TRIVE: A Tri-Fusion Verified Translation Agent

Mengpeng Yang; Tian Xia; Jingxu Yang; Chao Chen; Yabo Sun; Qiang Liu

TRIVE: A Tri-Fusion Verified Translation Agent

Mengpeng Yang, Tian Xia, Jingxu Yang, Chao Chen, Yabo Sun, Qiang Liu

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

For the WMT26 General Machine Translation Task (Unconstrained Track), we propose TRIVE (Tri-Fusion Verified Translation Agent), a model-agnostic multi-agent translation framework. TRIVE organizes translation as a closed loop of Generation–Fusion–Verification–Correction: it first generates candidates in parallel under three complementary strategies—Faithful, Fluent, and Free—then synthesizes a draft that balances semantic fidelity and natural expression, and finally iteratively refines the output via rule-based validation and large language model evaluation. Because each module is implemented primarily via prompting, TRIVE can be directly overlaid onto diverse backbone models without task-specific fine-tuning. Experiments show that overlaying TRIVE on multiple backbones consistently improves instruction-following translation quality. Finally, we submitted Gemini-3.5-Flash + TRIVE as our primary system to WMT26.

Anthology ID:: 2026.wmt-main.84
Volume:: Proceedings of the Eleventh Conference on Machine Translation
Month:: October
Year:: 2026
Address:: Budapest, Hungary
Editors:: Barry Haddow, Tom Kocmi, Philipp Koehn, Christof Monz
Venues:: WMT | WS
SIG:: SIGMT
Publisher:: Association for Computational Linguistics
Note:
Pages:: 1463–1476
Language:
URL:: https://aclanthology.org/2026.wmt-main.84/
DOI:
Bibkey:
Cite (ACL):: Mengpeng Yang, Tian Xia, Jingxu Yang, Chao Chen, Yabo Sun, and Qiang Liu. 2026. TRIVE: A Tri-Fusion Verified Translation Agent. In Proceedings of the Eleventh Conference on Machine Translation, pages 1463–1476, Budapest, Hungary. Association for Computational Linguistics.
Cite (Informal):: TRIVE: A Tri-Fusion Verified Translation Agent (Yang et al., WMT 2026)
Copy Citation:
PDF:: https://aclanthology.org/2026.wmt-main.84.pdf

PDF Cite Search Fix data

Export citation

BibTeX
MODS XML
Endnote
Preformatted

@inproceedings{yang-etal-2026-trive,
    title = "{TRIVE}: A Tri-Fusion Verified Translation Agent",
    author = "Yang, Mengpeng  and
      Xia, Tian  and
      Yang, Jingxu  and
      Chen, Chao  and
      Sun, Yabo  and
      Liu, Qiang",
    editor = "Haddow, Barry  and
      Kocmi, Tom  and
      Koehn, Philipp  and
      Monz, Christof",
    booktitle = "Proceedings of the Eleventh Conference on Machine Translation",
    month = oct,
    year = "2026",
    address = "Budapest, Hungary",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2026.wmt-main.84/",
    pages = "1463--1476",
    ISBN = "979-8-89176-474-3",
    abstract = "For the WMT26 General Machine Translation Task (Unconstrained Track), we propose TRIVE (Tri-Fusion Verified Translation Agent), a model-agnostic multi-agent translation framework. TRIVE organizes translation as a closed loop of Generation{--}Fusion{--}Verification{--}Correction: it first generates candidates in parallel under three complementary strategies{---}Faithful, Fluent, and Free{---}then synthesizes a draft that balances semantic fidelity and natural expression, and finally iteratively refines the output via rule-based validation and large language model evaluation. Because each module is implemented primarily via prompting, TRIVE can be directly overlaid onto diverse backbone models without task-specific fine-tuning. Experiments show that overlaying TRIVE on multiple backbones consistently improves instruction-following translation quality. Finally, we submitted Gemini-3.5-Flash + TRIVE as our primary system to WMT26."
}

Download as File

<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="yang-etal-2026-trive">
    <titleInfo>
        <title>TRIVE: A Tri-Fusion Verified Translation Agent</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Mengpeng</namePart>
        <namePart type="family">Yang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Tian</namePart>
        <namePart type="family">Xia</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jingxu</namePart>
        <namePart type="family">Yang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Chao</namePart>
        <namePart type="family">Chen</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yabo</namePart>
        <namePart type="family">Sun</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Qiang</namePart>
        <namePart type="family">Liu</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 Eleventh Conference on Machine Translation</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Barry</namePart>
            <namePart type="family">Haddow</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Tom</namePart>
            <namePart type="family">Kocmi</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Philipp</namePart>
            <namePart type="family">Koehn</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Christof</namePart>
            <namePart type="family">Monz</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">979-8-89176-474-3</identifier>
    </relatedItem>
    <abstract>For the WMT26 General Machine Translation Task (Unconstrained Track), we propose TRIVE (Tri-Fusion Verified Translation Agent), a model-agnostic multi-agent translation framework. TRIVE organizes translation as a closed loop of Generation–Fusion–Verification–Correction: it first generates candidates in parallel under three complementary strategies—Faithful, Fluent, and Free—then synthesizes a draft that balances semantic fidelity and natural expression, and finally iteratively refines the output via rule-based validation and large language model evaluation. Because each module is implemented primarily via prompting, TRIVE can be directly overlaid onto diverse backbone models without task-specific fine-tuning. Experiments show that overlaying TRIVE on multiple backbones consistently improves instruction-following translation quality. Finally, we submitted Gemini-3.5-Flash + TRIVE as our primary system to WMT26.</abstract>
    <identifier type="citekey">yang-etal-2026-trive</identifier>
    <location>
        <url>https://aclanthology.org/2026.wmt-main.84/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>1463</start>
            <end>1476</end>
        </extent>
    </part>
</mods>
</modsCollection>

Download as File

%0 Conference Proceedings
%T TRIVE: A Tri-Fusion Verified Translation Agent
%A Yang, Mengpeng
%A Xia, Tian
%A Yang, Jingxu
%A Chen, Chao
%A Sun, Yabo
%A Liu, Qiang
%Y Haddow, Barry
%Y Kocmi, Tom
%Y Koehn, Philipp
%Y Monz, Christof
%S Proceedings of the Eleventh Conference on Machine Translation
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-474-3
%F yang-etal-2026-trive
%X For the WMT26 General Machine Translation Task (Unconstrained Track), we propose TRIVE (Tri-Fusion Verified Translation Agent), a model-agnostic multi-agent translation framework. TRIVE organizes translation as a closed loop of Generation–Fusion–Verification–Correction: it first generates candidates in parallel under three complementary strategies—Faithful, Fluent, and Free—then synthesizes a draft that balances semantic fidelity and natural expression, and finally iteratively refines the output via rule-based validation and large language model evaluation. Because each module is implemented primarily via prompting, TRIVE can be directly overlaid onto diverse backbone models without task-specific fine-tuning. Experiments show that overlaying TRIVE on multiple backbones consistently improves instruction-following translation quality. Finally, we submitted Gemini-3.5-Flash + TRIVE as our primary system to WMT26.
%U https://aclanthology.org/2026.wmt-main.84/
%P 1463-1476

Download as File

Markdown (Informal)

[TRIVE: A Tri-Fusion Verified Translation Agent](https://aclanthology.org/2026.wmt-main.84/) (Yang et al., WMT 2026)

TRIVE: A Tri-Fusion Verified Translation Agent (Yang et al., WMT 2026)

ACL

Mengpeng Yang, Tian Xia, Jingxu Yang, Chao Chen, Yabo Sun, and Qiang Liu. 2026. TRIVE: A Tri-Fusion Verified Translation Agent. In Proceedings of the Eleventh Conference on Machine Translation, pages 1463–1476, Budapest, Hungary. Association for Computational Linguistics.