---
title: "Less Languages, Less Tokens — paper by Zhang et al. (2026)"
description: "The authors propose UL-XCoT to address both issues. 53 references. Among them: 0."
authors: ["Chenyuan Zhang", "Qiguang Chen", "Xie Chen", "Zhuotao Tian", "Bowen Xing", "Meishan Zhang", "Libo Qin", "Baotian Hu", "Min Zhang"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2026.acl-long.1379/"
doi: "10.18653/v1/2026.acl-long.1379"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/zhang-2026-less-languages-less-tokens-efficient/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zhang-2026-less-languages-less-tokens-efficient/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/39/) / Less Languages, Less Tokens: An Efficient Unified Logic Cross-lingual Chain-of-Thought Reasoning Framework

# Less Languages, Less Tokens: An Efficient Unified Logic Cross-lingual Chain-of-Thought Reasoning Framework

**Authors:** Chenyuan Zhang, [Qiguang Chen](https://md.research.elamur.ai/free/en/authors/qiguang-chen/), [Xie Chen](https://md.research.elamur.ai/free/en/authors/xie-chen/), [Zhuotao Tian](https://md.research.elamur.ai/free/en/authors/zhuotao-tian/), Bowen Xing, [Meishan Zhang](https://md.research.elamur.ai/free/en/authors/meishan-zhang/), [Libo Qin](https://md.research.elamur.ai/free/en/authors/libo-qin/), [Baotian Hu](https://md.research.elamur.ai/free/en/authors/baotian-hu/) and [Min Zhang](https://md.research.elamur.ai/free/en/authors/min-zhang/) · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2026](https://md.research.elamur.ai/free/en/venues/acl/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.acl-long.1379/) · [PDF](https://aclanthology.org/2026.acl-long.1379.pdf) · [DOI 10.18653/v1/2026.acl-long.1379](https://doi.org/10.18653/v1/2026.acl-long.1379) · [OpenAlex](https://openalex.org/W7166856755)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Combining cross-lingual chain-of-thought (XCoT) with self-consistency improves multilingual reasoning, but it is costly. Earlier methods generate full reasoning traces in every candidate language, so compute grows with the number of languages. Pruning or comparing these traces is also difficult, because LLM hidden states differ systematically by language. The authors propose UL-XCoT to address both issues. It first builds a language-invariant "unified logic" space: language centers are averaged from validation-set hidden states, decomposed with SVD, and the main language-variation directions are projected out. Within this space, Candidate Language Selection keeps only the top-k languages whose query representations are closest to the source query. Dynamic CoT Pruning then tracks trajectory dynamics during decoding and stops weak paths early. The remaining answers are combined by voting. With DeepSeek-R1-Distill-Qwen-7B, UL-XCoT stays competitive in accuracy on PolyMath and MMLU-ProX-Lite while using far fewer tokens and less latency, and it is more robust for low-resource languages. This makes multilingual test-time scaling considerably cheaper.

## Key points
- Evaluation covers PolyMath (18 languages) and MMLU-ProX-Lite (29 languages), using DeepSeek-R1-Distill-Qwen-7B.
- On PolyMath, average decoding tokens fall by more than 50% versus AUTO-CAP and by more than 65% versus standard self-consistency (SC), with the lowest latency across languages.
- Varying the DCP pruning ratio ρ shows that values around 0.55–0.70 give the best balance between accuracy and efficiency. An LLM-judge analysis confirms that pruned traces are weaker, especially in step validity and completeness.
- Language selection adapts to each query rather than favoring a fixed list. Each language appears in the selected set about 5.6% of the time on average, and lower-frequency languages such as bn, ar and te are still chosen.
- The method requires white-box access to hidden states, so its use with black-box APIs is untested. Code is available at github.com/chenyuanTKCY/UL-XCoT.

*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

- [Harbin Institute of Technology](https://md.research.elamur.ai/free/en/universities/harbin-institute-of-technology/) — University, Harbin, China · 185 works · 29,140 citations
- [Shanghai Jiao Tong University](https://md.research.elamur.ai/free/en/universities/shanghai-jiao-tong-university/) — University, Shanghai, China · 212 works · 19,544 citations
- [Central South University](https://md.research.elamur.ai/free/en/universities/central-south-university/) — University, Changsha, China · 23 works · 657 citations
- [Shanghai Innovation Institute](https://md.research.elamur.ai/free/en/universities/shanghai-innovation-institute/) — University, Shanghai, China · 34 works · 325 citations
- [University of Science and Technology Beijing](https://md.research.elamur.ai/free/en/universities/university-of-science-and-technology-beijing/) — University, Beijing, China · 10 works · 315 citations

📚 Shelves (5): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,450 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,296 works · [👁️ Computer Vision and Pattern Recognition](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-vision-and-pattern-recognition/) · 3,486 works

🧭 Topics (2): [🧮 Reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/) · 1,685 works · [🌍 Multilingual & cross-lingual NLP](https://md.research.elamur.ai/free/en/topics/multilingual-nlp/) · 882 works

## Cite

Cite as: Chenyuan Zhang, Qiguang Chen, Xie Chen, Zhuotao Tian, Bowen Xing, Meishan Zhang, Libo Qin, Baotian Hu and Min Zhang. 2026. Less Languages, Less Tokens: An Efficient Unified Logic Cross-lingual Chain-of-Thought Reasoning Framework. Proceedings of the 64th Annual Meeting 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.

```bibtex
@inproceedings{zhang2026less,
  title = {{Less Languages, Less Tokens: An Efficient Unified Logic Cross-lingual Chain-of-Thought Reasoning Framework}},
  author = {Chenyuan Zhang, Qiguang Chen, Xie Chen, Zhuotao Tian, Bowen Xing, Meishan Zhang, Libo Qin, Baotian Hu and Min Zhang},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1379},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1379}
}
```

- [← previous Less is More: Knowledge-Aware Compression for Long Legal Judgment Prediction](https://md.research.elamur.ai/free/en/papers/lou-2026-less-is-more-knowledge-aware/)
- [next → Let Retrievers Think Before Action: Thought-Augmented Embedding for Dense Retrieval](https://md.research.elamur.ai/free/en/papers/yan-2026-let-retrievers-think-before-action/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/39/)

## References

53 references, 20 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Pranjal Aggarwal, Aman Madaan, Yiming Yang, et al. 2023. Let’s sample step by step: Adaptive-consistency for efficient reasoning and coding with llms. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 12375–12396. [find](https://scholar.google.com/scholar?q=Pranjal+Aggarwal%2C+Aman+Madaan%2C+Yiming+Yang%2C+et+al%2E+2023%2E+Let%E2%80%99s+sample+step+by+step%3A+Adaptive%2Dconsistency+for+efficient+reasoning+and+coding+with+llms%2E+In+Proceedings+of+the+2023+Conference+on+Empirica)
2. Sanchit Ahuja, Praneetha Vaddamanu, and Barun Patra. 2025. Efficientxlang: Towards improving token efficiency through cross-lingual reasoning. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 15612–15624. parallel reading comprehension dataset in 122 language variants.… [find](https://scholar.google.com/scholar?q=Sanchit+Ahuja%2C+Praneetha+Vaddamanu%2C+and+Barun+Patra%2E+2025%2E+Efficientxlang%3A+Towards+improving+token+efficiency+through+cross%2Dlingual+reasoning%2E+In+Findings+of+the+Association+for+Computational+Linguist)
3. Josh Barua, Seun Eisape, Kayo Yin, and Alane Suhr. 2025. Long chain-of-thought reasoning across languages. arXiv preprint arXiv:2508.14828. — [arXiv:2508.14828](https://arxiv.org/abs/2508.14828) [find](https://scholar.google.com/scholar?q=Josh+Barua%2C+Seun+Eisape%2C+Kayo+Yin%2C+and+Alane+Suhr%2E+2025%2E+Long+chain%2Dof%2Dthought+reasoning+across+languages%2E+arXiv+preprint+arXiv%3A2508%2E14828%2E+%E2%80%94+%5BarXiv%3A2508%2E14828%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2508%2E14828%29)
4. Linzheng Chai, Jian Yang, Tao Sun, Hongcheng Guo, Jiaheng Liu, Bing Wang, Xinnian Liang, Jiaqi Bai, Tongliang Li, Qiyao Peng, et al. 2025. xcot: Cross-lingual instruction tuning for cross-lingual chain-of-thought reasoning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39,… [find](https://scholar.google.com/scholar?q=Linzheng+Chai%2C+Jian+Yang%2C+Tao+Sun%2C+Hongcheng+Guo%2C+Jiaheng+Liu%2C+Bing+Wang%2C+Xinnian+Liang%2C+Jiaqi+Bai%2C+Tongliang+Li%2C+Qiyao+Peng%2C+et+al%2E+2025%2E+xcot%3A+Cross%2Dlingual+instruction+tuning+for+cross%2Dlingual+chai)
5. Kang Chen, Yaoning Wang, Kai Xiong, Zhuoka Feng, Wenhe Sun, Haotian Chen, and Yixin Cao. 2025a. Do llms signal when they’re right? evidence from neuron agreement. arXiv preprint arXiv:2510.26277. — [arXiv:2510.26277](https://arxiv.org/abs/2510.26277) [find](https://scholar.google.com/scholar?q=Kang+Chen%2C+Yaoning+Wang%2C+Kai+Xiong%2C+Zhuoka+Feng%2C+Wenhe+Sun%2C+Haotian+Chen%2C+and+Yixin+Cao%2E+2025a%2E+Do+llms+signal+when+they%E2%80%99re+right%3F+evidence+from+neuron+agreement%2E+arXiv+preprint+arXiv%3A2510%2E26277%2E+%E2%80%94+%5Ba)
6. Kang Chen, Mengdi Zhang, and Yixin Cao. 2025b. Less data less tokens: Multilingual unification learning for efficient test-time reasoning in llms. arXiv preprint arXiv:2506.18341. [find](https://scholar.google.com/scholar?q=Kang+Chen%2C+Mengdi+Zhang%2C+and+Yixin+Cao%2E+2025b%2E+Less+data+less+tokens%3A+Multilingual+unification+learning+for+efficient+test%2Dtime+reasoning+in+llms%2E+arXiv+preprint+arXiv%3A2506%2E18341%2E)
7. Nuo Chen, Zinan Zheng, Ning Wu, Ming Gong, Dongmei Zhang, and Jia Li. 2024a. [Breaking language barriers in multilingual mathematical reasoning: Insights and observations](https://md.research.elamur.ai/free/en/papers/chen-2024-breaking-language-barriers-multilingual-mathematical/). In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 7001–7016.
8. Qiguang Chen, Libo Qin, Jinhao Liu, Dengyun Peng, Jiannan Guan, Peng Wang, Mengkang Hu, Yuhang Zhou, Te Gao, and Wanxiang Che. 2025c. [Towards reasoning era: A survey of long chain-of-thought for reasoning large language models](https://md.research.elamur.ai/free/en/papers/chen-2025-towards-reasoning-era-survey-long/). arXiv preprint arXiv:2503.09567.
9. Qiguang Chen, Libo Qin, Jiaqi Wang, Jinxuan Zhou, and Wanxiang Che. 2024b. Unlocking the capabilities of thought: A reasoning boundary framework to quantify and optimize chain-of-thought. Advances in Neural Information Processing Systems, 37:54872– 54904. [find](https://scholar.google.com/scholar?q=Qiguang+Chen%2C+Libo+Qin%2C+Jiaqi+Wang%2C+Jinxuan+Zhou%2C+and+Wanxiang+Che%2E+2024b%2E+Unlocking+the+capabilities+of+thought%3A+A+reasoning+boundary+framework+to+quantify+and+optimize+chain%2Dof%2Dthought%2E+Advances+in+)
10. Jiahai Feng, Stuart Russell, and Jacob Steinhardt. 2024. Monitoring latent world states in language models with propositional probes. arXiv preprint arXiv:2406.19501. [find](https://scholar.google.com/scholar?q=Jiahai+Feng%2C+Stuart+Russell%2C+and+Jacob+Steinhardt%2E+2024%2E+Monitoring+latent+world+states+in+language+models+with+propositional+probes%2E+arXiv+preprint+arXiv%3A2406%2E19501%2E)
11. Sangmin Bae, Jongwoo Ko, Hwanjun Song, and Se-Young Yun. 2023. Fast and robust early-exiting framework for autoregressive language models with synchronized parallel decoding. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 5910–5924. [find](https://scholar.google.com/scholar?q=Sangmin+Bae%2C+Jongwoo+Ko%2C+Hwanjun+Song%2C+and+Se%2DYoung+Yun%2E+2023%2E+Fast+and+robust+early%2Dexiting+framework+for+autoregressive+language+models+with+synchronized+parallel+decoding%2E+In+Proceedings+of+the+202)
12. Asma Ghandeharioun, Avi Caciularu, Adam Pearce, Lucas Dixon, and Mor Geva. 2024. [Patchscopes: A unifying framework for inspecting hidden representations of language models](https://md.research.elamur.ai/free/en/papers/ghandeharioun-2024-patchscopes-unifying-framework-inspecting-hidden/). arXiv preprint arXiv:2401.06102.
13. Lucas Bandarkar, Davis Liang, Benjamin Muller, Mikel Artetxe, Satya Narayan Shukla, Donald Husa, Naman Goyal, Abhinandan Krishnan, Luke Zettlemoyer, and Madian Khabsa. 2024. The belebele benchmark: a Akash Ghosh, Debayan Datta, Sriparna Saha, and Chirag Agarwal. 2025. A survey of multilingual reason… [find](https://scholar.google.com/scholar?q=Lucas+Bandarkar%2C+Davis+Liang%2C+Benjamin+Muller%2C+Mikel+Artetxe%2C+Satya+Narayan+Shukla%2C+Donald+Husa%2C+Naman+Goyal%2C+Abhinandan+Krishnan%2C+Luke+Zettlemoyer%2C+and+Madian+Khabsa%2E+2024%2E+The+belebele+benchmark%3A+a+)
14. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, Runxin Xu, Ruoyu Zhang, Shirong Ma, Xiao Bi, et al. 2025. [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/). arXiv preprint arXiv:2501.12948.
15. Shibo Hao, Sainbayar Sukhbaatar, DiJia Su, Xian Li, Zhiting Hu, Jason Weston, and Yuandong Tian. 2024. Training large language models to reason in a continuous latent space. arXiv preprint arXiv:2412.06769. [find](https://scholar.google.com/scholar?q=Shibo+Hao%2C+Sainbayar+Sukhbaatar%2C+DiJia+Su%2C+Xian+Li%2C+Zhiting+Hu%2C+Jason+Weston%2C+and+Yuandong+Tian%2E+2024%2E+Training+large+language+models+to+reason+in+a+continuous+latent+space%2E+arXiv+preprint+arXiv%3A2412%2E)
16. Yifei He, Alon Benhaim, Barun Patra, Praneetha Vaddamanu, Sanchit Ahuja, Parul Chopra, Vishrav Chaudhary, Han Zhao, and Xia Song. 2025. Scaling laws for multilingual language models. In Findings of the Association for Computational Linguistics: ACL 2025, pages 4257–4273. [find](https://scholar.google.com/scholar?q=Yifei+He%2C+Alon+Benhaim%2C+Barun+Patra%2C+Praneetha+Vaddamanu%2C+Sanchit+Ahuja%2C+Parul+Chopra%2C+Vishrav+Chaudhary%2C+Han+Zhao%2C+and+Xia+Song%2E+2025%2E+Scaling+laws+for+multilingual+language+models%2E+In+Findings+of+th)
17. Haoyang Huang, Tianyi Tang, Dongdong Zhang, Wayne Xin Zhao, Ting Song, Yan Xia, and Furu Wei. 2023. [Not all languages are created equal in llms: Improving multilingual capability by cross-lingual-thought prompting](https://md.research.elamur.ai/free/en/papers/huang-2023-all-languages-created-equal-llms/). In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 12365…
18. Kaiyu Huang, Fengran Mo, Xinyu Zhang, Hongliang Li, You Li, Yuanchi Zhang, Weijian Yi, Yulong Mao, Jinchen Liu, Yuzhuang Xu, et al. 2026. A survey on large language models with multilingualism: Recent advances and new frontiers. Artificial Intelligence Review. [find](https://scholar.google.com/scholar?q=Kaiyu+Huang%2C+Fengran+Mo%2C+Xinyu+Zhang%2C+Hongliang+Li%2C+You+Li%2C+Yuanchi+Zhang%2C+Weijian+Yi%2C+Yulong+Mao%2C+Jinchen+Liu%2C+Yuzhuang+Xu%2C+et+al%2E+2026%2E+A+survey+on+large+language+models+with+multilingualism%3A+Recent)
19. Ammar Khairi, Daniel D’souza, Ye Shen, Julia Kreutzer, and Sara Hooker. 2025. [When life gives you samples: The benefits of scaling up inference compute for multilingual llms](https://md.research.elamur.ai/free/en/papers/khairi-2025-when-life-gives-you-samples/). In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 27547–27571.
20. Takeshi Kojima, Shixiang Shane Gu, Machel Reid, Yutaka Matsuo, and Yusuke Iwasawa. 2022. [Large language models are zero-shot reasoners](https://md.research.elamur.ai/free/en/papers/kojima-2022-large-language-models-zero-shot/). Advances in neural information processing systems, 35:22199– 22213.
21. Hele-Andra Kuulmets, Taido Purason, Agnes Luhtaru, and Mark Fishel. 2024. Teaching llama a new language through cross-lingual knowledge transfer. In Findings of the Association for Computational Linguistics: NAACL 2024, pages 3309–3325. [find](https://scholar.google.com/scholar?q=Hele%2DAndra+Kuulmets%2C+Taido+Purason%2C+Agnes+Luhtaru%2C+and+Mark+Fishel%2E+2024%2E+Teaching+llama+a+new+language+through+cross%2Dlingual+knowledge+transfer%2E+In+Findings+of+the+Association+for+Computational+Lingu)
22. Huiyuan Lai, Xiaoyu Zhang, and Malvina Nissim. 2025. Multidimensional consistency improves reasoning in language models. arXiv preprint arXiv:2503.02670. — [arXiv:2503.02670](https://arxiv.org/abs/2503.02670) [find](https://scholar.google.com/scholar?q=Huiyuan+Lai%2C+Xiaoyu+Zhang%2C+and+Malvina+Nissim%2E+2025%2E+Multidimensional+consistency+improves+reasoning+in+language+models%2E+arXiv+preprint+arXiv%3A2503%2E02670%2E+%E2%80%94+%5BarXiv%3A2503%2E02670%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F250)
23. Viet Dac Lai, Nghia Ngo, Amir Pouran Ben Veyseh, Hieu Man, Franck Dernoncourt, Trung Bui, and Thien Huu Nguyen. 2023. [Chatgpt beyond english: Towards a comprehensive evaluation of large language models in multilingual learning](https://md.research.elamur.ai/free/en/papers/lai-2023-chatgpt-beyond-english-comprehensive-evaluation/). In Findings of the Association for Computational Linguistics: EMNLP 2023…
24. Haitao Li, Qian Dong, Junjie Chen, Huixue Su, Yujia Zhou, Qingyao Ai, Ziyi Ye, and Yiqun Liu. 2024. [Llms-as-judges: a comprehensive survey on llm-based evaluation methods](https://md.research.elamur.ai/free/en/papers/li-2024-llms-judges-comprehensive-survey-llm/). arXiv preprint arXiv:2412.05579.
25. OpenAI. 2024. Gpt-4o mini: advancing cost-efficient intelligence. OpenAI Blog. Accessed: 2026-01-05. [find](https://scholar.google.com/scholar?q=OpenAI%2E+2024%2E+Gpt%2D4o+mini%3A+advancing+cost%2Defficient+intelligence%2E+OpenAI+Blog%2E+Accessed%3A+2026%2D01%2D05%2E)
26. Libo Qin, Qiguang Chen, Xiachong Feng, Yang Wu, Yongheng Zhang, Yinghui Li, Min Li, Wanxiang Che, and Philip S Yu. 2026. Large language models meet nlp: A survey. Frontiers of Computer Science, 20(11):2011361. [find](https://scholar.google.com/scholar?q=Libo+Qin%2C+Qiguang+Chen%2C+Xiachong+Feng%2C+Yang+Wu%2C+Yongheng+Zhang%2C+Yinghui+Li%2C+Min+Li%2C+Wanxiang+Che%2C+and+Philip+S+Yu%2E+2026%2E+Large+language+models+meet+nlp%3A+A+survey%2E+Frontiers+of+Computer+Science%2C+20%2811%29)
27. Libo Qin, Qiguang Chen, Fuxuan Wei, Shijue Huang, and Wanxiang Che. 2023. [Cross-lingual prompting: Improving zero-shot chain-of-thought reasoning across languages](https://md.research.elamur.ai/free/en/papers/qin-2023-cross-lingual-prompting-improving-zero/). In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 2695–2709.
28. Libo Qin, Qiguang Chen, Yuhang Zhou, Zhi Chen, Yinghui Li, Lizi Liao, Min Li, Wanxiang Che, and Philip S Yu. 2025. A survey of multilingual large language models. Patterns, 6(1). [find](https://scholar.google.com/scholar?q=Libo+Qin%2C+Qiguang+Chen%2C+Yuhang+Zhou%2C+Zhi+Chen%2C+Yinghui+Li%2C+Lizi+Liao%2C+Min+Li%2C+Wanxiang+Che%2C+and+Philip+S+Yu%2E+2025%2E+A+survey+of+multilingual+large+language+models%2E+Patterns%2C+6%281%29%2E)
29. Sara Rajaee, Rochelle Choenni, Ekaterina Shutova, et al. 2025. Best-of-l: Cross-lingual reward modeling for mathematical reasoning. arXiv preprint arXiv:2509.15811. [find](https://scholar.google.com/scholar?q=Sara+Rajaee%2C+Rochelle+Choenni%2C+Ekaterina+Shutova%2C+et+al%2E+2025%2E+Best%2Dof%2Dl%3A+Cross%2Dlingual+reward+modeling+for+mathematical+reasoning%2E+arXiv+preprint+arXiv%3A2509%2E15811%2E)
30. Leonardo Ranaldi, Giulia Pucci, Federico Ranaldi, Elena Sofia Ruzzetti, and Fabio Massimo Zanzotto. 2024. A tree-of-thoughts to broaden multi-step reasoning across languages. In Findings of the Association for Computational Linguistics: NAACL 2024, pages 1229–1241. [find](https://scholar.google.com/scholar?q=Leonardo+Ranaldi%2C+Giulia+Pucci%2C+Federico+Ranaldi%2C+Elena+Sofia+Ruzzetti%2C+and+Fabio+Massimo+Zanzotto%2E+2024%2E+A+tree%2Dof%2Dthoughts+to+broaden+multi%2Dstep+reasoning+across+languages%2E+In+Findings+of+the+Associ)
31. Matthew Renze and Erhan Guven. 2024. The benefits of a concise chain of thought on problem-solving in large language models. arXiv preprint arXiv:2401.05618. — [arXiv:2401.05618](https://arxiv.org/abs/2401.05618) [find](https://scholar.google.com/scholar?q=Matthew+Renze+and+Erhan+Guven%2E+2024%2E+The+benefits+of+a+concise+chain+of+thought+on+problem%2Dsolving+in+large+language+models%2E+arXiv+preprint+arXiv%3A2401%2E05618%2E+%E2%80%94+%5BarXiv%3A2401%2E05618%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs)
32. Lucas Resck, Isabelle Augenstein, and Anna Korhonen. 2025. [Explainability and interpretability of multilingual large language models: A survey](https://md.research.elamur.ai/free/en/papers/resck-2025-explainability-interpretability-multilingual-large-language/). In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 20465–20497.
33. Freda Shi, Mirac Suzgun, Markus Freitag, Xuezhi Wang, Suraj Srivats, Soroush Vosoughi, Hyung Won Chung, Yi Tay, Sebastian Ruder, Denny Zhou, et al. 2022. [Language models are multilingual chain-of-thought reasoners](https://md.research.elamur.ai/free/en/papers/shi-2022-language-models-multilingual-chain-thought/). arXiv preprint arXiv:2210.03057.
34. Oscar Skean, Md Rifat Arefin, Dan Zhao, Niket Patel, Jalal Naghiyev, Yann LeCun, and Ravid Shwartz-Ziv. 2025. [Layer by layer: Uncovering hidden representations in language models](https://md.research.elamur.ai/free/en/papers/skean-2025-layer-layer-uncovering-hidden-representations/). arXiv preprint arXiv:2502.02013.
35. Khanh-Tung Tran, Barry O’Sullivan, and Hoang D Nguyen. 2025a. Scaling test-time compute for low-resource languages: Multilingual reasoning in llms. arXiv e-prints, pages arXiv–2504. [find](https://scholar.google.com/scholar?q=Khanh%2DTung+Tran%2C+Barry+O%E2%80%99Sullivan%2C+and+Hoang+D+Nguyen%2E+2025a%2E+Scaling+test%2Dtime+compute+for+low%2Dresource+languages%3A+Multilingual+reasoning+in+llms%2E+arXiv+e%2Dprints%2C+pages+arXiv%E2%80%932504%2E)
36. Khanh-Tung Tran, Nguyet-Hang Vu, Barry O’Sullivan, and Harry Nguyen. 2025b. Disentangling language understanding and reasoning structures in cross-lingual chain-of-thought prompting. In EMNLP 2025. [find](https://scholar.google.com/scholar?q=Khanh%2DTung+Tran%2C+Nguyet%2DHang+Vu%2C+Barry+O%E2%80%99Sullivan%2C+and+Harry+Nguyen%2E+2025b%2E+Disentangling+language+understanding+and+reasoning+structures+in+cross%2Dlingual+chain%2Dof%2Dthought+prompting%2E+In+EMNLP+2025%2E)
37. Bibek Upadhayay and Vahid Behzadan. 2023. [Taco: Enhancing cross-lingual transfer for low-resource languages in llms through translation-assisted chain-of-thought processes](https://md.research.elamur.ai/free/en/papers/upadhayay-2023-taco-enhancing-cross-lingual-transfer/). arXiv preprint arXiv:2311.10797.
38. Teng Wang, Zhenqi He, Wing-Yin Yu, Xiaojin Fu, and Xiongwei Han. 2025a. Large language models are good multi-lingual learners: When llms meet cross-lingual prompts. In Proceedings of the 31st International Conference on Computational Linguistics, pages 4442–4456. [find](https://scholar.google.com/scholar?q=Teng+Wang%2C+Zhenqi+He%2C+Wing%2DYin+Yu%2C+Xiaojin+Fu%2C+and+Xiongwei+Han%2E+2025a%2E+Large+language+models+are+good+multi%2Dlingual+learners%3A+When+llms+meet+cross%2Dlingual+prompts%2E+In+Proceedings+of+the+31st+Internat)
39. Weiqin Wang, Yile Wang, and Hui Huang. 2025b. Ranked voting based self-consistency of large language models. In Findings of the Association for Computational Linguistics: ACL 2025, pages 14410– 14426. [find](https://scholar.google.com/scholar?q=Weiqin+Wang%2C+Yile+Wang%2C+and+Hui+Huang%2E+2025b%2E+Ranked+voting+based+self%2Dconsistency+of+large+language+models%2E+In+Findings+of+the+Association+for+Computational+Linguistics%3A+ACL+2025%2C+pages+14410%E2%80%93+14426%2E)
40. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc Le, Ed Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2022. [Self-consistency improves chain of thought reasoning in language models](https://md.research.elamur.ai/free/en/papers/wang-2022-self-consistency-improves-chain-thought/). arXiv preprint arXiv:2203.11171.
41. Yiming Wang, Pei Zhang, Siyuan Huang, Baosong Yang, Zhuosheng Zhang, Fei Huang, and Rui Wang. 2025c. Sampling-efficient test-time scaling: Self-estimating the best-of-n sampling in early decoding. arXiv preprint arXiv:2503.01422. [find](https://scholar.google.com/scholar?q=Yiming+Wang%2C+Pei+Zhang%2C+Siyuan+Huang%2C+Baosong+Yang%2C+Zhuosheng+Zhang%2C+Fei+Huang%2C+and+Rui+Wang%2E+2025c%2E+Sampling%2Defficient+test%2Dtime+scaling%3A+Self%2Destimating+the+best%2Dof%2Dn+sampling+in+early+decoding%2E+arX)
42. Yiming Wang, Pei Zhang, Jialong Tang, Haoran Wei, Baosong Yang, Rui Wang, Chenshu Sun, Feitong Sun, Jiran Zhang, Junxuan Wu, et al. 2025d. Polymath: Evaluating mathematical reasoning in multilingual contexts. arXiv preprint arXiv:2504.18428. — [arXiv:2504.18428](https://arxiv.org/abs/2504.18428) [find](https://scholar.google.com/scholar?q=Yiming+Wang%2C+Pei+Zhang%2C+Jialong+Tang%2C+Haoran+Wei%2C+Baosong+Yang%2C+Rui+Wang%2C+Chenshu+Sun%2C+Feitong+Sun%2C+Jiran+Zhang%2C+Junxuan+Wu%2C+et+al%2E+2025d%2E+Polymath%3A+Evaluating+mathematical+reasoning+in+multilingual+c)
43. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, et al. 2022. [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/). Advances in neural information processing systems, 35:24824–24837.
44. Marrese-Taylor, Shijian Lu, Yusuke Iwasawa, Yutaka Matsuo, and Irene Li. 2025. [MMLU-ProX: A multilingual benchmark for advanced large language model evaluation](https://md.research.elamur.ai/free/en/papers/xuan-2025-mmlu-prox-multilingual-benchmark-advanced/). In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 1513–1532, Suzhou, China. Association for…
45. Rui Yang, Ruomeng Ding, Yong Lin, Huan Zhang, and Tong Zhang. 2024. Regularizing hidden states enables learning generalizable reward model for llms. Advances in Neural Information Processing Systems, 37:62279–62309. [find](https://scholar.google.com/scholar?q=Rui+Yang%2C+Ruomeng+Ding%2C+Yong+Lin%2C+Huan+Zhang%2C+and+Tong+Zhang%2E+2024%2E+Regularizing+hidden+states+enables+learning+generalizable+reward+model+for+llms%2E+Advances+in+Neural+Information+Processing+Systems%2C+)
46. Anqi Zhang, Yulin Chen, Jane Pan, Chen Zhao, Aurojit Panda, Jinyang Li, and He He. 2025a. Reasoning models know when they’re right: Probing hidden states for self-verification. arXiv preprint arXiv:2504.05419. [find](https://scholar.google.com/scholar?q=Anqi+Zhang%2C+Yulin+Chen%2C+Jane+Pan%2C+Chen+Zhao%2C+Aurojit+Panda%2C+Jinyang+Li%2C+and+He+He%2E+2025a%2E+Reasoning+models+know+when+they%E2%80%99re+right%3A+Probing+hidden+states+for+self%2Dverification%2E+arXiv+preprint+arXiv%3A25)
47. Qiyuan Zhang, Fuyuan Lyu, Zexu Sun, Lei Wang, Weixu Zhang, Wenyue Hua, Haolun Wu, Zhihan Guo, Yufei Wang, Niklas Muennighoff, et al. 2025b. [A survey on test-time scaling in large language models: What, how, where, and how well?](https://md.research.elamur.ai/free/en/papers/zhang-2025-survey-test-time-scaling-large/) arXiv preprint arXiv:2503.24235.
48. Yongheng Zhang, Qiguang Chen, Min Li, Wanxiang Che, and Libo Qin. 2024. Autocap: Towards automatic cross-lingual alignment planning for zero-shot chain-of-thought. In Findings of the Association for Computational Linguistics: ACL 2024, pages 9191– 9200. [find](https://scholar.google.com/scholar?q=Yongheng+Zhang%2C+Qiguang+Chen%2C+Min+Li%2C+Wanxiang+Che%2C+and+Libo+Qin%2E+2024%2E+Autocap%3A+Towards+automatic+cross%2Dlingual+alignment+planning+for+zero%2Dshot+chain%2Dof%2Dthought%2E+In+Findings+of+the+Association+for+C)
49. Weixiang Zhao, Jiahe Guo, Yang Deng, Tongtong Wu, Wenxuan Zhang, Yulin Hu, Xingyu Sui, Yanyan Zhao, Wanxiang Che, Bing Qin, et al. 2025. When less language is more: Language-reasoning disentanglement makes llms better multilingual reasoners. arXiv preprint arXiv:2505.15257. — [arXiv:2505.15257](https://arxiv.org/abs/2505.15257) [find](https://scholar.google.com/scholar?q=Weixiang+Zhao%2C+Jiahe+Guo%2C+Yang+Deng%2C+Tongtong+Wu%2C+Wenxuan+Zhang%2C+Yulin+Hu%2C+Xingyu+Sui%2C+Yanyan+Zhao%2C+Wanxiang+Che%2C+Bing+Qin%2C+et+al%2E+2025%2E+When+less+language+is+more%3A+Language%2Dreasoning+disentanglement+)
50. Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric Xing, et al. 2023. [Judging llm-as-a-judge with mt-bench and chatbot arena](https://md.research.elamur.ai/free/en/papers/zheng-2023-judging-llm-judge-mt-bench/). Advances in neural information processing systems, 36:46595–46623.
51. Zheng Weihua, Roy Ka-Wei Lee, Zhengyuan Liu, Wu Kui, Aiti Aw, and Bowei Zou. 2025. Ccl-xcot: An efficient cross-lingual knowledge transfer method for mitigating hallucination generation. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 1768–1788. [find](https://scholar.google.com/scholar?q=Zheng+Weihua%2C+Roy+Ka%2DWei+Lee%2C+Zhengyuan+Liu%2C+Wu+Kui%2C+Aiti+Aw%2C+and+Bowei+Zou%2E+2025%2E+Ccl%2Dxcot%3A+An+efficient+cross%2Dlingual+knowledge+transfer+method+for+mitigating+hallucination+generation%2E+In+Findings+o)
52. Silei Xu, Wenhao Xie, Lingxiao Zhao, and Pengcheng He. 2025. [Chain of draft: Thinking faster by writing less](https://md.research.elamur.ai/free/en/papers/xu-2025-chain-draft-thinking-faster-writing/). arXiv preprint arXiv:2502.18600.
53. Weihao Xuan, Rui Yang, Heli Qi, Qingcheng Zeng, Yunze Xiao, Aosong Feng, Dairui Liu, Yun Xing, Junjue Wang, Fan Gao, Jinghui Lu, Yuang Jiang, Huitao Li, Xin Li, Kunyu Yu, Ruihai Dong, Shangding Gu, Yuekang Li, Xiaofei Xie, Felix Juefei-Xu, Foutse Khomh, Osamu Yoshie, Qingyu Chen, Douglas Teodoro, Na… [find](https://scholar.google.com/scholar?q=Weihao+Xuan%2C+Rui+Yang%2C+Heli+Qi%2C+Qingcheng+Zeng%2C+Yunze+Xiao%2C+Aosong+Feng%2C+Dairui+Liu%2C+Yun+Xing%2C+Junjue+Wang%2C+Fan+Gao%2C+Jinghui+Lu%2C+Yuang+Jiang%2C+Huitao+Li%2C+Xin+Li%2C+Kunyu+Yu%2C+Ruihai+Dong%2C+Shangding+Gu%2C+Yu)

*Updated 2 November 2026*