Auditable Multi-View Hosted Multimodal Judging for AstroCLIMB

Kewen Zhu; Junyuan Tan

Auditable Multi-View Hosted Multimodal Judging for AstroCLIMB

Correct Metadata for

Use this form to create a GitHub issue with structured data describing the correction. You will need a GitHub account. Once you create that issue, the correction will be reviewed by a staff member.

⚠️ Mobile Users: Submitting this form to create a new issue will only work with github.com, not the GitHub Mobile app.

Important: The Anthology treat PDFs as authoritative. Please use this form only to correct data that is out of line with the PDF. See our corrections guidelines if you need to change the PDF.

Title Adjust the title. Retain tags such as <fixed-case>.

Authors Adjust author names and order to match the PDF.

Abstract Correct abstract if needed. Retain XML formatting tags such as <tex-math>. You may use ... for bold, ... for italic, ... for underline, <sc>...</sc> for small-caps, <tt>...<tt> for typewriter text, <url>...</url> for URLs, <a href=...> for hyperlinks, and <par/> for paragraph breaks.

Verification against PDF Ensure that the new title/authors match the snapshot below. (If there is no snapshot or it is too small, consult the PDF.)

Authors concatenated from the text boxes above:

ALL author names match the snapshot above—including middle initials, hyphens, and accents.

Abstract

We study AstroCLIMB, a four-way relationship-classification task over pairs of astronomy figure/caption endpoints. Starting from a leakage-audited lexical and low-level image-statistics baseline (M7), we evaluate a hosted multimodal judge under a fixed, auditable protocol. GPT-5.6 Luna receives only the two endpoint contents and produces integer class probabilities for three views in both endpoint directions. On 2,000 frozen validation pairs, all 12,000 calls succeeded. Equal-weight fusion reaches macro-F1 0.67, versus M7’s 0.40 (paired component-bootstrap delta +0.27, 95% CI [+0.24,+0.29]); Luna-only reaches 0.68. These are conditional local-validation results, not an official Luna leaderboard score. We therefore report an auditable system study without claiming paper-level generalization.

Anthology ID:: 2026.wasp-1.5
Volume:: Proceedings of the 4th Workshop on Artificial Intelligence for Scientific Publications
Month:: November
Year:: 2026
Address:: Hengqin, Zhuhai, China
Editors:: Alberto Accomazzi, Tirthankar Ghosal, Felix Grezes, Kelly Lockhart, Atilla Kaan Alkan
Venues:: WASP | WS
SIG:
Publisher:: Association for Computational Linguistics
Note:
Pages:: 34–37
Language:
URL:: https://aclanthology.org/2026.wasp-1.5/
DOI:
Bibkey:
Cite (ACL):: Kewen Zhu and Junyuan Tan. 2026. Auditable Multi-View Hosted Multimodal Judging for AstroCLIMB. In Proceedings of the 4th Workshop on Artificial Intelligence for Scientific Publications, pages 34–37, Hengqin, Zhuhai, China. Association for Computational Linguistics.
Cite (Informal):: Auditable Multi-View Hosted Multimodal Judging for AstroCLIMB (Zhu & Tan, WASP 2026)
Copy Citation:
PDF:: https://aclanthology.org/2026.wasp-1.5.pdf

PDF Cite Search Fix data

Export citation

BibTeX
MODS XML
Endnote
Preformatted

@inproceedings{zhu-tan-2026-auditable,
    title = "Auditable Multi-View Hosted Multimodal Judging for {A}stro{CLIMB}",
    author = "Zhu, Kewen  and
      Tan, Junyuan",
    editor = "Accomazzi, Alberto  and
      Ghosal, Tirthankar  and
      Grezes, Felix  and
      Lockhart, Kelly  and
      Alkan, Atilla Kaan",
    booktitle = "Proceedings of the 4th Workshop on Artificial Intelligence for Scientific Publications",
    month = nov,
    year = "2026",
    address = "Hengqin, Zhuhai, China",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2026.wasp-1.5/",
    pages = "34--37",
    ISBN = "979-8-89176-456-9",
    abstract = "We study AstroCLIMB, a four-way relationship-classification task over pairs of astronomy figure/caption endpoints. Starting from a leakage-audited lexical and low-level image-statistics baseline (M7), we evaluate a hosted multimodal judge under a fixed, auditable protocol. GPT-5.6 Luna receives only the two endpoint contents and produces integer class probabilities for three views in both endpoint directions. On 2,000 frozen validation pairs, all 12,000 calls succeeded. Equal-weight fusion reaches macro-F1 0.67, versus M7{'}s 0.40 (paired component-bootstrap delta +0.27, 95{\%} CI [+0.24,+0.29]); Luna-only reaches 0.68. These are conditional local-validation results, not an official Luna leaderboard score. We therefore report an auditable system study without claiming paper-level generalization."
}

Download as File

<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhu-tan-2026-auditable">
    <titleInfo>
        <title>Auditable Multi-View Hosted Multimodal Judging for AstroCLIMB</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Kewen</namePart>
        <namePart type="family">Zhu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Junyuan</namePart>
        <namePart type="family">Tan</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 4th Workshop on Artificial Intelligence for Scientific Publications</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Alberto</namePart>
            <namePart type="family">Accomazzi</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Tirthankar</namePart>
            <namePart type="family">Ghosal</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Felix</namePart>
            <namePart type="family">Grezes</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Kelly</namePart>
            <namePart type="family">Lockhart</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Atilla</namePart>
            <namePart type="given">Kaan</namePart>
            <namePart type="family">Alkan</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Hengqin, Zhuhai, China</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-456-9</identifier>
    </relatedItem>
    <abstract>We study AstroCLIMB, a four-way relationship-classification task over pairs of astronomy figure/caption endpoints. Starting from a leakage-audited lexical and low-level image-statistics baseline (M7), we evaluate a hosted multimodal judge under a fixed, auditable protocol. GPT-5.6 Luna receives only the two endpoint contents and produces integer class probabilities for three views in both endpoint directions. On 2,000 frozen validation pairs, all 12,000 calls succeeded. Equal-weight fusion reaches macro-F1 0.67, versus M7’s 0.40 (paired component-bootstrap delta +0.27, 95% CI [+0.24,+0.29]); Luna-only reaches 0.68. These are conditional local-validation results, not an official Luna leaderboard score. We therefore report an auditable system study without claiming paper-level generalization.</abstract>
    <identifier type="citekey">zhu-tan-2026-auditable</identifier>
    <location>
        <url>https://aclanthology.org/2026.wasp-1.5/</url>
    </location>
    <part>
        <date>2026-11</date>
        <extent unit="page">
            <start>34</start>
            <end>37</end>
        </extent>
    </part>
</mods>
</modsCollection>

Download as File

%0 Conference Proceedings
%T Auditable Multi-View Hosted Multimodal Judging for AstroCLIMB
%A Zhu, Kewen
%A Tan, Junyuan
%Y Accomazzi, Alberto
%Y Ghosal, Tirthankar
%Y Grezes, Felix
%Y Lockhart, Kelly
%Y Alkan, Atilla Kaan
%S Proceedings of the 4th Workshop on Artificial Intelligence for Scientific Publications
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, Zhuhai, China
%@ 979-8-89176-456-9
%F zhu-tan-2026-auditable
%X We study AstroCLIMB, a four-way relationship-classification task over pairs of astronomy figure/caption endpoints. Starting from a leakage-audited lexical and low-level image-statistics baseline (M7), we evaluate a hosted multimodal judge under a fixed, auditable protocol. GPT-5.6 Luna receives only the two endpoint contents and produces integer class probabilities for three views in both endpoint directions. On 2,000 frozen validation pairs, all 12,000 calls succeeded. Equal-weight fusion reaches macro-F1 0.67, versus M7’s 0.40 (paired component-bootstrap delta +0.27, 95% CI [+0.24,+0.29]); Luna-only reaches 0.68. These are conditional local-validation results, not an official Luna leaderboard score. We therefore report an auditable system study without claiming paper-level generalization.
%U https://aclanthology.org/2026.wasp-1.5/
%P 34-37

Download as File

Markdown (Informal)

[Auditable Multi-View Hosted Multimodal Judging for AstroCLIMB](https://aclanthology.org/2026.wasp-1.5/) (Zhu & Tan, WASP 2026)

Auditable Multi-View Hosted Multimodal Judging for AstroCLIMB (Zhu & Tan, WASP 2026)

ACL

Kewen Zhu and Junyuan Tan. 2026. Auditable Multi-View Hosted Multimodal Judging for AstroCLIMB. In Proceedings of the 4th Workshop on Artificial Intelligence for Scientific Publications, pages 34–37, Hengqin, Zhuhai, China. Association for Computational Linguistics.