# Unlocking Reasoning Capability on Machine Translation in Large Language Models

arXiv: https://arxiv.org/abs/2602.14763 · HTML: https://mlsanity.com/paper/2602.14763/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 15 of its 44 references that are in the map.
- First posted on arXiv 16 Feb 2026.

## Placed in

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

## Builds on (most cited first, 15 of 15)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [Kimi K2: Open Agentic Intelligence](https://mlsanity.com/paper/2507.20534/) (2025)
- [NVIDIA Nemotron Nano 2: An Accurate and Efficient Hybrid Mamba-Transformer Reasoning Model](https://mlsanity.com/paper/2508.14444/) (2025)
- [SealQA: Raising the Bar for Reasoning in Search-Augmented Language Models](https://mlsanity.com/paper/2506.01062/) (2025)
- [Magistral](https://mlsanity.com/paper/2506.10910/) (2025)
- [Command A: An Enterprise-Ready Large Language Model](https://mlsanity.com/paper/2504.00698/) (2025)
- [NaturalThoughts: Selecting and Distilling Reasoning Traces for General Reasoning Tasks](https://mlsanity.com/paper/2507.01921/) (2025)
- [New Trends for Modern Machine Translation with Large Reasoning Models](https://mlsanity.com/paper/2503.10351/) (2025)
- [Lost at the Beginning of Reasoning](https://mlsanity.com/paper/2506.22058/) (2025)
- [Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought](https://mlsanity.com/paper/2510.04230/) (2025)
- [Beyond English-Centric Training: How Reinforcement Learning Improves Cross-Lingual Reasoning in LLMs](https://mlsanity.com/paper/2509.23657/) (2025)
- [Test-Time Scaling of Reasoning Models for Machine Translation](https://mlsanity.com/paper/2510.06471/) (2025)
- [LLM Reasoning for Machine Translation: Synthetic Data Generation over Thinking Tokens](https://mlsanity.com/paper/2510.11919/) (2025)

## Built on by (most cited first, 6 of 6)

- [Should We be Pedantic About Reasoning Errors in Machine Translation?](https://mlsanity.com/paper/2604.09890/) (2026)
- [Reasoning over Grammar: Can Synthetic Linguistic Reasoning Traces Enhance Low-Resource Machine Translation?](https://mlsanity.com/paper/2606.03782/) (2026)
- [PluraMath: Extending Mathematical Reasoning Evaluation Beyond High-Resource Languages](https://mlsanity.com/paper/2607.05992/) (2026)
- [LatentMT: Machine Translation with Latent Reasoning](https://mlsanity.com/paper/2607.18618/) (2026)
- [Reasoning about In-Context Samples for Machine-Translation](https://mlsanity.com/paper/2608.27036/) (2026)
- [Ladders in Chaos: When, How, (and Perhaps Why) Does Test-Time Scaling Improve LLM Machine Translation](https://mlsanity.com/paper/2608.28496/) (2026)

## Abstract (arXiv)

Reasoning-oriented large language models (RLMs) achieve strong gains on tasks such as mathematics and coding by generating explicit intermediate reasoning. However, their impact on machine translation (MT) remains underexplored. We systematically evaluate several open- and closed-weights RLMs on the WMT24++ benchmark and find that enabling explicit reasoning consistently degrades translation quality across languages and models. Analysis reveals that MT reasoning traces are highly linear, lacking revision, self-correction and exploration of alternative translations, which limits their usefulness. Furthermore, injecting higher-quality reasoning traces from stronger models does not reliably improve weaker models' performance. To address this mismatch, we propose a structured reasoning framework tailored to translation, based on multi-step drafting, adequacy refinement, fluency improvement, and selective iterative revision. We curate a synthetic dataset of dynamic structured reasoning traces and post-train a large reasoning model on this data. Experiments show significant improvements over standard translation fine-tuning and injected generic reasoning baselines. Our findings demonstrate that reasoning must be task-structured to benefit MT.