Elamur Research⌕
☰
home / papers / 2025 / Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought

Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought

Paper · 2025 · Guijin Son, Donghun Yang, Hitesh Laxmichand Patel et al.

Summary

Most open work on distilling long chain-of-thought reasoning into smaller models targets English or Chinese. Pipelines built on translated data tend to carry translation artifacts and handle colloquial input poorly. Using Korean as a mid-resource case study, the authors propose Language-Mixed CoT. During the thinking phase, the model may alternate between English, which anchors the reasoning, and Korean, which keeps the meaning of the original query intact. They assembled 5.79M prompts written natively in Korean, sourced from online question-and-answer sites, exam questions, STEM material and programming tasks. With Qwen3-32B as the teacher, they generated 3.7M long traces under this schema to build the Yi-Sang dataset. After more than 100 ablations, they selected a 260k high-yield subset, Yi-Sang-HQ. Fine-tuning nine models (4B–35B) on it produced the KO-REAson series. The largest variant achieves the best average across nine Korean benchmarks, compared with similarly sized reasoners trained on closed data. The work shows a practical, RL-free path to competitive reasoning for mid-resource languages. Data, models and evaluation tooling are released.

Key points

  • Yi-Sang dataset: 5.79M prompts, 3.7M reasoning traces and about 6.77B tokens, which the authors describe as the largest publicly documented Korean post-training resource. Yi-Sang-HQ (260k examples) is the curated subset used for final training.
  • Translation baseline: Training Qwen2.5-1.5B on translated OpenThoughts raised MATH from 25.48 to 74.35. However, HAE-RAE Bench dropped from 35.24 to 15.34, which motivated the native-prompt approach.
  • KO-REAson-35B (built on A.X-3.1-35B): It reaches an average of 64.0 ±2.5, ranking first on 5 of 9 benchmarks and second on the remaining four. Comparison models include GPT-OSS-20B, R1-Distill-32B, QwQ-32B and EXAONE-Deep-32B.
  • Gains across model families: Smaller and mid-sized models gain about +18.6 points on average. Tested families include Gemma-3, Qwen2.5, Llama-3.1, Kanana-1.5, A.X-3.1 and KONI-Llama-3.1. Language-Mixed CoT beats both Korean-only and English-only CoT.
  • Evaluation and transfer: Held-in benchmarks were MCLM, KMMLU-Redux, HAE-RAE Bench and ClinicalQA. Held-out benchmarks were KMMLU-Hard, KMMLU-Pro, KSM, CLIcK and KoBALT-700. Despite Korean-only training, every model also improved on AIME24/25 and GPQA. Multimodal gains appeared on reasoning-oriented HAERAE-Vision and KAIO-2, but not on the factoid-style K-Viscuit. Models are released under Apache-2.0.

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

 Fields, topics and concepts

📚 Shelves 5The 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 worksComputer Vision and Pattern Recognition · 2,869 works

🧭 Topics 2Research subjects and languages the work is about.Reasoning · 1,426 worksMultilingual & cross-lingual NLP · 828 works

Cite

Cite as Guijin Son, Donghun Yang, Hitesh Laxmichand Patel, Amit Agarwal, Hyunwoo Ko, Chanuk Lim, Srikant Panda, Minhyuk Kim, Nikunj Drolia, Dasol Choi, Kyong-Ha Lee and Youngjae Yu. 2025. Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought. arXiv:2510.04230.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2510.04230 · arXiv 2510.04230

@misc{son2025pushing,
  title = {{Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought}},
  author = {Guijin Son, Donghun Yang, Hitesh Laxmichand Patel, Amit Agarwal, Hyunwoo Ko, Chanuk Lim, Srikant Panda, Minhyuk Kim, Nikunj Drolia, Dasol Choi, Kyong-Ha Lee and Youngjae Yu},
  year = {2025},
  doi = {10.48550/arxiv.2510.04230},
  eprint = {2510.04230},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2510.04230}
}

References

75 references, 30 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Sandhini Agarwal, Lama Ahmad, Jason Ai, Sam Altman, Andy Applebaum, Edwin Arbus, Rahul K Arora, Yu Bai, Bowen Baker, Haiming Bao, et al. gpt-oss-120b & gpt-oss-20b model card. arXiv preprint arXiv:2508.10925, 2025.
  2. Yunju Bak, Hojin Lee, Minho Ryu, Jiyeon Ham, Seungjae Jung, Daniel Wontae Nam, Taegyeong Eo, Donghun Lee, Doohae Jung, Boseop Kim, et al. Kanana: Compute-efficient bilingual language models. arXiv preprint arXiv:2502.18934, 2025. — [arXiv:2502.18934](https://arxiv.org/abs/2502.18934) find
  3. Akhiad Bercovich, Itay Levy, Izik Golan, Mohammad Dabbah, Ran El-Yaniv, Omri Puny, Ido Galil, Zach Moshe, Tomer Ronen, Najeeb Nabwani, et al. Llama-nemotron: Efficient reasoning models. arXiv preprint arXiv:2505.00949, 2025.
  4. Nuo Chen, Zinan Zheng, Ning Wu, Ming Gong, Dongmei Zhang, and Jia Li. Breaking language barriers in multilingual mathematical reasoning: Insights and observations. In Findings of the Association for Computational Linguistics: EMNLP 2024, pp. 7001–7016, 2024. find
  5. Yang Chen, Zhuolin Yang, Zihan Liu, Chankyu Lee, Peng Xu, Mohammad Shoeybi, Bryan Catanzaro, and Wei Ping. Acereason-nemotron: Advancing math and code reasoning through reinforcement learning. arXiv preprint arXiv:2505.16400, 2025. find
  6. Dasol Choi, Guijin Son, Soo Yong Kim, Gio Paik, and Seunghyeok Hong. Improving fine-grained visual understanding in vlms through text-only training. arXiv preprint arXiv:2412.12940, 2024. — [arXiv:2412.12940](https://arxiv.org/abs/2412.12940) find
  7. Dasol Choi, Guijin Son, Hanwool Lee, Minhyuk Kim, Hyunwoo Ko, Teabin Lim, Ahn Eungyeol, Jungwhan Kim, Seunghyeok Hong, and Youngsook Song. What users leave unsaid: Underspecified queries limit vision-language models, 2026. URL https://arxiv.org/abs/ 2601.06165. ↗
  8. Gheorghe Comanici, Eric Bieber, Mike Schaekermann, Ice Pasupat, Noveen Sachdeva, Inderjit Dhillon, Marcel Blistein, Ori Ram, Dan Zhang, Evan Rosen, et al. Gemini 2.5: Pushing the frontier with advanced reasoning, multimodality, long context, and next generation agentic capabilities. arXiv preprint a…
  9. Tri Dao, Dan Fu, Stefano Ermon, Atri Rudra, and Christopher Ré. Flashattention: Fast and memory-efficient exact attention with io-awareness. Advances in neural information processing systems, 35: 16344–16359, 2022. find
  10. Google Deepmind. Introducing gemini 2.0: our new ai model for the agentic era, 2024. URL https://blog.google/technology/google-deepmind/ google-gemini-ai-update-december-2024/\#project-astra. ↗
  11. Kanishk Gandhi, Ayush Chakravarthy, Anikait Singh, Nathan Lile, and Noah D Goodman. Cognitive behaviors that enable self-improving reasoners, or, four habits of highly effective stars. arXiv preprint arXiv:2503.01307, 2025.
  12. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, et al. The llama 3 herd of models. arXiv preprint arXiv:2407.21783, 2024.
  13. Xinyu Guan, Li Lyna Zhang, Yifei Liu, Ning Shang, Youran Sun, Yi Zhu, Fan Yang, and Mao Yang. rstar-math: Small llms can master math reasoning with self-evolved deep thinking. arXiv preprint arXiv:2501.04519, 2025.
  14. Etash Guha, Ryan Marten, Sedrick Keh, Negin Raoof, Georgios Smyrnis, Hritik Bansal, Marianna Nezhurina, Jean Mercat, Trung Vu, Zayne Sprague, Ashima Suvarna, Benjamin Feuer, Liangyu Chen, Zaid Khan, Eric Frankel, Sachin Grover, Caroline Choi, Niklas Muennighoff, Shiye Su, Wanjia Zhao, John Yang, Shr… — OpenThoughts: Data Recipes for Reasoning Models
  15. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, et al. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948, 2025.
  16. Huy Hoang Ha. Pensez: Less data, better reasoning–rethinking french llm. arXiv:2503.13661, 2025. arXiv preprint Tao He, Rongchuan Mu, Lizi Liao, Yixin Cao, Ming Liu, and Bing Qin. Good learners think their thinking: Generative prm makes large reasoning model more efficient math learner. arXiv prepri… find
  17. Andreas Hochlehnert, Hardik Bhatnagar, Vishaal Udandarao, Samuel Albanie, Ameya Prabhu, and Matthias Bethge. A sober look at progress in language model reasoning: Pitfalls and paths to reproducibility. arXiv preprint arXiv:2504.07086, 2025. find
  18. Jiwoo Hong, Noah Lee, Rodrigo Martı́nez-Castaño, César Rodrı́guez, and James Thorne. Cross-lingual transfer of reward models in multilingual alignment. arXiv preprint arXiv:2410.18027, 2024. find
  19. Seokhee Hong, Sunkyoung Kim, Guijin Son, Soyeon Kim, Yeonjung Hong, and Jinsik Lee. From kmmlu-redux to kmmlu-pro: A professional korean benchmark suite for llm evaluation. arXiv preprint arXiv:2507.08924, 2025. — [arXiv:2507.08924](https://arxiv.org/abs/2507.08924) find
  20. Pin-Lun Hsu, Yun Dai, Vignesh Kothapalli, Qingquan Song, Shao Tang, Siyu Zhu, Steven Shimizu, Shivam Sahni, Haowen Ning, and Yanning Chen. Liger kernel: Efficient triton kernels for llm training. arXiv preprint arXiv:2410.10989, 2024. find
  21. Jingcheng Hu, Yinmin Zhang, Qi Han, Daxin Jiang, Xiangyu Zhang, and Heung-Yeung Shum. Open-reasoner-zero: An open source approach to scaling up reinforcement learning on the base model. arXiv preprint arXiv:2503.24290, 2025.
  22. Hugging Face. Open r1: A fully open reproduction of deepseek-r1, January 2025. URL https://github.com/huggingface/open-r1. ↗
  23. Aaron Hurst, Adam Lerer, Adam P Goucher, Adam Perelman, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Welihinda, Alan Hayes, Alec Radford, et al. Gpt-4o system card. arXiv preprint arXiv:2410.21276, 2024.
  24. Aaron Jaech, Adam Kalai, Adam Lerer, Adam Richardson, Ahmed El-Kishky, Aiden Low, Alec Helyar, Aleksander Madry, Alex Beutel, Alex Carney, et al. Openai o1 system card. arXiv preprint arXiv:2412.16720, 2024.
  25. Yunjie Ji, Xiaoyu Tian, Sitong Zhao, Haotian Wang, Shuaiting Chen, Yiping Peng, Han Zhao, and Xiangang Li. Am-thinking-v1: Advancing the frontier of reasoning at 32b scale. arXiv preprint arXiv:2505.08311, 2025. — [arXiv:2505.08311](https://arxiv.org/abs/2505.08311) find
  26. Andy L Jones. Scaling scaling laws with board games. arXiv preprint arXiv:2104.03113, 2021. — [arXiv:2104.03113](https://arxiv.org/abs/2104.03113) find
  27. Eunsu Kim, Juyoung Suk, Philhoon Oh, Haneul Yoo, James Thorne, and Alice Oh. Click: A benchmark dataset of cultural and linguistic intelligence in korean. arXiv preprint arXiv:2403.06412, 2024. find
  28. KISTI. Kisti-koni/koni-llama3.1-70b-instruct-preview. 2024. URL https://huggingface.co/KISTI-KONI/KONI-Llama3.1-70B-Instruct-preview. ↗
  29. Hyunwoo Ko, Guijin Son, and Dasol Choi. Understand, solve and translate: Bridging the multilingual mathematical reasoning gap. arXiv preprint arXiv:2501.02448, 2025. find
  30. Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph E. Gonzalez, Hao Zhang, and Ion Stoica. Efficient memory management for large language model serving with pagedattention. In Proceedings of the ACM SIGOPS 29th Symposium on Operating Systems Principles, 2023.
  31. SKT AI Model Lab. A.x 3.1. 2025. URL https://huggingface.co/skt/A.X-3.1. ↗
  32. Huiyuan Lai and Malvina Nissim. mcot: Multilingual instruction tuning for reasoning consistency in language models. arXiv preprint arXiv:2406.02301, 2024.
  33. Jungyup Lee, Jemin Kim, Sang Park, and SeungJae Lee. //huggingface.co/dnotitia/DNA-R1. Dna r1, 2025a. find
  34. URL https: Nahyun Lee, Guijin Son, Hyunwoo Ko, and Kyubeen Han. Kaio: A collection of more challenging korean questions. arXiv preprint arXiv:2509.14752, 2025b. — [arXiv:2509.14752](https://arxiv.org/abs/2509.14752) find
  35. Jia LI, Edward Beeching, Lewis Tunstall, Ben Lipkin, Roman Soletskyi, Shengyi Costa Huang, Kashif Rasul, Longhui Yu, Albert Jiang, Ziju Shen, Zihan Qin, Bin Dong, Li Zhou, Yann Fleureau, Guillaume Lample, and Stanislas Polu. Numinamath. [https://huggingface. co/AI-MO/NuminaMath-1.5](https://github.c… find
  36. Yafu Li, Ronghao Zhang, Zhilin Wang, Huajian Zhang, Leyang Cui, Yongjing Yin, Tong Xiao, and Yue Zhang. Lost in literalism: How supervised training shapes translationese in llms. arXiv preprint arXiv:2503.04369, 2025a.
  37. Yu Li, Zhuoshi Pan, Honglin Lin, Mengyuan Sun, Conghui He, and Lijun Wu. Can one domain help others? a data-centric study on multi-domain reasoning via reinforcement learning. arXiv preprint arXiv:2507.17512, 2025b. lightblue. lightblue/DeepSeek-R1-Distill-Qwen-1.5B-Multilingual. https://huggingface… — [arXiv:2507.17512](https://arxiv.org/abs/2507.17512) find
  38. Cong Liu, Zhong Wang, ShengYu Shen, Jialiang Peng, Xiaoli Zhang, ZhenDong Du, and YaFang Wang. The chinese dataset distilled from deepseek-r1-671b. https://huggingface.co/ datasets/Congliu/Chinese-DeepSeek-R1-Distill-data-110k, 2025a. ↗
  39. Yifei Liu, Li Lyna Zhang, Yi Zhu, Bingcheng Dong, Xudong Zhou, Ning Shang, Fan Yang, and Mao Yang. rstar-coder: Scaling competitive code reasoning with a large-scale verified dataset. arXiv preprint arXiv:2505.21297, 2025b.
  40. Zijun Liu, Peiyi Wang, Runxin Xu, Shirong Ma, Chong Ruan, Peng Li, Yang Liu, and Yu Wu. Inference-time scaling for generalist reward modeling. arXiv preprint arXiv:2504.02495, 2025c.
  41. Michael Luo, Sijun Tan, Justin Wong, Xiaoxiang Shi, William Y. Tang, Manan Roongta, Colin Cai, Jeffrey Luo, Li Erran Li, Raluca Ada Popa, and Ion Stoica. Deepscaler: Surpassing o1-preview with a 1.5b model by scaling rl. https://pretty-radio-b75.notion.site/ DeepScaleR-Surpassing-O1-Preview-with-a-1… ↗
  42. MeCab-KO Contributors. Mecab-ko: Korean morphological analyzer. https://github.com/ hephaex/mecab-ko. Accessed: 2025-12-02. ↗
  43. Swaroop Mishra, Daniel Khashabi, Chitta Baral, and Hannaneh Hajishirzi. Cross-task generalization via natural language crowdsourcing instructions. arXiv preprint arXiv:2104.08773, 2021. find
  44. Niklas Muennighoff, Zitong Yang, Weijia Shi, Xiang Lisa Li, Li Fei-Fei, Hannaneh Hajishirzi, Luke Zettlemoyer, Percy Liang, Emmanuel Candès, and Tatsunori Hashimoto. s1: Simple test-time scaling. arXiv preprint arXiv:2501.19393, 2025.
  45. Team OLMo, Pete Walsh, Luca Soldaini, Dirk Groeneveld, Kyle Lo, Shane Arora, Akshita Bhagia, Yuling Gu, Shengyi Huang, Matt Jordan, et al. 2 olmo 2 furious. arXiv preprint arXiv:2501.00656, 2024. find
  46. ChaeHun Park, Yujin Baek, Jaeseok Kim, Yu-Jung Heo, Du-Seong Chang, and Jaegul Choo. Evaluating visual and cultural interpretation: The k-viscuit benchmark with human-vlm collaboration. arXiv preprint arXiv:2406.16469, 2024. find
  47. Cheonbok Park, Jeonghoon Kim, Joosung Lee, Sanghwan Bae, Jaegul Choo, and Kang Min Yoo. Cross-lingual collapse: How language-centric foundation models shape reasoning in large language models. arXiv preprint arXiv:2506.05850, 2025.
  48. Kunat Pipatanakul, Pittawat Taveekitworachai, Potsawee Manakul, and Kasima Tharnpipitchai. Adapting language-specific llms to a reasoning model in one day via model merging–an open recipe. arXiv preprint arXiv:2502.09056, 2025. — [arXiv:2502.09056](https://arxiv.org/abs/2502.09056) find
  49. Jirui Qi, Shan Chen, Zidi Xiong, Raquel Fernández, Danielle S Bitterman, and Arianna Bisazza. When models reason in your language: Controlling thinking trace language comes at the cost of accuracy. arXiv preprint arXiv:2505.22888, 2025. — [arXiv:2505.22888](https://arxiv.org/abs/2505.22888) find
  50. Qwen, :, An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, Keming Lu, Keqin Bao, Kexin Yang, Le Yu, Mei Li, Mingfeng… find
  51. Leonardo Ranaldi and Giulia Pucci. Multilingual reasoning via self-training. In Luis Chiruzzo, Alan Ritter, and Lu Wang (eds.), Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Pape… find
  52. Abhinav Rastogi, Albert Q Jiang, Andy Lo, Gabrielle Berrada, Guillaume Lample, Jason Rute, Joep Barmentlo, Karmesh Yadav, Kartik Khandelwal, Khyathi Raghavi Chandu, et al. Magistral. arXiv preprint arXiv:2506.10910, 2025. find
  53. David Rein, Betty Li Hou, Asa Cooper Stickland, Jackson Petty, Richard Yuanzhe Pang, Julien Dirani, Julian Michael, and Samuel R Bowman. Gpqa: A graduate-level google-proof q&a benchmark. In First Conference on Language Modeling, 2024.
  54. LG Research, Kyunghoon Bae, Eunbi Choi, Kibong Choi, Stanley Jungkyu Choi, Yemuk Choi, Seokhee Hong, Junwon Hwang, Hyojin Jeon, Kijeong Jeon, et al. Exaone deep: Reasoning enhanced language models. arXiv preprint arXiv:2503.12524, 2025. — [arXiv:2503.12524](https://arxiv.org/abs/2503.12524) find
  55. Stephen E Robertson, Steve Walker, Susan Jones, Micheline M Hancock-Beaulieu, Mike Gatford, et al. Okapi at TREC-3. British Library Research and Development Department, 1995.
  56. Rulin Shao, Shuyue Stella Li, Rui Xin, Scott Geng, Yiping Wang, Sewoong Oh, Simon Shaolei Du, Nathan Lambert, Sewon Min, Ranjay Krishna, et al. Spurious rewards: Rethinking training signals in rlvr. arXiv preprint arXiv:2506.10947, 2025. find
  57. Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, YK Li, Yang Wu, et al. Deepseekmath: Pushing the limits of mathematical reasoning in open language models. arXiv preprint arXiv:2402.03300, 2024.
  58. Hyopil Shin, Sangah Lee, Dongjun Jang, Wooseok Song, Jaeyoon Kim, Chaeyoung Oh, Hyemi Jo, Youngchae Ahn, Sihyun Oh, Hyohyeong Chang, et al. Kobalt: Korean benchmark for advanced linguistic tasks. arXiv preprint arXiv:2505.16125, 2025. — [arXiv:2505.16125](https://arxiv.org/abs/2505.16125) find
  59. Guijin Son, Hanwool Lee, Suwan Kim, Huiseo Kim, Jaecheol Lee, Je Won Yeom, Jihyu Jung, Jung Woo Kim, and Songseong Kim. Hae-rae bench: Evaluation of korean knowledge in language models. arXiv preprint arXiv:2309.02706, 2023. find
  60. Guijin Son, Hanwool Lee, Sungdong Kim, Seungone Kim, Niklas Muennighoff, Taekyoon Choi, Cheonbok Park, Kang Min Yoo, and Stella Biderman. Kmmlu: Measuring massive multitask language understanding in korean. arXiv preprint arXiv:2402.11548, 2024.
  61. Guijin Son, Jiwoo Hong, Hyunwoo Ko, and James Thorne. Linguistic generalizability of test-time scaling in mathematical reasoning. arXiv preprint arXiv:2502.17407, 2025.
  62. Zayne Sprague, Fangcong Yin, Juan Diego Rodriguez, Dongwei Jiang, Manya Wadhwa, Prasann Singhal, Xinyu Zhao, Xi Ye, Kyle Mahowald, and Greg Durrett. To cot or not to cot? chain-of-thought helps mainly on math and symbolic reasoning. arXiv preprint arXiv:2409.12183, 2024.
  63. Gemma Team, Aishwarya Kamath, Johan Ferret, Shreya Pathak, Nino Vieillard, Ramona Merhej, Sarah Perrin, Tatiana Matejovicova, Alexandre Ramé, Morgane Rivière, et al. Gemma 3 technical report. arXiv preprint arXiv:2503.19786, 2025. find
  64. Qwen Team. Qwq-32b: Embracing the power of reinforcement learning, March 2025. URL https://qwenlm.github.io/blog/qwq-32b/. ↗
  65. Shubham Toshniwal, Wei Du, Ivan Moshkov, Branislav Kisacanin, Alexan Ayrapetyan, and Igor Gitman. Openmathinstruct-2: Accelerating ai for math with massive open-source instruction data. arXiv preprint arXiv:2410.01560, 2024. find
  66. Zengzhi Wang, Fan Zhou, Xuefeng Li, and Pengfei Liu. Octothinker: Mid-training incentivizes reinforcement learning scaling. arXiv preprint arXiv:2506.20512, 2025. — [arXiv:2506.20512](https://arxiv.org/abs/2506.20512) find
  67. Liang Wen, Yunke Cai, Fenrui Xiao, Xin He, Qi An, Zhenyu Duan, Yimin Du, Junchen Liu, Lifu Tang, Xiaowei Lv, et al. Light-r1: Curriculum sft, dpo and rl for long cot from scratch and beyond. arXiv preprint arXiv:2503.10460, 2025. find
  68. Can Xu, Qingfeng Sun, Kai Zheng, Xiubo Geng, Pu Zhao, Jiazhan Feng, Chongyang Tao, and Daxin Jiang. Wizardlm: Empowering large language models to follow complex instructions. arXiv preprint arXiv:2304.12244, 2023.
  69. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al. Qwen3 technical report. arXiv preprint arXiv:2505.09388, 2025.
  70. Yixin Ye, Zhen Huang, Yang Xiao, Ethan Chern, Shijie Xia, and Pengfei Liu. Limo: Less is more for reasoning. arXiv preprint arXiv:2502.03387, 2025.
  71. Edward Yeo, Yuxuan Tong, Morry Niu, Graham Neubig, and Xiang Yue. Demystifying long chain-of-thought reasoning in llms. arXiv preprint arXiv:2502.03373, 2025.
  72. 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. Crosslingual reasoning through test-time scaling. arXiv preprint arXiv:2505.05408, 2025. — [arXiv:2505.05408](https://arxiv.org/abs/2505.05408) find
  73. Qiying Yu, Zheng Zhang, Ruofei Zhu, Yufeng Yuan, Xiaochen Zuo, Yu Yue, Weinan Dai, Tiantian Fan, Gaohong Liu, Lingjun Liu, et al. Dapo: An open-source llm reinforcement learning system at scale. arXiv preprint arXiv:2503.14476, 2025.
  74. Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric Xing, et al. Judging llm-as-a-judge with mt-bench and chatbot arena. Advances in neural information processing systems, 36:46595–46623, 2023.
  75. Wenhao Zhu, Shujian Huang, Fei Yuan, Shuaijie She, Jiajun Chen, and Alexandra Birch. Question translation training for better multilingual reasoning. arXiv preprint arXiv:2401.07817, 2024. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/son-2025-pushing-multilingual-reasoning-models-language/