@inproceedings{steingrimsson-etal-2026-drag,
title = "What a {DRAG} (it is being small): The {AMI} Submission to the {WMT} 2026 General Translation Shared Task",
author = {Steingr{\'i}msson, Steinth{\'o}r and
Th{\'o}rdharson, Sveinbj{\"o}rn and
Dadason, J{\'o}n},
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.79/",
pages = "1414--1424",
ISBN = "979-8-89176-474-3",
abstract = "We describe the AMI team{'}s submission to the WMT 2026 General Translation shared task, participating in the English-Icelandic translation direction. Building on our 2025 submission, which relied on a 3B-parameter model and substantial rule-based post-processing, this year we aim to reduce that dependency and rely more on the translation model itself. We continually pre-train a Qwen3-4B checkpoint on a mixture of Icelandic, English, code, math, and structured data, then fine-tune it with LoRA on a retrieval-augmented, domain-conditioned instruction-tuning dataset covering five domains. At both training and inference time, source sentences are enriched with dictionary entries and back-translated example translations similar to the sentence being translated, retrieved by a dedicated RAG server, and served through a self-healing inference client that retries and validates generations. We report COMET and chrF++ scores comparing this system against last year{'}s submission, a larger Llama-3.1-8B model adapted with the same recipe, and an instruction-tuned Qwen3-4B baseline, alongside ablations that isolate the contribution of each part of the retrieval-augmented prompt."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="steingrimsson-etal-2026-drag">
<titleInfo>
<title>What a DRAG (it is being small): The AMI Submission to the WMT 2026 General Translation Shared Task</title>
</titleInfo>
<name type="personal">
<namePart type="given">Steinthór</namePart>
<namePart type="family">Steingrímsson</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sveinbjörn</namePart>
<namePart type="family">Thórdharson</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jón</namePart>
<namePart type="family">Dadason</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>We describe the AMI team’s submission to the WMT 2026 General Translation shared task, participating in the English-Icelandic translation direction. Building on our 2025 submission, which relied on a 3B-parameter model and substantial rule-based post-processing, this year we aim to reduce that dependency and rely more on the translation model itself. We continually pre-train a Qwen3-4B checkpoint on a mixture of Icelandic, English, code, math, and structured data, then fine-tune it with LoRA on a retrieval-augmented, domain-conditioned instruction-tuning dataset covering five domains. At both training and inference time, source sentences are enriched with dictionary entries and back-translated example translations similar to the sentence being translated, retrieved by a dedicated RAG server, and served through a self-healing inference client that retries and validates generations. We report COMET and chrF++ scores comparing this system against last year’s submission, a larger Llama-3.1-8B model adapted with the same recipe, and an instruction-tuned Qwen3-4B baseline, alongside ablations that isolate the contribution of each part of the retrieval-augmented prompt.</abstract>
<identifier type="citekey">steingrimsson-etal-2026-drag</identifier>
<location>
<url>https://aclanthology.org/2026.wmt-main.79/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>1414</start>
<end>1424</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T What a DRAG (it is being small): The AMI Submission to the WMT 2026 General Translation Shared Task
%A Steingrímsson, Steinthór
%A Thórdharson, Sveinbjörn
%A Dadason, Jón
%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 steingrimsson-etal-2026-drag
%X We describe the AMI team’s submission to the WMT 2026 General Translation shared task, participating in the English-Icelandic translation direction. Building on our 2025 submission, which relied on a 3B-parameter model and substantial rule-based post-processing, this year we aim to reduce that dependency and rely more on the translation model itself. We continually pre-train a Qwen3-4B checkpoint on a mixture of Icelandic, English, code, math, and structured data, then fine-tune it with LoRA on a retrieval-augmented, domain-conditioned instruction-tuning dataset covering five domains. At both training and inference time, source sentences are enriched with dictionary entries and back-translated example translations similar to the sentence being translated, retrieved by a dedicated RAG server, and served through a self-healing inference client that retries and validates generations. We report COMET and chrF++ scores comparing this system against last year’s submission, a larger Llama-3.1-8B model adapted with the same recipe, and an instruction-tuned Qwen3-4B baseline, alongside ablations that isolate the contribution of each part of the retrieval-augmented prompt.
%U https://aclanthology.org/2026.wmt-main.79/
%P 1414-1424
Markdown (Informal)
[What a DRAG (it is being small): The AMI Submission to the WMT 2026 General Translation Shared Task](https://aclanthology.org/2026.wmt-main.79/) (Steingrímsson et al., WMT 2026)
ACL