# Aligning Large Language Models to Low-Resource Languages through LLM-Based Selective Translation: A Systematic Study

arXiv: https://arxiv.org/abs/2507.14304 · HTML: https://mlsanity.com/paper/2507.14304/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in LLM preference optimization methods, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 14 of its 26 references that are in the map.
- First posted on arXiv 18 Jul 2025.

## Placed in

- LLM alignment / RLHF / preference optimisation / [LLM preference optimization methods](https://mlsanity.com/alignment/llm-preference-optimization-methods/)

## Builds on (most cited first, 14 of 14)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [The State and Fate of Linguistic Diversity and Inclusion in the NLP World](https://mlsanity.com/paper/2004.09095/) (2020)
- [Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluation](https://mlsanity.com/paper/2412.03304/) (2024)
- [Llama-Nemotron: Efficient Reasoning Models](https://mlsanity.com/paper/2505.00949/) (2025)
- [Multilingual LLMs are Better Cross-lingual In-context Learners with Alignment](https://mlsanity.com/paper/2305.05940/) (2023)
- [LLMs Are Few-Shot In-Context Low-Resource Language Learners](https://mlsanity.com/paper/2403.16512/) (2024)
- [LLMs Beyond English: Scaling the Multilingual Capability of LLMs with Cross-Lingual Feedback](https://mlsanity.com/paper/2406.01771/) (2024)
- [NeMo-Aligner: Scalable Toolkit for Efficient Model Alignment](https://mlsanity.com/paper/2405.01481/) (2024)
- [MILU: A Multi-task Indic Language Understanding Benchmark](https://mlsanity.com/paper/2411.02538/) (2024)
- [Adapting Multilingual LLMs to Low-Resource Languages using Continued Pre-training and Synthetic Corpus](https://mlsanity.com/paper/2410.14815/) (2024)
- [InstructAlign: High-and-Low Resource Language Alignment via Continual Crosslingual Instruction Tuning](https://mlsanity.com/paper/2305.13627/) (2023)
- [X-Instruction: Aligning Language Model in Low-resource Languages with Self-curated Cross-lingual Instructions](https://mlsanity.com/paper/2405.19744/) (2024)
- [On Limitations of LLM as Annotator for Low Resource Languages](https://mlsanity.com/paper/2411.17637/) (2024)
- [Llama-3-Nanda-10B-Chat: An Open Generative Large Language Model for Hindi](https://mlsanity.com/paper/2504.06011/) (2025)

## Built on by (most cited first, 4 of 4)

- [CultureGuard: Towards Culturally-Aware Dataset and Guard Model for Multilingual Safety Applications](https://mlsanity.com/paper/2508.01710/) (2025)
- [No One Fits All: From Fixed Prompting to Learned Routing in Multilingual LLMs](https://mlsanity.com/paper/2604.16937/) (2026)
- [LLM Safety Alignment in Low-Resource Languages: A Systematic Literature Review](https://mlsanity.com/paper/2608.14626/) (2026)
- [Pak3H: Evaluating the Cost of Cultural Mismatch in LLM Alignment with a Human-Contextualized Urdu Benchmark](https://mlsanity.com/paper/2608.30065/) (2026)

## Abstract (arXiv)

Multilingual large language models (LLMs) often demonstrate a performance gap between English and non-English languages, particularly in low-resource settings. Aligning these models to low-resource languages is essential yet challenging due to limited high-quality data. While English alignment datasets are readily available, curating equivalent data in other languages is expensive and time-consuming. A common workaround is to translate existing English alignment data; however, standard translation techniques often fail to preserve critical elements such as code, mathematical expressions, and structured formats like JSON. In this work, we investigate LLM-based selective translation, a technique that selectively translates only the translatable parts of a text while preserving non-translatable content and sentence structure. We conduct a systematic study to explore key questions around this approach, including its effectiveness compared to vanilla translation, the importance of filtering noisy outputs, and the benefits of mixing translated samples with original English data during alignment. Our experiments focus on the low-resource Indic language Hindi and compare translations generated by Google Cloud Translation (GCP) and Llama-3.1-405B. The results highlight the promise of selective translation as a practical and effective method for improving multilingual alignment in LLMs.