@InProceedings{rajaee-EtAl:2026:wmt,
  author    = {Rajaee, Sara  and  Vincent, Sebastian  and  Berard, Alexandre  and  Fadaee, Marzieh  and  Marchisio, Kelly  and  Kocmi, Tom},
  title     = {Unlocking Reasoning Capability on Machine Translation in Large Language Models},
  booktitle      = {Proceedings of the Eleventh Conference on Machine Translation},
  month          = {October},
  year           = {2026},
  address        = {Budapest, Hungary},
  publisher      = {Association for Computational Linguistics},
  pages     = {597--613},
  abstract  = {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 RLMs on the WMT24++ benchmark and find that enabling explicit reasoning consistently degrades translation quality across languages and models. Our structural analysis shows that MT reasoning traces are highly linear, lacking revision, self-correction, and exploration of alternative translations, which limits their usefulness. Controlled reasoning-injection experiments demonstrate that providing high-quality reasoning traces from stronger models does not reliably improve weaker models' performance, indicating the failure lies within the format of MT reasoning. To address this mismatch, we propose a structured reasoning framework tailored to translation, based on multi-step improvements and dynamic iterative revision over difficult segments. Post-training a 111B RLM on such structured reasoning traces yields consistent gains over standard translation fine-tuning and injected generic reasoning baselines. Our findings demonstrate that reasoning must be task-structured to benefit MT.},
  url       = {https://aclanthology.org/2026.wmt-1.33}
}