# Exploring Performance Variations in Finetuned Translators of Ultra-Low Resource Languages: Do Linguistic Differences Matter?

arXiv: https://arxiv.org/abs/2511.22482 · HTML: https://mlsanity.com/paper/2511.22482/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 33 references that are in the map.
- First posted on arXiv 27 Nov 2025.

## Builds on (most cited first, 9 of 9)

- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [On the Opportunities and Risks of Foundation Models](https://mlsanity.com/paper/2108.07258/) (2021)
- [No Language Left Behind: Scaling Human-Centered Machine Translation](https://mlsanity.com/paper/2207.04672/) (2022)
- [Facebook FAIR's WMT19 News Translation Task Submission](https://mlsanity.com/paper/1907.06616/) (2019)
- [Fixing MoE Over-Fitting on Low-Resource Languages in Multilingual Machine Translation](https://mlsanity.com/paper/2212.07571/) (2022)
- [Synthetic Pre-Training Tasks for Neural Machine Translation](https://mlsanity.com/paper/2212.09864/) (2022)
- [Toucan: Many-to-Many Translation for 150 African Language Pairs](https://mlsanity.com/paper/2407.04796/) (2024)
- [Sentence-level Aggregation of Lexical Metrics Correlates Stronger with Human Judgements than Corpus-level Aggregation](https://mlsanity.com/paper/2407.12832/) (2024)
- [Harnessing the Power of Artificial Intelligence to Vitalize Endangered Indigenous Languages: Technologies and Experiences](https://mlsanity.com/paper/2407.12620/) (2024)

## Built on by (most cited first, 1 of 1)

- [The Constitution Speaks Nheengatu: An Open MT System and Reproducible Corpus Pipeline for the Amazonian L'ingua Geral](https://mlsanity.com/paper/2610.14760/) (2026)

## Abstract (arXiv)

Finetuning pre-trained language models with small amounts of data is a commonly-used method to create translators for ultra-low resource languages such as endangered Indigenous languages. However, previous works have reported substantially different performances with translators created using similar methodology and data. In this work we systematically explored possible causes of the performance difference, aiming to determine whether it was a product of different cleaning procedures, limitations of the pre-trained models, the size of the base model, or the size of the training dataset, studying both directions of translation. Our studies, using two Brazilian Indigenous languages, related but with significant structural linguistic characteristics, indicated none or very limited influence from those training factors, suggesting differences between languages may play a significant role in the ability to produce translators by fine-tuning pre-trained models.