Elamur Research⌕
☰
home / papers / 2026 / Test-Time Scaling of Reasoning Models for Machine Translation

Test-Time Scaling of Reasoning Models for Machine Translation

Paper · 2026 · Zihao Li, Shaoxiong Ji and Jörg Tiedemann

Summary

The paper asks whether giving reasoning models more inference-time compute (test-time scaling, TTS) improves machine translation, as it does for math and coding. The authors test 12 reasoning models on eight MT benchmarks covering literary, biomedical, cultural, commonsense, terminology-constrained and retrieval-augmented translation. They study three settings: direct translation under controlled thinking budgets, forced extension of reasoning past the model's own stopping point, and two-stage post-editing, where a model revises its own draft. Budgets are enforced with a custom logits processor that closes the reasoning span at a fixed token limit.

For general-purpose models, extra thinking helps only slightly at very small budgets, then levels off. Models fine-tuned for a domain behave differently: their in-domain quality keeps improving until they naturally stop using more reasoning tokens, but this advantage mostly disappears on other domains. Forcing longer reasoning consistently hurts quality, while TTS reliably improves post-editing. The practical lesson is to spend compute on task-specific fine-tuning and draft-then-revise pipelines, not on longer single-pass reasoning.

Key points

  • Models (12 total):
    • Qwen-3 in six sizes (0.6B–32B).
    • Cogito-3B and Cogito-8B, built on Llama-3.2/3.1.
    • DRT-7B, DRT-8B and DRT-14B, fine-tuned on MetaphorTrans.
    • Proprietary Grok-3-Mini, whose reasoning_effort offers just two options, low and high.
  • Benchmarks: WMT24-Literary, LitEval-Corpus, MetaphorTrans, WMT23/24-Biomedical, CAMT, Commonsense-MT (three ambiguity subsets), RTT and RAGTrans. Most language pairs involve English or Chinese.
  • Metrics:
    • COMET-22 and COMETKiwi-22.
    • Gemini-2.0-Flash used as a judge, giving reference-based (GRB) and reference-free (GRF) scores on a 0–100 scale.
    • GEA, applied only to the literary sets, on a 0–100 scale and a 1–5 scale.
  • Forced extrapolation: a single "wait" token is inserted when the model tries to end its reasoning early. This degrades translation quality across the board.
  • Limitations: no human evaluation, no qualitative analysis of the reasoning chains, and low-resource languages are largely absent.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓University of HelsinkiHelsinki · Finland50 works · 5,318 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 2Research subjects and languages the work is about.Machine translation · 1,087 worksReasoning · 1,426 works

Cite

Cite as Zihao Li, Shaoxiong Ji and Jörg Tiedemann. 2026. Test-Time Scaling of Reasoning Models for Machine Translation. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.eacl-long.133

@inproceedings{li2026test,
  title = {{Test-Time Scaling of Reasoning Models for Machine Translation}},
  author = {Zihao Li, Shaoxiong Ji and Jörg Tiedemann},
  booktitle = {Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.eacl-long.133},
  url = {https://doi.org/10.18653/v1/2026.eacl-long.133}
}

References

48 references, 19 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Sweta Agrawal, Chunting Zhou, Mike Lewis, Luke Zettlemoyer, and Marjan Ghazvininejad. 2023. In-context examples selection for machine translation. In Findings of the Association for Computational Linguistics: ACL 2023, pages 8857–8873, Toronto, Canada. Association for Computational Linguistics.
  2. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel Ziegler, Jeffrey Wu, Clemens Winter, and… find
  3. Shanbo Cheng, Yu Bao, Qian Cao, Luyang Huang, Liyan Kang, Zhicheng Liu, Yu Lu, Wenhao Zhu, Jingwen Chen, Zhichao Huang, Tao Li, Yifu Li, Huiying Lin, Sitong Liu, Ningxin Peng, Shuaijie She, Lu Xu, Nuo Xu, Sen Yang, and 7 others. 2025. Seed-x: Building strong multilingual translation llm with 7b para… — [arXiv:2507.13618](https://arxiv.org/abs/2507.13618) find
  4. Gheorghe Comanici, Eric Bieber, Mike Schaekermann, Ice Pasupat, Noveen Sachdeva, Inderjit Dhillon, Marcel Blistein, Ori Ram, Dan Zhang, Evan Rosen, Luke Marris, Sam Petulla, Colin Gaffney, Asaf Aharoni, Nathan Lintz, Tiago Cardal Pais, Henrik Jacobsson, Idan Szpektor, Nan-Jiang Jiang, and 3416 other…
  5. Menglong Cui, Pengzhi Gao, Wei Liu, Jian Luan, and Bin Wang. 2025. Multilingual machine translation with open large language models at practical scale: An empirical study. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: H… find
  6. Deep Cogito. 2025. Cogito v1 Preview Introducing IDA as a path to general superintelligence. https://www.deepcogito.com/research/ cogito-v1-preview. Accessed: 2025-09-09. ↗
  7. Zhaopeng Feng, Yan Zhang, Hao Li, Bei Wu, Jiayu Liao, Wenqiang Liu, Jun Lang, Yang Feng, Jian Wu, and Zuozhu Liu. 2025. TEaR: Improving LLM-based machine translation with systematic self-refinement. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 3922–3938, Albuquerqu… find
  8. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, Amy Yang, Angela Fan, Anirudh Goyal, Anthony Hartshorn, Aobo Yang, Archi Mitra, Archie Sravankumar, Artem Korenev, Arthur Hinsvark, and 542 o… — The Llama 3 Herd of Models
  9. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, Runxin Xu, Ruoyu Zhang, Shirong Ma, Xiao Bi, Xiaokang Zhang, Xingkai Yu, Yu Wu, Z. F. Wu, Zhibin Gou, Zhihong Shao, Zhuoshu Li, Ziyi Gao, Aixin Liu, and 175 others. 2025. Deepseek-r1 incentivizes reasoning in llms through rein… find
  10. Jie He, Tao Wang, Deyi Xiong, and Qun Liu. 2020. The box is in the pen: Evaluating commonsense reasoning in neural machine translation. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 3662–3672, Online. Association for Computational Linguistics. find
  11. Minggui He, Yilun Liu, Shimin Tao, Yuanchang Luo, Hongyong Zeng, Chang Su, Li Zhang, Hongxia Ma, Daimeng Wei, Weibin Meng, Hao Yang, Boxing Chen, and Osamu Yoshie. 2025. R1-t1: Fully incentivizing translation capability in llms via reasoning learning. Preprint, arXiv:2502.19735.
  12. Amr Hendy, Mohamed Abdelrehim, Amr Sharaf, Vikas Raunak, Mohamed Gabr, Hitokazu Matsushita, Young Jin Kim, Mohamed Afify, and Hany Hassan Awadalla. 2023. How good are gpt models at machine translation? a comprehensive evaluation. arXiv preprint arXiv:2302.09210. find
  13. Aaron Hurst, Adam Lerer, Adam P. Goucher, Adam Perelman, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Welihinda, Alan Hayes, Alec Radford, Aleksander Mkadry, Alex Baker-Whitcomb, Alex Beutel, Alex Borzunov, Alex Carney, Alex Chow, Alexander Kirillov, Alex Nichol, Alex Paino, and 397 others. 2024. Gp… — GPT-4o System Card
  14. Aaron Jaech, Adam Kalai, Adam Lerer, Adam Richardson, Ahmed El-Kishky, Aiden Low, Alec Helyar, Aleksander Madry, Alex Beutel, Alex Carney, Alex Iftimie, Alex Karpenko, Alex Tachard Passos, Alexander Neitz, Alexander Prokofiev, Alexander Wei, Allison Tam, Ally Bennett, Ananya Kumar, and 242 others. 2… — OpenAI o1 System Card
  15. Tom Kocmi, Ekaterina Artemova, Eleftherios Avramidis, Rachel Bawden, Ondřej Bojar, Konstantin Dranch, Anton Dvorkovich, Sergey Dukanov, Mark Fishel, Markus Freitag, Thamme Gowda, Roman Grundkiewicz, Barry Haddow, Marzena Karpinska, Philipp Koehn, Howard Lakougna, Jessica Lundin, Christof Monz, Kento… — Findings of the WMT25 General Machine Translation Shared Task: Time to Stop Evaluating on Easy Test Sets
  16. Tom Kocmi, Eleftherios Avramidis, Rachel Bawden, Ondřej Bojar, Anton Dvorkovich, Christian Federmann, Mark Fishel, Markus Freitag, Thamme Gowda, Roman Grundkiewicz, Barry Haddow, Marzena Karpinska, Philipp Koehn, Benjamin Marie, Christof Monz, Kenton Murray, Masaaki Nagata, Martin Popel, Maja Popovi… — Findings of the WMT24 General Machine Translation Shared Task: The LLM Era Is Here but MT Is Not Solved Yet
  17. Tom Kocmi and Christian Federmann. 2023. Large language models are state-of-the-art evaluators of translation quality. In Proceedings of the 24th Annual Conference of the European Association for Machine Translation, pages 193–203, Tampere, Finland. European Association for Machine Translation. find
  18. Dacheng Li, Shiyi Cao, Chengkun Cao, Xiuyu Li, Shangyin Tan, Kurt Keutzer, Jiarong Xing, Joseph E Gonzalez, and Ion Stoica. 2025. S*: Test time scaling for code generation. arXiv preprint arXiv:2502.14382. find
  19. Xinnuo Li, Yunxiang Zhang, and Lu Wang. 2024. Improving language model self-correction capability with meta-feedback. OpenReview. find
  20. Sinuo Liu, Chenyang Lyu, Minghao Wu, Longyue Wang, Weihua Luo, Kaifu Zhang, and Zifu Shang. 2025. New trends for modern machine translation with large reasoning models. arXiv preprint arXiv:2503.10351.
  21. Chenyang Lyu, Zefeng Du, Jitao Xu, Yitao Duan, Minghao Wu, Teresa Lynn, Alham Fikri Aji, Derek F Wong, Siyou Liu, and Longyue Wang. 2023. A paradigm shift: The future of machine translation lies with large language models. arXiv preprint arXiv:2305.01181. find
  22. Niklas Muennighoff, Zitong Yang, Weijia Shi, Xiang Lisa Li, Li Fei-Fei, Hannaneh Hajishirzi, Luke Zettlemoyer, Percy Liang, Emmanuel Candès, and Tatsunori Hashimoto. 2025. s1: Simple test-time scaling. arXiv preprint arXiv:2501.19393.
  23. Mariana Neves, Cristian Grozea, Philippe Thomas, Roland Roller, Rachel Bawden, Aurélie Névéol, Steffen Castle, Vanessa Bonato, Giorgio Maria Di Nunzio, Federica Vezzani, Maika Vicente Navarro, Lana Yeganova, and Antonio Jimeno Yepes. 2024. Findings of the WMT 2024 biomedical translation shared task:… find
  24. Mariana Neves, Antonio Jimeno Yepes, Aurélie Névéol, Rachel Bawden, Giorgio Maria Di Nunzio, Roland Roller, Philippe Thomas, Federica Vezzani, Maika Vicente Navarro, Lana Yeganova, Dina Wiemann, and Cristian Grozea. 2023. Findings of the WMT 2023 biomedical translation shared task: Evaluation of Cha… find
  25. OpenAI. 2025. Openai o3 and o4-mini system card. System card, OpenAI. Published April 16, 2025. find
  26. Ricardo Rei, José G. C. de Souza, Duarte Alves, Chrysoula Zerva, Ana C Farinha, Taisiya Glushkova, Alon Lavie, Luisa Coheur, and André F. T. Martins. 2022a. COMET-22: Unbabel-IST 2022 submission for the metrics shared task. In Proceedings of the Seventh Conference on Machine Translation (WMT), pages…
  27. Ricardo Rei, Nuno M Guerreiro, José Pombal, João Alves, Pedro Teixeirinha, Amin Farajian, and André FT Martins. 2025. Tower+: Bridging generality and translation specialization in multilingual llms. arXiv preprint arXiv:2506.17080. find
  28. Ricardo Rei, Craig Stewart, Ana C Farinha, and Alon Lavie. 2020. COMET: A neural framework for MT evaluation. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 2685–2702, Online. Association for Computational Linguistics.
  29. Ricardo Rei, Marcos Treviso, Nuno M. Guerreiro, Chrysoula Zerva, Ana C Farinha, Christine Maroti, José G. C. de Souza, Taisiya Glushkova, Duarte Alves, Luisa Coheur, Alon Lavie, and André F. T. Martins. 2022b. find
  30. CometKiwi: IST-unbabel 2022 submission for the quality estimation shared task. In Proceedings of the Seventh Conference on Machine Translation (WMT), pages 634–645, Abu Dhabi, United Arab Emirates (Hybrid). Association for Computational Linguistics.
  31. Charlie Snell, Jaehoon Lee, Kelvin Xu, and Aviral Kumar. 2024. Scaling llm test-time compute optimally can be more effective than scaling model parameters. arXiv preprint arXiv:2408.03314.
  32. Guijin Son, Jiwoo Hong, Hyunwoo Ko, and James Thorne. 2025. Linguistic generalizability of test-time scaling in mathematical reasoning. arXiv preprint arXiv:2502.17407.
  33. Shaomu Tan, Ryosuke Mitani, Ritvik Choudhary, and Toshiyuki Sekiya. 2025. Investigating test-time scaling with reranking for machine translation. arXiv preprint arXiv:2509.19020. — [arXiv:2509.19020](https://arxiv.org/abs/2509.19020) find
  34. Khanh-Tung Tran, Barry O’Sullivan, and Hoang D Nguyen. 2025. Scaling test-time compute for low-resource languages: Multilingual reasoning in llms. arXiv preprint arXiv:2504.02890. — [arXiv:2504.02890](https://arxiv.org/abs/2504.02890) find
  35. David Vilar, Markus Freitag, Colin Cherry, Jiaming Luo, Viresh Ratnakar, and George Foster. 2023. Prompting PaLM for translation: Assessing strategies and performance. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 15406–1542… find
  36. Jiaan Wang, Fandong Meng, Yunlong Liang, and Jie Zhou. 2025a. DRT: Deep reasoning translation via long chain-of-thought. In Findings of the Association for Computational Linguistics: ACL 2025, pages 6770–6782, Vienna, Austria. Association for Computational Linguistics. find
  37. Jiaan Wang, Fandong Meng, Yingxue Zhang, and Jie Zhou. 2024a. Retrieval-augmented machine translation with unstructured knowledge. arXiv preprint arXiv:2412.04342. find
  38. Jiaan Wang, Fandong Meng, and Jie Zhou. 2025b. Deep reasoning translation via reinforcement learning. arXiv preprint arXiv:2504.10187. — [arXiv:2504.10187](https://arxiv.org/abs/2504.10187) find
  39. Jiaan Wang, Fandong Meng, and Jie Zhou. 2025c. Extrans: Multilingual deep reasoning translation via exemplar-enhanced reinforcement learning. arXiv preprint arXiv:2505.12996. — [arXiv:2505.12996](https://arxiv.org/abs/2505.12996) find
  40. Longyue Wang, Siyou Liu, Chenyang Lyu, Wenxiang Jiao, Xing Wang, Jiahao Xu, Zhaopeng Tu, Yan Gu, Weiyu Chen, Minghao Wu, Liting Zhou, Philipp Koehn, Andy Way, and Yulin Yuan. 2024b. Findings of the WMT 2024 shared task on discourse-level literary translation. In Proceedings of the Ninth Conference o… find
  41. Yutong Wang, Jiali Zeng, Xuebo Liu, Fandong Meng, Jie Zhou, and Min Zhang. 2024c. TasTe: Teaching large language models to translate through self-reflection. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 6144–6158, Bangkok,… find
  42. Di Wu, Seth Aycock, and Christof Monz. 2025. Please translate again: Two simple experiments on whether human-like reasoning helps translation. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 20424–20440, Suzhou, China. Association for Computational Li…
  43. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, Chujie Zheng, Dayiheng Liu, Fan Zhou, Fei Huang, Feng Hu, Hao Ge, Haoran Wei, Huan Lin, Jialong Tang, and 41 others. 2025. Qwen3 technical report. Preprint, arXiv:2505.09388.
  44. Zheng-Xin Yong, M Farid Adilazuarda, Jonibek Mansurov, Ruochen Zhang, Niklas Muennighoff, Carsten Eickhoff, Genta Indra Winata, Julia Kreutzer, Stephen H Bach, and Alham Fikri Aji. 2025. Crosslingual reasoning through test-time scaling. arXiv preprint arXiv:2505.05408. — [arXiv:2505.05408](https://arxiv.org/abs/2505.05408) find
  45. Huaao Zhang, Qiang Wang, Bo Qin, Zelin Shi, Haibo Wang, and Ming Chen. 2023. Understanding and improving the robustness of terminology constraints in neural machine translation. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages… find
  46. Ran Zhang, Wei Zhao, and Steffen Eger. 2025. How good are LLMs for literary translation, really? literary translation evaluation with humans and LLMs. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technol… find
  47. Mao Zheng, Zheng Li, Bingxin Qu, Mingyang Song, Yang Du, Mingrui Sun, and Di Wang. 2025. Hunyuan-mt technical report. arXiv preprint arXiv:2509.05209.
  48. Wenhao Zhu, Hongyi Liu, Qingxiu Dong, Jingjing Xu, Shujian Huang, Lingpeng Kong, Jiajun Chen, and Lei Li. 2024. Multilingual machine translation with large language models: Empirical results and analysis. In Findings of the Association for Computational Linguistics: NAACL 2024, pages 2765–2781, Mexi…

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/li-2026-test-time-scaling-reasoning-models/