# Lost in Literalism: How Supervised Training Shapes Translationese in LLMs

arXiv: https://arxiv.org/abs/2503.04369 · HTML: https://mlsanity.com/paper/2503.04369/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 31 of its 57 references that are in the map.
- First posted on arXiv 6 Mar 2025.

## Builds on (most cited first, 30 of 31)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Models](https://mlsanity.com/paper/2403.13372/) (2024)
- [No Language Left Behind: Scaling Human-Centered Machine Translation](https://mlsanity.com/paper/2207.04672/) (2022)
- [Stanza: A Python Natural Language Processing Toolkit for Many Human Languages](https://mlsanity.com/paper/2003.07082/) (2020)
- [Large Language Models Are State-of-the-Art Evaluators of Translation Quality](https://mlsanity.com/paper/2302.14520/) (2023)
- [Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation](https://mlsanity.com/paper/2401.08417/) (2024)
- [How Good Are GPT Models at Machine Translation? A Comprehensive Evaluation](https://mlsanity.com/paper/2302.09210/) (2023)
- [CometKiwi: IST-Unbabel 2022 Submission for the Quality Estimation Shared Task](https://mlsanity.com/paper/2209.06243/) (2022)
- [Is ChatGPT A Good Translator? Yes With GPT-4 As The Engine](https://mlsanity.com/paper/2301.08745/) (2023)
- [Achieving Human Parity on Automatic Chinese to English News Translation](https://mlsanity.com/paper/1803.05567/) (2018)
- [Dictionary-based Phrase-level Prompting of Large Language Models for Machine Translation](https://mlsanity.com/paper/2302.07856/) (2023)
- [TIM: Teaching Large Language Models to Translate with Comparison](https://mlsanity.com/paper/2307.04408/) (2023)
- [The Effect of Translationese in Machine Translation Test Sets](https://mlsanity.com/paper/1906.08069/) (2019)
- [TEaR: Improving LLM-based Machine Translation with Systematic Self-Refinement](https://mlsanity.com/paper/2402.16379/) (2024)
- [Translationese as a Language in "Multilingual" NMT](https://mlsanity.com/paper/1911.03823/) (2019)
- [On Compositional Generalization of Neural Machine Translation](https://mlsanity.com/paper/2105.14802/) (2021)
- [Do GPTs Produce Less Literal Translations?](https://mlsanity.com/paper/2305.16806/) (2023)
- [Post-editese: an Exacerbated Translationese](https://mlsanity.com/paper/1907.00900/) (2019)
- [Guiding Large Language Models to Post-Edit Machine Translation with Error Annotations](https://mlsanity.com/paper/2404.07851/) (2024)
- [The Fine-Tuning Paradox: Boosting Translation Quality Without Sacrificing LLM Abilities](https://mlsanity.com/paper/2405.20089/) (2024)
- [Using Monolingual Data in Neural Machine Translation: a Systematic Study](https://mlsanity.com/paper/1903.11437/) (2019)
- [Understanding Translationese in Cross-Lingual Summarization](https://mlsanity.com/paper/2212.07220/) (2022)
- [Comparing Feature-Engineering and Feature-Learning Approaches for Multilingual Translationese Classification](https://mlsanity.com/paper/2109.07604/) (2021)
- [Original or Translated? A Causal Analysis of the Impact of Translationese on Machine Translation Performance](https://mlsanity.com/paper/2205.02293/) (2022)
- [Lost in Translationese? Reducing Translation Effect Using Abstract Meaning Representation](https://mlsanity.com/paper/2304.11501/) (2023)
- [Translating away Translationese without Parallel Data](https://mlsanity.com/paper/2310.18830/) (2023)
- [Is Preference Alignment Always the Best Option to Enhance LLM-Based Translation? An Empirical Analysis](https://mlsanity.com/paper/2409.20059/) (2024)

## Built on by (most cited first, 11 of 11)

- [Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought](https://mlsanity.com/paper/2510.04230/) (2025)
- [The Art of Asking: Multilingual Prompt Optimization for Synthetic Data](https://mlsanity.com/paper/2510.19806/) (2025)
- [Preferences for Idiomatic Language are Acquired Slowly -- and Forgotten Quickly: A Case Study on Swedish](https://mlsanity.com/paper/2602.03484/) (2026)
- [NeoBabel: A Multilingual Open Tower for Visual Generation](https://mlsanity.com/paper/2507.06137/) (2025)
- [GAIA-v2-LILT: Multilingual Adaptation of Agent Benchmark beyond Translation](https://mlsanity.com/paper/2604.24929/) (2026)
- [Translationese-index: Using Likelihood Ratios for Graded and Generalizable Measurement of Translationese](https://mlsanity.com/paper/2507.12260/) (2025)
- [Benchmarking Concept-Spilling Across Languages in LLMs](https://mlsanity.com/paper/2601.12549/) (2026)
- [A Dataset for Probing Translationese Preferences in English-to-Swedish Translation](https://mlsanity.com/paper/2603.08450/) (2026)
- [Testing the Deliteralization Hypothesis in Human and Machine Translation](https://mlsanity.com/paper/2605.25686/) (2026)
- [An Investigation of Translationese in the Generations of Multilingual Large Language Models](https://mlsanity.com/paper/2608.17399/) (2026)
- [EuroAlpaca: Task-Preserving Localisation of Instruction Data for European Languages](https://mlsanity.com/paper/2609.05043/) (2026)

## Abstract (arXiv)

Large language models (LLMs) have achieved remarkable success in machine translation, demonstrating impressive performance across diverse languages. However, translationese, characterized by overly literal and unnatural translations, remains a persistent challenge in LLM-based translation systems. Despite their pre-training on vast corpora of natural utterances, LLMs exhibit translationese errors and generate unexpected unnatural translations, stemming from biases introduced during supervised fine-tuning (SFT). In this work, we systematically evaluate the prevalence of translationese in LLM-generated translations and investigate its roots during supervised training. We introduce methods to mitigate these biases, including polishing golden references and filtering unnatural training instances. Empirical evaluations demonstrate that these approaches significantly reduce translationese while improving translation naturalness, validated by human evaluations and automatic metrics. Our findings highlight the need for training-aware adjustments to optimize LLM translation outputs, paving the way for more fluent and target-language-consistent translations. We release the data and code at https://github.com/yafuly/LLM_Translationese.