva</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Aparna</namePart>
            <namePart type="family">Garimella</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Yufang</namePart>
            <namePart type="family">Hou</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">979-8-89176-477-4</identifier>
    </relatedItem>
    <abstract>DocSem requires systems to answer quantitative questions over document images while returning the evidence supporting each prediction. We present Retrieve, Reason, Verify, a modular system that separates document transcription, evidence localization, numerical reasoning, deterministic execution, and selective multi-model verification. Under controlled evidence grounding on 217 validation instances, Qwen3.8-Max achieves 98.62% Joint Exact Accuracy. Combining Qwen3.8-Max and GPT-5.6 Sol as independent solvers, with Gemini 3.1 Pro invoked only when their numerical predictions disagree, improves validation performance to 99.08%. On the 1,730-instance hidden test set, which exhibits more challenging visual and grounding conditions, including degraded document quality, non-uniform evidence identifiers, and multiple plausible evidence passages, the complete end-to-end pipeline achieves 32.7% Joint Exact Accuracy. A manual audit of 50 hidden-test instances finds that the system-selected evidence differs from the manually identified evidence in 34 cases, whereas only two cases produce different numerical answers when the evidence agrees. These observations suggest that, within the audited sample, errors are concentrated more strongly in document grounding than in downstream numerical computation once the relevant evidence has been established.</abstract>
    <identifier type="citekey">tadivalasa-2026-retrieve</identifier>
    <location>
        <url>https://aclanthology.org/2026.docinsights-1.33/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>465</start>
            <end>476</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Retrieve, Reason, Verify: A Modular Multi-Model System for DocSem
%A Tadivalasa, Pranavi
%Y T.Y.S.S., Santosh
%Y Agarwal, Amit
%Y Gupta, Vivek
%Y Kementchedjhieva, Yova
%Y Garimella, Aparna
%Y Hou, Yufang
%S The Proceedings of Workshop on Document Intelligence and Understanding
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-477-4
%F tadivalasa-2026-retrieve
%X DocSem requires systems to answer quantitative questions over document images while returning the evidence supporting each prediction. We present Retrieve, Reason, Verify, a modular system that separates document transcription, evidence localization, numerical reasoning, deterministic execution, and selective multi-model verification. Under controlled evidence grounding on 217 validation instances, Qwen3.8-Max achieves 98.62% Joint Exact Accuracy. Combining Qwen3.8-Max and GPT-5.6 Sol as independent solvers, with Gemini 3.1 Pro invoked only when their numerical predictions disagree, improves validation performance to 99.08%. On the 1,730-instance hidden test set, which exhibits more challenging visual and grounding conditions, including degraded document quality, non-uniform evidence identifiers, and multiple plausible evidence passages, the complete end-to-end pipeline achieves 32.7% Joint Exact Accuracy. A manual audit of 50 hidden-test instances finds that the system-selected evidence differs from the manually identified evidence in 34 cases, whereas only two cases produce different numerical answers when the evidence agrees. These observations suggest that, within the audited sample, errors are concentrated more strongly in document grounding than in downstream numerical computation once the relevant evidence has been established.
%U https://aclanthology.org/2026.docinsights-1.33/
%P 465-476

Markdown (Informal)

[Retrieve, Reason, Verify: A Modular Multi-Model System for DocSem](https://aclanthology.org/2026.docinsights-1.33/) (Tadivalasa, DocInsights 2026)

• Retrieve, Reason, Verify: A Modular Multi-Model System for DocSem (Tadivalasa, DocInsights 2026)

ACL

• Pranavi Tadivalasa. 2026. Retrieve, Reason, Verify: A Modular Multi-Model System for DocSem. In The Proceedings of Workshop on Document Intelligence and Understanding, pages 465–476, Budapest, Hungary. Association for Computational Linguistics.