---
title: "Cross-lingual Collapse — paper by Choo; and Yoo (2025)"
description: "The authors study how reinforcement learning with verifiable rewards (RLVR) affects the language of… 38 references, cited by 10 papers here."
authors: ["Cheonbok Park; Jeonghoon Kim; Joosung Lee; Sanghwan Bae; Jaegul Choo;", "Kang Min Yoo"]
year: 2025
venue: "arXiv:2506.05850"
source: "https://arxiv.org/abs/2506.05850"
doi: "10.48550/arxiv.2506.05850"
arxiv: "2506.05850"
licence: "CC BY 4.0 (arXiv)"
licence_url: "http://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/park-2025-cross-lingual-collapse-language-centric/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/park-2025-cross-lingual-collapse-language-centric/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2025](https://md.research.elamur.ai/free/en/papers/2025/page/8/) / Cross-lingual Collapse: How Language-Centric Foundation Models Shape Reasoning in Large Language Models

# Cross-lingual Collapse: How Language-Centric Foundation Models Shape Reasoning in Large Language Models

**Authors:** Cheonbok Park; Jeonghoon Kim; Joosung Lee; Sanghwan Bae; Jaegul Choo; and [Kang Min Yoo](https://md.research.elamur.ai/free/en/authors/kang-min-yoo/) · **Year:** 2025 · **Venue:** arXiv:2506.05850 · **Licence:** [CC BY 4.0 (arXiv)](http://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2506.05850) · [PDF](https://arxiv.org/pdf/2506.05850) · [DOI 10.48550/arxiv.2506.05850](https://doi.org/10.48550/arxiv.2506.05850) · [arXiv 2506.05850](https://arxiv.org/abs/2506.05850) · [OpenAlex](https://openalex.org/W4417114674)

**Citations:** 0 citations (OpenAlex, 2026-10-14) · [10](https://md.research.elamur.ai/free/en/papers/park-2025-cross-lingual-collapse-language-centric/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
The authors study how reinforcement learning with verifiable rewards (RLVR) affects the language of a model's reasoning. They find that when multilingual LLMs are trained to reason in a non-English language, the chain-of-thought drifts back to English, the dominant pre-training language. They call this "Cross-lingual Collapse." GSM8K and a subset of SimpleRL-Zoo were machine-translated with GPT-4o into Korean, Ukrainian, Chinese, Thai and Japanese and quality-filtered. Three small instruct models (OLMo 2, Llama-3.2, Qwen-2.5) were then trained with GRPO in the verl framework. Throughout training the authors measured two things: answer accuracy, using math-verify on translated GSM8K and MATH500, and a script-based word ratio showing which language the reasoning uses. Accuracy usually improves exactly as the reasoning switches to English. Several factors make the switch stronger: English-heavy backbones, long-CoT GRPO, harder problems and high-entropy sampling. Language-consistency rewards, decoding controls and more balanced backbones keep the target language but cost accuracy. For low-resource languages such as Ukrainian, this shows that RL-trained reasoning models can look accurate while no longer reasoning in the user's language.

## Key points
- **Ukrainian pilot.** Llama-3.2-3B Instruct was trained on Ukrainian GSM8K. Between roughly steps 100 and 250, the Ukrainian share of rollout reasoning dropped sharply while English took over and validation accuracy jumped.
- **Ukrainian GSM8K results for Llama after GRPO.** Accuracy reached 70.9% (+17.1 points over the base model). The Ukrainian word ratio fell to 0.3% (−82.6), and English made up 96.8% of the reasoning.
- **Mitigation trade-off (Llama, Ukrainian).** Three interventions were tested: a proportion reward, a threshold reward (both with λ = 0.5) and rollout-level filtering, which discards rollouts that are at least 50% English. All three raised the Ukrainian ratio to 82–88% but lowered accuracy to about 62–64%. Rollout filtering gave the best balance (63.8% accuracy, 88.0% target-language ratio).
- **Harder data amplifies collapse (Qwen-2.5-1.5B, Korean).** Adding the hard SimpleRL-Zoo data raised GSM8K accuracy to 47.5%, but the Korean ratio fell from 94.0% to 14.5%. SFT kept about 97% Korean but reached lower accuracy (40.9–42.5%).
- **Note for cataloguing.** Version v3 of arXiv:2506.05850 is retitled "When Verifiable Rewards Switch the Language: Cross-lingual Collapse in Chain-of-Thought." In a related test, further GRPO on DeepSeek-R1-Distill Qwen also collapsed.

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

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works

🧭 Topics (3): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works · [🧮 Reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/) · 1,426 works · [🌍 Multilingual & cross-lingual NLP](https://md.research.elamur.ai/free/en/topics/multilingual-nlp/) · 828 works

## Cite

Cite as: Cheonbok Park; Jeonghoon Kim; Joosung Lee; Sanghwan Bae; Jaegul Choo; and Kang Min Yoo. 2025. Cross-lingual Collapse: How Language-Centric Foundation Models Shape Reasoning in Large Language Models. arXiv:2506.05850.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{choo2025cross,
  title = {{Cross-lingual Collapse: How Language-Centric Foundation Models Shape Reasoning in Large Language Models}},
  author = {Cheonbok Park; Jeonghoon Kim; Joosung Lee; Sanghwan Bae; Jaegul Choo; and Kang Min Yoo},
  year = {2025},
  doi = {10.48550/arxiv.2506.05850},
  eprint = {2506.05850},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2506.05850}
}
```

- [← previous Crisp: Cognitive Restructuring of Negative Thoughts through Multi-turn Supportive Dialogues](https://md.research.elamur.ai/free/en/papers/zhou-2025-crisp-cognitive-restructuring-negative-thoughts/)
- [next → Decompose-ToM: Enhancing Theory of Mind Reasoning in Large Language Models through Simulation and Task Decomposition](https://md.research.elamur.ai/free/en/papers/sarangi-2025-decompose-tom-enhancing-theory-mind/)
- [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/page/8/)

## References

38 references, 21 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Frederic Blain, Chrysoula Zerva, Ricardo Ribeiro, Nuno Miguel Guerreiro, Diptesh Kanojia, José GC de Souza, Beatriz Silva, Tânia Vaz, Yan Jingxuan, Fatemeh Azadi, et al. [Findings of the wmt 2023 shared task on quality estimation](https://md.research.elamur.ai/free/en/papers/blain-2023-findings-wmt-2023-shared-task/). In Eight conference on machine translation, pp. 629–653. Association…
2. Nuo Chen, Zinan Zheng, Ning Wu, Ming Gong, Dongmei Zhang, and Jia Li. Breaking language barriers in multilingual mathematical reasoning: Insights and observations. In Yaser Al-Onaizan, Mohit Bansal, and Yun-Nung Chen (eds.), Findings of the Association for Computational Linguistics: EMNLP 2024, pp.… [find](https://scholar.google.com/scholar?q=Nuo+Chen%2C+Zinan+Zheng%2C+Ning+Wu%2C+Ming+Gong%2C+Dongmei+Zhang%2C+and+Jia+Li%2E+Breaking+language+barriers+in+multilingual+mathematical+reasoning%3A+Insights+and+observations%2E+In+Yaser+Al%2DOnaizan%2C+Mohit+Bansal%2C+a)
3. Pinzhen Chen, Shaoxiong Ji, Nikolay Bogoychev, Andrey Kutuzov, Barry Haddow, and Kenneth Heafield. Monolingual or multilingual instruction tuning: Which makes a better alpaca. In Yvette Graham and Matthew Purver (eds.), Findings of the Association for Computational Linguistics: EACL 2024, pp. 1347–1… [find](https://scholar.google.com/scholar?q=Pinzhen+Chen%2C+Shaoxiong+Ji%2C+Nikolay+Bogoychev%2C+Andrey+Kutuzov%2C+Barry+Haddow%2C+and+Kenneth+Heafield%2E+Monolingual+or+multilingual+instruction+tuning%3A+Which+makes+a+better+alpaca%2E+In+Yvette+Graham+and+Mat)
4. Julian’s, Malta, March 2024b. Association for Computational Linguistics. URL https://aclanthology.org/2024.findings-eacl.90/. [↗](https://aclanthology.org/2024.findings-eacl.90/)
5. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, Christopher Hesse, and John Schulman. [Training verifiers to solve math word problems](https://md.research.elamur.ai/free/en/papers/cobbe-2021-training-verifiers-solve-math-word/). arXiv preprint arXiv:2110.14168, 2021.
6. Ganqu Cui, Yuchen Zhang, Jiacheng Chen, Lifan Yuan, Zhi Wang, Yuxin Zuo, Haozhan Li, Yuchen Fan, Huayu Chen, Weize Chen, et al. The entropy mechanism of reinforcement learning for reasoning language models. arXiv preprint arXiv:2505.22617, 2025. [find](https://scholar.google.com/scholar?q=Ganqu+Cui%2C+Yuchen+Zhang%2C+Jiacheng+Chen%2C+Lifan+Yuan%2C+Zhi+Wang%2C+Yuxin+Zuo%2C+Haozhan+Li%2C+Yuchen+Fan%2C+Huayu+Chen%2C+Weize+Chen%2C+et+al%2E+The+entropy+mechanism+of+reinforcement+learning+for+reasoning+language+m)
7. DeepSeek-AI, Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, Xiaokang Zhang, Xingkai Yu, Yu Wu, Z. F. Wu, Zhibin Gou, Zhihong Shao, Zhuoshu Li, Ziyi Gao, Aixin Liu, Bing Xue, Bingxuan Wang, Bochao Wu, Bei Feng, Chengda Lu, Chengg… — [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://md.research.elamur.ai/free/en/papers/deepseek-ai-2025-deepseek-r1-incentivizing-reasoning-capability/)
8. 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](https://md.research.elamur.ai/free/en/papers/grattafiori-2024-llama-3-herd-models/). arXiv preprint arXiv:2407.21783, 2024.
9. Nuno M Guerreiro, Ricardo Rei, Daan van Stigt, Luisa Coheur, Pierre Colombo, and André FT Martins. [xcomet: Transparent machine translation evaluation through fine-grained error detection](https://md.research.elamur.ai/free/en/papers/guerreiro-2024-xcomet-transparent-machine-translation-evaluation/). Transactions of the Association for Computational Linguistics, 12:979–995, 2024.
10. Tuomas Haarnoja, Aurick Zhou, Pieter Abbeel, and Sergey Levine. Soft actor-critic: Off-policy maximum entropy deep reinforcement learning with a stochastic actor. In International conference on machine learning, pp. 1861–1870. [find](https://scholar.google.com/scholar?q=Tuomas+Haarnoja%2C+Aurick+Zhou%2C+Pieter+Abbeel%2C+and+Sergey+Levine%2E+Soft+actor%2Dcritic%3A+Off%2Dpolicy+maximum+entropy+deep+reinforcement+learning+with+a+stochastic+actor%2E+In+International+conference+on+machin)
11. Tannon Kew, Florian Schottmann, and Rico Sennrich. Turning english-centric llms into polyglots: How much multilinguality is needed? In Yaser Al-Onaizan, Mohit Bansal, and Yun-Nung Chen (eds.), Findings of the Association for Computational Linguistics: EMNLP 2024, Miami, Florida, USA, November 12-16,… [find](https://scholar.google.com/scholar?q=Tannon+Kew%2C+Florian+Schottmann%2C+and+Rico+Sennrich%2E+Turning+english%2Dcentric+llms+into+polyglots%3A+How+much+multilinguality+is+needed%3F+In+Yaser+Al%2DOnaizan%2C+Mohit+Bansal%2C+and+Yun%2DNung+Chen+%28eds%2E%29%2C+Finding)
12. Hunter Lightman, Vineet Kosaraju, Yuri Burda, Harrison Edwards, Bowen Baker, Teddy Lee, Jan Leike, John Schulman, Ilya Sutskever, and Karl Cobbe. Let’s verify step by step. In The Twelfth International Conference on Learning Representations, ICLR 2024, [Vienna, Austria, May 7-11, 2024](https://md.research.elamur.ai/free/en/papers/lightman-2023-let-s-verify-step-step/).
13. OpenReview.net, 2024. URL [https://openreview.net/forum?id=](https://openreview.net/forum?id=) v8L0pN6EOi. [↗](https://openreview.net/forum?id=)
14. Jack Lindsey, Wes Gurnee, Emmanuel Ameisen, Brian Chen, Adam Pearce, Nicholas L. Turner, Craig Citro, David Abrahams, Shan Carter, Basil Hosmer, Jonathan Marcus, Michael Sklar, Adly Templeton, Trenton Bricken, Callum McDougall, Hoagy Cunningham, Thomas Henighan, Adam Jermyn, Andy Jones, Andrew Persi… [find](https://scholar.google.com/scholar?q=Jack+Lindsey%2C+Wes+Gurnee%2C+Emmanuel+Ameisen%2C+Brian+Chen%2C+Adam+Pearce%2C+Nicholas+L%2E+Turner%2C+Craig+Citro%2C+David+Abrahams%2C+Shan+Carter%2C+Basil+Hosmer%2C+Jonathan+Marcus%2C+Michael+Sklar%2C+Adly+Templeton%2C+Trenton)
15. Kelly Marchisio, Wei-Yin Ko, Alexandre Berard, Théo Dehaze, and Sebastian Ruder. [Understanding and mitigating language confusion in LLMs](https://md.research.elamur.ai/free/en/papers/marchisio-2024-understanding-mitigating-language-confusion-llms/). In Yaser Al-Onaizan, Mohit Bansal, and Yun-Nung Chen (eds.), Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pp. 6653–667…
16. Team OLMo, Pete Walsh, Luca Soldaini, Dirk Groeneveld, Kyle Lo, Shane Arora, Akshita Bhagia, Yuling Gu, Shengyi Huang, Matt Jordan, Nathan Lambert, Dustin Schwenk, Oyvind Tafjord, Taira Anderson, David Atkinson, Faeze Brahman, Christopher Clark, Pradeep Dasigi, Nouha Dziri, Michal Guerquin, Hamish I… [find](https://scholar.google.com/scholar?q=Team+OLMo%2C+Pete+Walsh%2C+Luca+Soldaini%2C+Dirk+Groeneveld%2C+Kyle+Lo%2C+Shane+Arora%2C+Akshita+Bhagia%2C+Yuling+Gu%2C+Shengyi+Huang%2C+Matt+Jordan%2C+Nathan+Lambert%2C+Dustin+Schwenk%2C+Oyvind+Tafjord%2C+Taira+Anderson%2C+Davi)
17. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul F. Christiano, Jan Leike, and Ryan Lowe. Train… — [Training language models to follow instructions with human feedback](https://md.research.elamur.ai/free/en/papers/ouyang-2022-training-language-models-follow-instructions/)
18. Lisa Schut, Yarin Gal, and Sebastian Farquhar. Do multilingual llms think in english? CoRR, abs/2502.15603, 2025. doi: 10.48550/ARXIV.2502.15603. URL https://doi.org/10. 48550/arXiv.2502.15603. — [doi:10.48550/arxiv.2502.15603](https://doi.org/10.48550/arxiv.2502.15603) [find](https://scholar.google.com/scholar?q=Lisa+Schut%2C+Yarin+Gal%2C+and+Sebastian+Farquhar%2E+Do+multilingual+llms+think+in+english%3F+CoRR%2C+abs%2F2502%2E15603%2C+2025%2E+doi%3A+10%2E48550%2FARXIV%2E2502%2E15603%2E+URL+https%3A%2F%2Fdoi%2Eorg%2F10%2E+48550%2FarXiv%2E2502%2E15603%2E+%E2%80%94+%5Bdoi)
19. Uri Shaham, Jonathan Herzig, Roee Aharoni, Idan Szpektor, Reut Tsarfaty, and Matan Eyal. [Multilingual instruction tuning with just a pinch of multilinguality](https://md.research.elamur.ai/free/en/papers/shaham-2024-multilingual-instruction-tuning-just-pinch/). In Lun-Wei Ku, Andre Martins, and Vivek Srikumar (eds.), Findings of the Association for Computational Linguistics, ACL 2024, Bangkok, Thaila…
20. Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Mingchuan Zhang, Y. K. Li, Y. Wu, and Daya Guo. [Deepseekmath: Pushing the limits of mathematical reasoning in open language models](https://md.research.elamur.ai/free/en/papers/shao-2024-deepseekmath-pushing-limits-mathematical-reasoning/). CoRR, abs/2402.03300, 2024. doi: 10.48550/ARXIV.2402.03300. URL https://doi.org/10.48550/arXiv.2402.03300. [↗](https://doi.org/10.48550/arXiv.2402.03300)
21. Shuaijie She, Wei Zou, Shujian Huang, Wenhao Zhu, Xiang Liu, Xiang Geng, and Jiajun Chen. [MAPO: Advancing multilingual reasoning through multilingual-alignment-as-preference optimization](https://md.research.elamur.ai/free/en/papers/she-2024-mapo-advancing-multilingual-reasoning-through/). In Lun-Wei Ku, Andre Martins, and Vivek Srikumar (eds.), Proceedings of the 62nd Annual Meeting of the Associati…
22. Guangming Sheng, Chi Zhang, Zilingfeng Ye, Xibin Wu, Wang Zhang, Ru Zhang, Yanghua Peng, Haibin Lin, and Chuan Wu. [Hybridflow: A flexible and efficient rlhf framework](https://md.research.elamur.ai/free/en/papers/sheng-2024-hybridflow-flexible-efficient-rlhf-framework/). arXiv preprint arXiv: 2409.19256, 2024.
23. Shivalika Singh, Angelika Romanou, Clémentine Fourrier, David I. Adelani, Jian Gang Ngui, Daniel Vila-Suero, Peerat Limkonchotiwat, Kelly Marchisio, Wei Qi Leong, Yosephine Susanto, Raymond Ng, Shayne Longpre, Wei-Yin Ko, Madeline Smith, Antoine Bosselut, Alice Oh, Andre F. T. Martins, Leshem Chosh… — [Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluation](https://md.research.elamur.ai/free/en/papers/singh-2025-global-mmlu-understanding-addressing-cultural/)
24. 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](https://scholar.google.com/scholar?q=Gemma+Team%2C+Aishwarya+Kamath%2C+Johan+Ferret%2C+Shreya+Pathak%2C+Nino+Vieillard%2C+Ramona+Merhej%2C+Sarah+Perrin%2C+Tatiana+Matejovicova%2C+Alexandre+Rame%CC%81%2C+Morgane+Rivie%CC%80re%2C+et+al%2E+Gemma+3+technical+report%2E+arXiv+)
25. Qwen Team. Qwen2.5: A party of foundation models, September 2024. URL [https://qwenlm.github.io/blog/qwen2.5/](https://qwenlm.github.io/blog/qwen2.5/). [↗](https://qwenlm.github.io/blog/qwen2.5/)
26. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc V Le, Ed H Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. Self-consistency improves chain of thought reasoning in language models. In The Eleventh International Conference on Learning Representations. Zhijun Wang, Jiahuan Li, Hao Zhou, Rongxian… [find](https://scholar.google.com/scholar?q=Xuezhi+Wang%2C+Jason+Wei%2C+Dale+Schuurmans%2C+Quoc+V+Le%2C+Ed+H+Chi%2C+Sharan+Narang%2C+Aakanksha+Chowdhery%2C+and+Denny+Zhou%2E+Self%2Dconsistency+improves+chain+of+thought+reasoning+in+language+models%2E+In+The+Eleven)
27. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, brian ichter, Fei Xia, Ed H. Chi, Quoc V Le, and Denny Zhou. Chain of thought prompting elicits reasoning in large language models. In Alice H. Oh, Alekh Agarwal, Danielle Belgrave, and Kyunghyun Cho (eds.), Advances in Neural Information Proce… — [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://md.research.elamur.ai/free/en/papers/wei-2022-chain-thought-prompting-elicits-reasoning/)
28. Guillaume Wenzek, Marie-Anne Lachaux, Alexis Conneau, Vishrav Chaudhary, Francisco Guzmán, Armand Joulin, and Edouard Grave. [CCNet: Extracting high quality monolingual datasets from web crawl data](https://md.research.elamur.ai/free/en/papers/wenzek-2020-ccnet-extracting-high-quality-monolingual/). In Nicoletta Calzolari, Frédéric Béchet, Philippe Blache, Khalid Choukri, Christopher Cieri, Thier…
29. Jianhao Yan, Pingchuan Yan, Yulong Chen, Judy Li, Xianchao Zhu, and Yue Zhang. Gpt-4 vs. human translators: A comprehensive evaluation of translation quality across languages, domains, and expertise levels. arXiv preprint arXiv:2407.03658, 2024. [find](https://scholar.google.com/scholar?q=Jianhao+Yan%2C+Pingchuan+Yan%2C+Yulong+Chen%2C+Judy+Li%2C+Xianchao+Zhu%2C+and+Yue+Zhang%2E+Gpt%2D4+vs%2E+human+translators%3A+A+comprehensive+evaluation+of+translation+quality+across+languages%2C+domains%2C+and+expertise+l)
30. 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](https://md.research.elamur.ai/free/en/papers/yang-2025-qwen3-technical-report/). arXiv preprint arXiv:2505.09388, 2025.
31. Shunyu Yao, Dian Yu, Jeffrey Zhao, Izhak Shafran, Thomas L. Griffiths, Yuan Cao, and Karthik R Narasimhan. Tree of thoughts: Deliberate problem solving with large language models. In Thirty-seventh Conference on Neural Information Processing Systems, 2023. URL [https://openreview.net/forum?id=5Xc1ec…](https://openreview.net/forum?id=5Xc1ec…) [↗](https://openreview.net/forum?id=5Xc1ec…)
32. Haneul Yoo, Cheonbok Park, Sangdoo Yun, Alice Oh, and Hwaran Lee. Code-switching curriculum learning for multilingual transfer in llms. arXiv preprint arXiv:2411.02460, 2024a. [find](https://scholar.google.com/scholar?q=Haneul+Yoo%2C+Cheonbok+Park%2C+Sangdoo+Yun%2C+Alice+Oh%2C+and+Hwaran+Lee%2E+Code%2Dswitching+curriculum+learning+for+multilingual+transfer+in+llms%2E+arXiv+preprint+arXiv%3A2411%2E02460%2C+2024a%2E)
33. Kang Min Yoo, Jaegeun Han, Sookyo In, Heewon Jeon, Jisu Jeong, Jaewook Kang, Hyunwook Kim, Kyung-Min Kim, Munhyong Kim, Sungju Kim, et al. [Hyperclova x technical report](https://md.research.elamur.ai/free/en/papers/yoo-2024-hyperclova-x-technical-report/). arXiv preprint arXiv:2404.01954, 2024b.
34. Dongkeun Yoon, Joel Jang, Sungdong Kim, Seungone Kim, Sheikh Shafayat, and Minjoon Seo. [LangBridge: Multilingual reasoning without multilingual supervision](https://md.research.elamur.ai/free/en/papers/yoon-2024-langbridge-multilingual-reasoning-without-multilingual/). In Lun-Wei Ku, Andre Martins, and Vivek Srikumar (eds.), Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistic…
35. Qiying Yu, Zheng Zhang, Ruofei Zhu, Yufeng Yuan, Xiaochen Zuo, Yu Yue, Tiantian Fan, Gaohong Liu, Lingjun Liu, Xin Liu, et al. [Dapo: An open-source llm reinforcement learning system at scale](https://md.research.elamur.ai/free/en/papers/yu-2025-dapo-open-source-llm-reinforcement/). arXiv preprint arXiv:2503.14476, 2025.
36. Weihao Zeng, Yuzhen Huang, Qian Liu, Wei Liu, Keqing He, Zejun Ma, and Junxian He. [Simplerl-zoo: Investigating and taming zero reinforcement learning for open base models in the wild](https://md.research.elamur.ai/free/en/papers/zeng-2025-simplerl-zoo-investigating-taming-zero/), 2025. URL https://arxiv.org/abs/2503.18892. [↗](https://arxiv.org/abs/2503.18892)
37. Chengzhi Zhong, Fei Cheng, Qianying Liu, Junfeng Jiang, Zhen Wan, Chenhui Chu, Yugo Murawaki, and Sadao Kurohashi. [Beyond english-centric llms: What language do multilingual language models think in?](https://md.research.elamur.ai/free/en/papers/zhong-2024-english-centric-llms-language-multilingual/) CoRR, abs/2408.10811, 2024. doi: 10.48550/ARXIV.2408.10811. URL https://doi.org/10.48550/arXiv.2408.… [↗](https://doi.org/10.48550/arXiv.2408.…)
38. Wenhao Zhu, Shujian Huang, Fei Yuan, Shuaijie She, Jiajun Chen, and Alexandra Birch. Question translation training for better multilingual reasoning. In Lun-Wei Ku, Andre Martins, and Vivek Srikumar (eds.), Findings of the Association for Computational Linguistics: ACL 2024, pp. 8411–8423, Bangkok,… [find](https://scholar.google.com/scholar?q=Wenhao+Zhu%2C+Shujian+Huang%2C+Fei+Yuan%2C+Shuaijie+She%2C+Jiajun+Chen%2C+and+Alexandra+Birch%2E+Question+translation+training+for+better+multilingual+reasoning%2E+In+Lun%2DWei+Ku%2C+Andre+Martins%2C+and+Vivek+Srikumar+%28)

*Updated 30 October 2026*