# Reinforcement Learning Elicits Contextual Learning of Unseen Language Translation

arXiv: https://arxiv.org/abs/2606.06428 · HTML: https://mlsanity.com/paper/2606.06428/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 28 of its 43 references that are in the map.
- First posted on arXiv 4 Jun 2026.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 28 of 28)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks](https://mlsanity.com/paper/1703.03400/) (2017)
- [ZeRO: Memory Optimizations Toward Training Trillion Parameter Models](https://mlsanity.com/paper/1910.02054/) (2019)
- [Tulu 3: Pushing Frontiers in Open Language Model Post-Training](https://mlsanity.com/paper/2411.15124/) (2024)
- [The State and Fate of Linguistic Diversity and Inclusion in the NLP World](https://mlsanity.com/paper/2004.09095/) (2020)
- [What Can Transformers Learn In-Context? A Case Study of Simple Function Classes](https://mlsanity.com/paper/2208.01066/) (2022)
- [MetaICL: Learning to Learn In Context](https://mlsanity.com/paper/2110.15943/) (2021)
- [Meta-learning via Language Model In-context Tuning](https://mlsanity.com/paper/2110.07814/) (2021)
- [Meta-Learning for Low-Resource Neural Machine Translation](https://mlsanity.com/paper/1808.08437/) (2018)
- [BLOOM+1: Adding Language Support to BLOOM for Zero-Shot Prompting](https://mlsanity.com/paper/2212.09535/) (2022)
- [A Benchmark for Learning to Translate a New Language from One Grammar Book](https://mlsanity.com/paper/2309.16575/) (2023)
- [Dictionary-based Phrase-level Prompting of Large Language Models for Machine Translation](https://mlsanity.com/paper/2302.07856/) (2023)
- [MT-R1-Zero: Advancing LLM-based Machine Translation via R1-Zero-like Reinforcement Learning](https://mlsanity.com/paper/2504.10160/) (2025)
- [Improving Machine Translation with Human Feedback: An Exploration of Quality Estimation as a Reward Model](https://mlsanity.com/paper/2401.12873/) (2024)
- [Teaching Large Language Models an Unseen Language on the Fly](https://mlsanity.com/paper/2402.19167/) (2024)
- [LINGOLY: A Benchmark of Olympiad-Level Linguistic Reasoning Puzzles in Low-Resource and Extinct Languages](https://mlsanity.com/paper/2406.06196/) (2024)
- [Can LLMs Really Learn to Translate a Low-Resource Language from One Grammar Book?](https://mlsanity.com/paper/2409.19151/) (2024)
- [Understanding In-Context Machine Translation for Low-Resource Languages: A Case Study on Manchu](https://mlsanity.com/paper/2502.11862/) (2025)
- [SSR-Zero: Simple Self-Rewarding Reinforcement Learning for Machine Translation](https://mlsanity.com/paper/2505.16637/) (2025)
- [Back to School: Translation Using Grammar Books](https://mlsanity.com/paper/2410.15263/) (2024)
- [LLM-Assisted Rule Based Machine Translation for Low/No-Resource Languages](https://mlsanity.com/paper/2405.08997/) (2024)
- [Explicit Learning and the LLM in Machine Translation](https://mlsanity.com/paper/2503.09454/) (2025)
- [Expanding the WMT24++ Benchmark with Rumantsch Grischun, Sursilvan, Sutsilvan, Surmiran, Puter, and Vallader](https://mlsanity.com/paper/2509.03148/) (2025)
- [Improving Low-Resource Machine Translation via Round-Trip Reinforcement Learning](https://mlsanity.com/paper/2601.12535/) (2026)
- [Translation Asymmetry in LLMs as a Data Augmentation Factor: A Case Study for 6 Romansh Language Varieties](https://mlsanity.com/paper/2603.25489/) (2026)
- [Improving Low-Resource Translation with Dictionary-Guided Fine-Tuning and RL: A Spanish-to-Wayuunaiki Study](https://mlsanity.com/paper/2508.19481/) (2025)
- [RUMLEM: A Dictionary-Based Lemmatizer for Romansh](https://mlsanity.com/paper/2604.11233/) (2026)

## Abstract (arXiv)

Prior work has shown that large language models (LLMs) can translate unseen or low-resource languages by undergoing continued training or even by encoding a grammar book in their context. However, both methods typically overfit specific languages, with limited zero-shot transfer at test time. To translate extremely low-resource languages at scale, we argue that LLMs must acquire the meta-skill of utilizing in-context linguistic knowledge rather than memorizing specific languages. In this paper, we propose a reinforcement learning (RL) approach to unseen language translation given rich linguistic context, using a surface-level translation metric (chrF) as the reward. Empirically, despite the lightweight reward, our RL-trained models effectively extract and apply relevant linguistic information from the provided context, leading to better translations on completely unseen languages than in-context learning or supervised fine-tuning. Our analyses suggest that outcome-based RL can extend beyond conventional reasoning tasks like math and coding to serve as a recipe for language learning from context.