@inproceedings{chen-etal-2026-translation,
title = "Translation or Recitation? Calibrating Evaluation Scores for Machine Translation of Extremely Low-Resource Languages",
author = "Chen, Danlu and
He, Ka Sing and
Tian, Jiahe and
Xiao, Chenghao and
Wu, Zhaofeng and
Berg-Kirkpatrick, Taylor and
Shi, Freda",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 2: Short Papers)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-short.29/",
pages = "350--363",
ISBN = "979-8-89176-391-3",
abstract = "The landscape of extremely low-resource machine translation (MT) is characterized by perplexing variability in reported performance, often making results across different language pairs difficult to contextualize. For researchers focused on specific language groups{---}such as ancient languages{---}it is nearly impossible to determine if breakthroughs reported in other contexts (e.g., African or American languages) result from superior methodologies or are merely artifacts of benchmark collection. To address this, we introduce the \textbf{FRED Difficulty Metrics}{---}\textit{Fertility Ratio (F)}, \textit{Retrieval Proxy} ($R$) \textit{Pre-training Exposure} ($E$) and \textit{Corpus Diversity} ($D$) {---}that serve as dataset-intrinsic metrics to contextualize reported scores. Our findings reveal that a significant portion of result variability is explained by train-test overlap and pre-training exposure rather than model capability. Additionally, we identify that underperforming XLR languages{---}particularly extinct and non-Latin indigenous languages{---}suffer from poor tokenization coverage (high token fertility), highlighting structural limitations of transfer learning for languages outside pre-trained models' representation space. By providing these indices alongside performance scores, we enable more transparent evaluation of cross-lingual transfer and provide a more reliable foundation for the XLR MT community."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="chen-etal-2026-translation">
<titleInfo>
<title>Translation or Recitation? Calibrating Evaluation Scores for Machine Translation of Extremely Low-Resource Languages</title>
</titleInfo>
<name type="personal">
<namePart type="given">Danlu</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ka</namePart>
<namePart type="given">Sing</namePart>
<namePart type="family">He</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiahe</namePart>
<namePart type="family">Tian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chenghao</namePart>
<namePart type="family">Xiao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhaofeng</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Taylor</namePart>
<namePart type="family">Berg-Kirkpatrick</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Freda</namePart>
<namePart type="family">Shi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-391-3</identifier>
</relatedItem>
<abstract>The landscape of extremely low-resource machine translation (MT) is characterized by perplexing variability in reported performance, often making results across different language pairs difficult to contextualize. For researchers focused on specific language groups—such as ancient languages—it is nearly impossible to determine if breakthroughs reported in other contexts (e.g., African or American languages) result from superior methodologies or are merely artifacts of benchmark collection. To address this, we introduce the FRED Difficulty Metrics—Fertility Ratio (F), Retrieval Proxy (R) Pre-training Exposure (E) and Corpus Diversity (D) —that serve as dataset-intrinsic metrics to contextualize reported scores. Our findings reveal that a significant portion of result variability is explained by train-test overlap and pre-training exposure rather than model capability. Additionally, we identify that underperforming XLR languages—particularly extinct and non-Latin indigenous languages—suffer from poor tokenization coverage (high token fertility), highlighting structural limitations of transfer learning for languages outside pre-trained models’ representation space. By providing these indices alongside performance scores, we enable more transparent evaluation of cross-lingual transfer and provide a more reliable foundation for the XLR MT community.</abstract>
<identifier type="citekey">chen-etal-2026-translation</identifier>
<location>
<url>https://aclanthology.org/2026.acl-short.29/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>350</start>
<end>363</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Translation or Recitation? Calibrating Evaluation Scores for Machine Translation of Extremely Low-Resource Languages
%A Chen, Danlu
%A He, Ka Sing
%A Tian, Jiahe
%A Xiao, Chenghao
%A Wu, Zhaofeng
%A Berg-Kirkpatrick, Taylor
%A Shi, Freda
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-391-3
%F chen-etal-2026-translation
%X The landscape of extremely low-resource machine translation (MT) is characterized by perplexing variability in reported performance, often making results across different language pairs difficult to contextualize. For researchers focused on specific language groups—such as ancient languages—it is nearly impossible to determine if breakthroughs reported in other contexts (e.g., African or American languages) result from superior methodologies or are merely artifacts of benchmark collection. To address this, we introduce the FRED Difficulty Metrics—Fertility Ratio (F), Retrieval Proxy (R) Pre-training Exposure (E) and Corpus Diversity (D) —that serve as dataset-intrinsic metrics to contextualize reported scores. Our findings reveal that a significant portion of result variability is explained by train-test overlap and pre-training exposure rather than model capability. Additionally, we identify that underperforming XLR languages—particularly extinct and non-Latin indigenous languages—suffer from poor tokenization coverage (high token fertility), highlighting structural limitations of transfer learning for languages outside pre-trained models’ representation space. By providing these indices alongside performance scores, we enable more transparent evaluation of cross-lingual transfer and provide a more reliable foundation for the XLR MT community.
%U https://aclanthology.org/2026.acl-short.29/
%P 350-363
Markdown (Informal)
[Translation or Recitation? Calibrating Evaluation Scores for Machine Translation of Extremely Low-Resource Languages](https://aclanthology.org/2026.acl-short.29/) (Chen et al., ACL 2026)
ACL
- Danlu Chen, Ka Sing He, Jiahe Tian, Chenghao Xiao, Zhaofeng Wu, Taylor Berg-Kirkpatrick, and Freda Shi. 2026. Translation or Recitation? Calibrating Evaluation Scores for Machine Translation of Extremely Low-Resource Languages. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 350–363, San Diego, California, United States. Association for Computational Linguistics.