--- title: "Language of Thought Shapes Output Diversity in Large" description: "This offers a training-free diversity technique that is relevant to multilingual and non-English settings. 62 references, cited by 2 papers here." authors: ["Shaoyang Xu", "Wenxuan 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.628/" doi: "10.18653/v1/2026.acl-long.628" 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/xu-2026-language-thought-shapes-output-diversity/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/xu-2026-language-thought-shapes-output-diversity/) [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/8/) / Language of Thought Shapes Output Diversity in Large Language Models # Language of Thought Shapes Output Diversity in Large Language Models **Authors:** [Shaoyang Xu](https://md.research.elamur.ai/free/en/authors/shaoyang-xu/) and [Wenxuan Zhang](https://md.research.elamur.ai/free/en/authors/wenxuan-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.628/) · [PDF](https://aclanthology.org/2026.acl-long.628.pdf) · [DOI 10.18653/v1/2026.acl-long.628](https://doi.org/10.18653/v1/2026.acl-long.628) · [OpenAlex](https://openalex.org/W7166831138) **Citations:** 0 citations (OpenAlex, 2026-10-14) · [2](https://md.research.elamur.ai/free/en/papers/xu-2026-language-thought-shapes-output-diversity/cited-by/) in our library, of the 149,439 papers whose bibliographies we read ## Summary Large language models often give near-identical answers to open-ended questions. This homogenization limits creativity and pluralistic alignment. The authors test whether the language a reasoning model thinks in, inside its \ block, can act as a controllable source of diversity. They steer the thinking language by inserting a translated opening phrase right after the \ token. They also force the final answer into English so that diversity can be compared fairly. A preliminary analysis of Qwen3-8B hidden states, using cosine distance to English and PCA, shows that 15 thinking languages occupy separate regions at different distances from English. The study then compares two strategies on two benchmarks with two diversity metrics. Single-Language Sampling repeats samples in one thinking language, while Mixed-Language Sampling pools samples across languages. Thinking in a non-English language reliably increases diversity, and languages farther from English help more, with only minor quality loss. Mixing languages brings further gains and raises the diversity ceiling as sampling scales. On cultural tasks, mixed thinking languages cover more cultural knowledge and values than common baselines. This offers a training-free diversity technique that is relevant to multilingual and non-English settings. ## Key points - **Language control:** The thinking language is set with a translated version of the prefix "Okay, the user is asking". A separate English prefix after \ keeps outputs in English. The experiments cover 15 thinking languages, which are English plus 14 languages officially supported by Qwen3-8B. - **Distance predicts diversity:** Thinking-space distance is measured as the cosine distance between averaged hidden states of each language and those of English. This distance correlates positively with output diversity, so languages "farther" from English give larger gains. - **Compositional effects:** In Mixed-Language Sampling, dropping any single language has little effect, but removing several languages degrades diversity substantially. - **Interaction with temperature:** Language of thought and temperature combine. Mixed-Language Sampling at temperature 1.0 reaches roughly the diversity of English sampling at 2.0, with M=15 samples. - **Pluralistic alignment test:** The evaluation uses BLEnD and WVS, scoring entropy over country and answer distributions. Models tested are Qwen3-8B/14B/32B and DeepSeek-R1-Distill-Llama-8B. Mixed-Language Sampling beats English sampling, high temperature, explicit diversity requests, and Multilingual Prompting. Code is available at github.com/iNLP-Lab/Multilingual-LoT-Diversity. *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 - [Singapore University of Technology and Design](https://md.research.elamur.ai/free/en/universities/singapore-university-of-technology-and-design/) — University, Singapore, Singapore · 49 works · 11,074 citations 📚 Shelves (5): [🖥️ 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 · [👁️ Computer Vision and Pattern Recognition](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-vision-and-pattern-recognition/) · 2,869 works 🧭 Topics (1): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works ## Cite Cite as: Shaoyang Xu and Wenxuan Zhang. 2026. Language of Thought Shapes Output Diversity in Large Language Models. 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{xu2026language, title = {{Language of Thought Shapes Output Diversity in Large Language Models}}, author = {Shaoyang Xu and Wenxuan 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.628}, url = {https://doi.org/10.18653/v1/2026.acl-long.628} } ``` - [← previous LADR: Locality-Aware Dynamic Rescue for Efficient Text-to-Image Generation with Diffusion Large Language Models](https://md.research.elamur.ai/free/en/papers/wang-2026-ladr-locality-aware-dynamic-rescue/) - [next → Language Reconstruction with Brain Predictive Coding from fMRI Data](https://md.research.elamur.ai/free/en/papers/yin-2026-language-reconstruction-brain-predictive-coding/) - [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/8/) ## References 62 references, 26 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. 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, Suzhou, China, November 4-9, 2025, pages 15612–15624. Association for Computational… [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) 2. Badr AlKhamissi, Muhammad N. ElNokrashy, Mai Alkhamissi, and Mona T. Diab. 2024. [Investigating cultural alignment of large language models](https://md.research.elamur.ai/free/en/papers/alkhamissi-2024-investigating-cultural-alignment-large-language/). In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2024, Bangkok, Thailand, August 11-16, 2… 3. Prasoon Bajpai and Tanmoy Chakraborty. 2025. Multilingual test-time scaling via initial thought transfer. CoRR, abs/2505.15508. — [arXiv:2505.15508](https://arxiv.org/abs/2505.15508) [find](https://scholar.google.com/scholar?q=Prasoon+Bajpai+and+Tanmoy+Chakraborty%2E+2025%2E+Multilingual+test%2Dtime+scaling+via+initial+thought+transfer%2E+CoRR%2C+abs%2F2505%2E15508%2E+%E2%80%94+%5BarXiv%3A2505%2E15508%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2505%2E15508%29) 4. Damián E. Blasi, Joseph Henrich, Evangelia Adamou, David Kemmerer, and Asifa Majid. 2022. Over-reliance on english hinders cognitive science. Trends in Cognitive Sciences, 26(12):1153–1170. [find](https://scholar.google.com/scholar?q=Dami%C3%A1n+E%2E+Blasi%2C+Joseph+Henrich%2C+Evangelia+Adamou%2C+David+Kemmerer%2C+and+Asifa+Majid%2E+2022%2E+Over%2Dreliance+on+english+hinders+cognitive+science%2E+Trends+in+Cognitive+Sciences%2C+26%2812%29%3A1153%E2%80%931170%2E) 5. Yong Cao, Li Zhou, Seolhwa Lee, Laura Cabello, Min Chen, and Daniel Hershcovich. 2023. [Assessing cross-cultural alignment between chatgpt and human societies: An empirical study](https://md.research.elamur.ai/free/en/papers/cao-2023-assessing-cross-cultural-alignment-between/). CoRR, abs/2303.17466. 6. Kang Chen, Mengdi Zhang, and Yixin Cao. 2025a. Less data less tokens: Multilingual unification learning for efficient test-time reasoning in llms. CoRR, abs/2506.18341. [find](https://scholar.google.com/scholar?q=Kang+Chen%2C+Mengdi+Zhang%2C+and+Yixin+Cao%2E+2025a%2E+Less+data+less+tokens%3A+Multilingual+unification+learning+for+efficient+test%2Dtime+reasoning+in+llms%2E+CoRR%2C+abs%2F2506%2E18341%2E) 7. Xiaoyang Chen, Xinan Dai, Yu Du, Qian Feng, Naixu Guo, Tingshuo Gu, Yuting Gao, Yingyi Gao, Xudong Han, Xiang Jiang, Yilin Jin, Hongyi Lin, Shisheng Lin, Xiangnan Li, Yuante Li, Yixing Li, Zhentao Lai, Zilu Ma, Yingrong Peng, and 12 others. 2025b. Deepmath-creative: A benchmark for evaluating mathem… — [arXiv:2505.08744](https://arxiv.org/abs/2505.08744) [find](https://scholar.google.com/scholar?q=Xiaoyang+Chen%2C+Xinan+Dai%2C+Yu+Du%2C+Qian+Feng%2C+Naixu+Guo%2C+Tingshuo+Gu%2C+Yuting+Gao%2C+Yingyi+Gao%2C+Xudong+Han%2C+Xiang+Jiang%2C+Yilin+Jin%2C+Hongyi+Lin%2C+Shisheng+Lin%2C+Xiangnan+Li%2C+Yuante+Li%2C+Yixing+Li%2C+Zhentao+Lai) 8. Yuxin Chen, Yiran Zhao, Yang Zhang, An Zhang, Kenji Kawaguchi, Shafiq Joty, Junnan Li, Tat-Seng Chua, Michael Qizhe Shieh, and Wenxuan Zhang. 2025c. The emergence of abstract thought in large language models beyond any language. CoRR, abs/2506.09890. — [arXiv:2506.09890](https://arxiv.org/abs/2506.09890) [find](https://scholar.google.com/scholar?q=Yuxin+Chen%2C+Yiran+Zhao%2C+Yang+Zhang%2C+An+Zhang%2C+Kenji+Kawaguchi%2C+Shafiq+Joty%2C+Junnan+Li%2C+Tat%2DSeng+Chua%2C+Michael+Qizhe+Shieh%2C+and+Wenxuan+Zhang%2E+2025c%2E+The+emergence+of+abstract+thought+in+large+language) 9. Vincent Conitzer, Rachel Freedman, Jobst Heitzig, Wesley H. Holliday, Bob M. Jacobs, Nathan Lambert, Milan Mossé, Eric Pacuit, Stuart Russell, Hailey Schoelkopf, Emanuel Tewolde, and William S. Zwicker. 2024. Position: Social choice should guide AI alignment in dealing with diverse human feedback. I… [find](https://scholar.google.com/scholar?q=Vincent+Conitzer%2C+Rachel+Freedman%2C+Jobst+Heitzig%2C+Wesley+H%2E+Holliday%2C+Bob+M%2E+Jacobs%2C+Nathan+Lambert%2C+Milan+Moss%C3%A9%2C+Eric+Pacuit%2C+Stuart+Russell%2C+Hailey+Schoelkopf%2C+Emanuel+Tewolde%2C+and+William+S%2E+Zwicke) 10. 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, and 81 others. 2025. [Deepseek-r1: Incentivizing reasoning capability in llm…](https://md.research.elamur.ai/free/en/papers/deepseek-ai-2025-deepseek-r1-incentivizing-reasoning-capability/) 11. Changjiang Gao, Xu Huang, Wenhao Zhu, Shujian Huang, Lei Li, and Fei Yuan. 2025. [Could thinking multilingually empower LLM reasoning?](https://md.research.elamur.ai/free/en/papers/gao-2025-could-thinking-multilingually-empower-llm/) CoRR, abs/2504.11833. 12. Changjiang Gao, Zixian Huang, Kaichen Yang, Jiajun Chen, Jixing Li, and Shujian Huang. 2026. [Explang: Improved exploration and exploitation in LLM reasoning with on-policy thinking language selection](https://md.research.elamur.ai/free/en/papers/gao-2026-explang-improved-exploration-exploitation-llm/). CoRR, abs/2602.21887. 13. Salvatore Giorgi, Tingting Liu, Ankit Aich, Kelsey Isman, Garrick Sherman, Zachary Fried, João Sedoc, Lyle H. Ungar, and Brenda Curtis. 2024. Modeling human subjectivity in llms using explicit and implicit human factors in personas. In Findings of the Association for Computational Linguistics: EMNLP… [find](https://scholar.google.com/scholar?q=Salvatore+Giorgi%2C+Tingting+Liu%2C+Ankit+Aich%2C+Kelsey+Isman%2C+Garrick+Sherman%2C+Zachary+Fried%2C+Jo%C3%A3o+Sedoc%2C+Lyle+H%2E+Ungar%2C+and+Brenda+Curtis%2E+2024%2E+Modeling+human+subjectivity+in+llms+using+explicit+and+imp) 14. Sikun Guo, Amir Hassan Shariatmadari, Guangzhi Xiong, Albert Huang, Myles Kim, Corey M. Williams, Stefan Bekiranov, and Aidong Zhang. 2025a. Ideabench: Benchmarking large language models for research idea generation. In Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mi… [find](https://scholar.google.com/scholar?q=Sikun+Guo%2C+Amir+Hassan+Shariatmadari%2C+Guangzhi+Xiong%2C+Albert+Huang%2C+Myles+Kim%2C+Corey+M%2E+Williams%2C+Stefan+Bekiranov%2C+and+Aidong+Zhang%2E+2025a%2E+Ideabench%3A+Benchmarking+large+language+models+for+research+) 15. Yanzhu Guo, Guokan Shang, and Chloé Clavel. 2025b. [Benchmarking linguistic diversity of large language models](https://md.research.elamur.ai/free/en/papers/guo-2025-benchmarking-linguistic-diversity-large-language/). Trans. Assoc. Comput. Linguistics, 13:1507– 1526. 16. Yanzhu Guo, Guokan Shang, Michalis Vazirgiannis, and Chloé Clavel. 2024. The curious decline of linguistic diversity: Training language models on synthetic text. In Findings of the Association for Computational Linguistics: NAACL 2024, Mexico City, Mexico, June 16-21, 2024, pages 3589–3604. Associat… [find](https://scholar.google.com/scholar?q=Yanzhu+Guo%2C+Guokan+Shang%2C+Michalis+Vazirgiannis%2C+and+Chlo%C3%A9+Clavel%2E+2024%2E+The+curious+decline+of+linguistic+diversity%3A+Training+language+models+on+synthetic+text%2E+In+Findings+of+the+Association+for+Com) 17. Christian Haerpfer, Ronald Inglehart, Alejandro Moreno, Christian Welzel, Kseniya Kizilova, Jaime Diez-Medrano, Marta Lagos, Pippa Norris, Eduard Ponarin, and Bjorn Puranen. 2022. World values survey: Round seven-country-pooled datafile version 5.0. Madrid, Spain & Vienna, Austria: JD Systems Instit… [find](https://scholar.google.com/scholar?q=Christian+Haerpfer%2C+Ronald+Inglehart%2C+Alejandro+Moreno%2C+Christian+Welzel%2C+Kseniya+Kizilova%2C+Jaime+Diez%2DMedrano%2C+Marta+Lagos%2C+Pippa+Norris%2C+Eduard+Ponarin%2C+and+Bjorn+Puranen%2E+2022%2E+World+values+survey%3A) 18. Simeng Han, Stephen Xia, Grant Zhang, Howard Dai, Chen Liu, Lichang Chen, Hoang Huy Nguyen, Hongyuan Mei, Jiayuan Mao, and R. Thomas McCoy. 2025. Creativity or brute force? using brainteasers as a window into the problem-solving abilities of large language models. CoRR, abs/2505.10844. — [arXiv:2505.10844](https://arxiv.org/abs/2505.10844) [find](https://scholar.google.com/scholar?q=Simeng+Han%2C+Stephen+Xia%2C+Grant+Zhang%2C+Howard+Dai%2C+Chen+Liu%2C+Lichang+Chen%2C+Hoang+Huy+Nguyen%2C+Hongyuan+Mei%2C+Jiayuan+Mao%2C+and+R%2E+Thomas+McCoy%2E+2025%2E+Creativity+or+brute+force%3F+using+brainteasers+as+a+win) 19. Liwei Jiang, Yuanjun Chai, Margaret Li, Mickel Liu, Raymond Fok, Nouha Dziri, Yulia Tsvetkov, Maarten Sap, Alon Albalak, and Yejin Choi. 2025. Artificial hivemind: The open-ended homogeneity of language models (and beyond). CoRR, abs/2510.22954. — [arXiv:2510.22954](https://arxiv.org/abs/2510.22954) [find](https://scholar.google.com/scholar?q=Liwei+Jiang%2C+Yuanjun+Chai%2C+Margaret+Li%2C+Mickel+Liu%2C+Raymond+Fok%2C+Nouha+Dziri%2C+Yulia+Tsvetkov%2C+Maarten+Sap%2C+Alon+Albalak%2C+and+Yejin+Choi%2E+2025%2E+Artificial+hivemind%3A+The+open%2Dended+homogeneity+of+langua) 20. 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, EMNLP 2025, Suzhou, China, Novembe… 21. Anatoliy V. Kharkhurin, Valeriya Koncha, and Morteza Charkhabi. 2023. The effects of multilingual and multicultural practices on divergent thinking. implications for plurilingual creativity paradigm. Bilingualism: Language and cognition, 26(3):592–609. [find](https://scholar.google.com/scholar?q=Anatoliy+V%2E+Kharkhurin%2C+Valeriya+Koncha%2C+and+Morteza+Charkhabi%2E+2023%2E+The+effects+of+multilingual+and+multicultural+practices+on+divergent+thinking%2E+implications+for+plurilingual+creativity+paradigm%2E+) 22. Arash Lagzian, Srinivas Anumasa, and Dianbo Liu. 2025. Multi-novelty: Improve the diversity and novelty of contents generated by large language models via inference-time multi-views brainstorming. CoRR, abs/2502.12700. — [arXiv:2502.12700](https://arxiv.org/abs/2502.12700) [find](https://scholar.google.com/scholar?q=Arash+Lagzian%2C+Srinivas+Anumasa%2C+and+Dianbo+Liu%2E+2025%2E+Multi%2Dnovelty%3A+Improve+the+diversity+and+novelty+of+contents+generated+by+large+language+models+via+inference%2Dtime+multi%2Dviews+brainstorming%2E+CoR) 23. Yihao Li, Jiayi Xin, Miranda Muqing Miao, Qi Long, and Lyle H. Ungar. 2025a. The impact of language mixing on bilingual LLM reasoning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, EMNLP 2025, Suzhou, China, November 4-9, 2025, pages 32531–32548. Associat… [find](https://scholar.google.com/scholar?q=Yihao+Li%2C+Jiayi+Xin%2C+Miranda+Muqing+Miao%2C+Qi+Long%2C+and+Lyle+H%2E+Ungar%2E+2025a%2E+The+impact+of+language+mixing+on+bilingual+LLM+reasoning%2E+In+Proceedings+of+the+2025+Conference+on+Empirical+Methods+in+Nat) 24. Ziniu Li, Congliang Chen, Tian Xu, Zeyu Qin, Jiancong Xiao, Zhi-Quan Luo, and Ruoyu Sun. 2025b. Preserving diversity in supervised fine-tuning of large language models. In The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025. OpenReview.net. [find](https://scholar.google.com/scholar?q=Ziniu+Li%2C+Congliang+Chen%2C+Tian+Xu%2C+Zeyu+Qin%2C+Jiancong+Xiao%2C+Zhi%2DQuan+Luo%2C+and+Ruoyu+Sun%2E+2025b%2E+Preserving+diversity+in+supervised+fine%2Dtuning+of+large+language+models%2E+In+The+Thirteenth+International) 25. Tian Liang, Zhiwei He, Wenxiang Jiao, Xing Wang, Yan Wang, Rui Wang, Yujiu Yang, Shuming Shi, and Zhaopeng Tu. 2024a. [Encouraging divergent thinking in large language models through multi-agent debate](https://md.research.elamur.ai/free/en/papers/liang-2023-encouraging-divergent-thinking-large-language/). In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2… 26. Weixin Liang, Yaohui Zhang, Zhengxuan Wu, Haley Lepp, Wenlong Ji, Xuandong Zhao, Hancheng Cao, Sheng Liu, Siyu He, Zhi Huang, Diyi Yang, Christopher Potts, Christopher D. Manning, and James Y. Zou. 2024b. Mapping the increasing use of llms in scientific papers. CoRR, abs/2404.01268. [find](https://scholar.google.com/scholar?q=Weixin+Liang%2C+Yaohui+Zhang%2C+Zhengxuan+Wu%2C+Haley+Lepp%2C+Wenlong+Ji%2C+Xuandong+Zhao%2C+Hancheng+Cao%2C+Sheng+Liu%2C+Siyu+He%2C+Zhi+Huang%2C+Diyi+Yang%2C+Christopher+Potts%2C+Christopher+D%2E+Manning%2C+and+James+Y%2E+Zou%2E+20) 27. Jiaming Luo, Colin Cherry, and George F. Foster. 2024. [To diverge or not to diverge: A morphosyntactic perspective on machine translation vs human translation](https://md.research.elamur.ai/free/en/papers/luo-2024-diverge-or-not-diverge-morphosyntactic/). Trans. Assoc. Comput. Linguistics, 12:355–371. 28. Niklas Muennighoff, Zitong Yang, Weijia Shi, Xiang Lisa Li, Li Fei-Fei, Hannaneh Hajishirzi, Luke Zettlemoyer, Percy Liang, Emmanuel J. Candès, and Tatsunori Hashimoto. 2025. [s1: Simple test-time scaling](https://md.research.elamur.ai/free/en/papers/muennighoff-2025-s1-simple-test-time-scaling/). In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, EMNL… 29. Junho Myung, Nayeon Lee, Yi Zhou, Jiho Jin, Rifki Afina Putri, Dimosthenis Antypas, Hsuvas Borkakoty, Eunsu Kim, Carla Pérez-Almendros, Abinew Ali Ayele, Víctor Gutiérrez-Basulto, Yazmín Ibáñez-García, Hwaran Lee, Shamsuddeen Hassan Muhammad, Ki-Woong Park, Anar Rzayev, Nina White, Seid Muhie Yimam,… — [BLEnD: A Benchmark for LLMs on Everyday Knowledge in Diverse Cultures and Languages](https://md.research.elamur.ai/free/en/papers/myung-2024-blend-benchmark-llms-everyday-knowledge/) 30. Vishakh Padmakumar and He He. 2024. Does writing with language models reduce content diversity? In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024. OpenReview.net. [find](https://scholar.google.com/scholar?q=Vishakh+Padmakumar+and+He+He%2E+2024%2E+Does+writing+with+language+models+reduce+content+diversity%3F+In+The+Twelfth+International+Conference+on+Learning+Representations%2C+ICLR+2024%2C+Vienna%2C+Austria%2C+May+7%2D1) 31. Max Peeperkorn, Tom Kouwenhoven, Dan Brown, and Anna Jordanous. 2024. Is temperature the creativity parameter of large language models? In Proceedings of the 15th International Conference on Computational Creativity, ICCC 2024, Jönköping, Sweden, June 17-21, 2024, pages 226–235. Association for Comp… [find](https://scholar.google.com/scholar?q=Max+Peeperkorn%2C+Tom+Kouwenhoven%2C+Dan+Brown%2C+and+Anna+Jordanous%2E+2024%2E+Is+temperature+the+creativity+parameter+of+large+language+models%3F+In+Proceedings+of+the+15th+International+Conference+on+Computati) 32. Max Peeperkorn, Tom Kouwenhoven, Dan Brown, and Anna Jordanous. 2025. Mind the gap: Con-formative decoding to improve output diversity of instruction-tuned large language models. CoRR, abs/2507.20956. — [arXiv:2507.20956](https://arxiv.org/abs/2507.20956) [find](https://scholar.google.com/scholar?q=Max+Peeperkorn%2C+Tom+Kouwenhoven%2C+Dan+Brown%2C+and+Anna+Jordanous%2E+2025%2E+Mind+the+gap%3A+Con%2Dformative+decoding+to+improve+output+diversity+of+instruction%2Dtuned+large+language+models%2E+CoRR%2C+abs%2F2507%2E20956%2E) 33. Antoine Bellemare Pépin, François Lespinasse, Philipp Thölke, Yann Harel, Kory W. Mathewson, Jay A. Olson, Yoshua Bengio, and Karim Jerbi. 2024. Divergent creativity in humans and large language models. CoRR, abs/2405.13012. — [arXiv:2405.13012](https://arxiv.org/abs/2405.13012) [find](https://scholar.google.com/scholar?q=Antoine+Bellemare+P%C3%A9pin%2C+Fran%C3%A7ois+Lespinasse%2C+Philipp+Th%C3%B6lke%2C+Yann+Harel%2C+Kory+W%2E+Mathewson%2C+Jay+A%2E+Olson%2C+Yoshua+Bengio%2C+and+Karim+Jerbi%2E+2024%2E+Divergent+creativity+in+humans+and+large+language+model) 34. Telmo Pires, Eva Schlinger, and Dan Garrette. 2019. [How multilingual is multilingual bert?](https://md.research.elamur.ai/free/en/papers/pires-2019-how-multilingual-is-multilingual-bert/) In Proceedings of the 57th Conference of the Association for Computational Linguistics, ACL 2019, Florence, Italy, July 28- August 2, 2019, Volume 1: Long Papers, pages 4996–5001. Association for Computational… 35. Jirui Qi, Shan Chen, Zidi Xiong, Raquel Fernández, Danielle S. Bitterman, and Arianna Bisazza. 2025. When models reason in your language: Controlling thinking language comes at the cost of accuracy. In Findings of the Association for Computational Linguistics: EMNLP 2025, Suzhou, China, November 4-9… [find](https://scholar.google.com/scholar?q=Jirui+Qi%2C+Shan+Chen%2C+Zidi+Xiong%2C+Raquel+Fern%C3%A1ndez%2C+Danielle+S%2E+Bitterman%2C+and+Arianna+Bisazza%2E+2025%2E+When+models+reason+in+your+language%3A+Controlling+thinking+language+comes+at+the+cost+of+accuracy%2E+I) 36. Nils Reimers and Iryna Gurevych. 2019. [Sentence-bert: Sentence embeddings using siamese bert-networks](https://md.research.elamur.ai/free/en/papers/reimers-2019-sentence-bert-sentence-embeddings-siamese/). In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, EMNLP-IJCNLP 2019, Hong Kong, Ch… 37. Kai Ruan, Xuan Wang, Jixiang Hong, Peng Wang, Yang Liu, and Hao Sun. 2024. [Liveideabench: Evaluating llms’ divergent thinking for scientific idea generation with minimal context](https://md.research.elamur.ai/free/en/papers/ruan-2024-evaluating-llms-divergent-thinking-capabilities/). CoRR, abs/2412.17596. 38. Michal Shur-Ofry, Bar Horowitz-Amsalem, Adir Ra-hamim, and Yonatan Belinkov. 2024. Growing a tail: Increasing output diversity in large language models. CoRR, abs/2411.02989. — [arXiv:2411.02989](https://arxiv.org/abs/2411.02989) [find](https://scholar.google.com/scholar?q=Michal+Shur%2DOfry%2C+Bar+Horowitz%2DAmsalem%2C+Adir+Ra%2Dhamim%2C+and+Yonatan+Belinkov%2E+2024%2E+Growing+a+tail%3A+Increasing+output+diversity+in+large+language+models%2E+CoRR%2C+abs%2F2411%2E02989%2E+%E2%80%94+%5BarXiv%3A2411%2E02989%5D%28http) 39. Guijin Son, Jiwoo Hong, Hyunwoo Ko, and James Thorne. 2025. [Linguistic generalizability of test-time scaling in mathematical reasoning](https://md.research.elamur.ai/free/en/papers/son-2025-linguistic-generalizability-test-time-scaling/). In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2025, Vienna, Austria, July 27 - August 1, 2… 40. Taylor Sorensen, Jared Moore, Jillian Fisher, Mitchell L. Gordon, Niloofar Mireshghallah, Christopher Michael Rytting, Andre Ye, Liwei Jiang, Ximing Lu, Nouha Dziri, Tim Althoff, and Yejin Choi. 2024. Position: [A roadmap to pluralistic alignment](https://md.research.elamur.ai/free/en/papers/sorensen-2024-roadmap-pluralistic-alignment/). In Forty-first International Conference on Machine Le… 41. Haoran Sun, Yekun Chai, Shuohuan Wang, Yu Sun, Hua Wu, and Haifeng Wang. 2025. Curiosity-driven reinforcement learning from human feedback. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2025, Vienna, Austria, July 27 -August 1… [find](https://scholar.google.com/scholar?q=Haoran+Sun%2C+Yekun+Chai%2C+Shuohuan+Wang%2C+Yu+Sun%2C+Hua+Wu%2C+and+Haifeng+Wang%2E+2025%2E+Curiosity%2Ddriven+reinforcement+learning+from+human+feedback%2E+In+Proceedings+of+the+63rd+Annual+Meeting+of+the+Association) 42. Zhi Rui Tam, Cheng-Kuang Wu, Yu Ying Chiu, Chieh-Yen Lin, Yun-Nung Chen, and Hung-yi Lee. 2025. [Language matters: How do multilingual input and reasoning paths affect large reasoning models?](https://md.research.elamur.ai/free/en/papers/tam-2025-language-matters-multilingual-input-reasoning/) CoRR, abs/2505.17407. 43. Selim F. Tekin, Fatih Ilhan, Tiansheng Huang, Sihao Hu, and Ling Liu. 2024. LLM-TOPLA: efficient LLM ensemble by maximising diversity. In Findings of the Association for Computational Linguistics: EMNLP 2024, Miami, Florida, USA, November 12-16, 2024, pages 11951–11966. Association for Computational… [find](https://scholar.google.com/scholar?q=Selim+F%2E+Tekin%2C+Fatih+Ilhan%2C+Tiansheng+Huang%2C+Sihao+Hu%2C+and+Ling+Liu%2E+2024%2E+LLM%2DTOPLA%3A+efficient+LLM+ensemble+by+maximising+diversity%2E+In+Findings+of+the+Association+for+Computational+Linguistics%3A+EMN) 44. 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](https://md.research.elamur.ai/free/en/papers/yang-2025-qwen3-technical-report/). CoRR, abs/2505.09388. 45. Guy Tevet and Jonathan Berant. 2021. [Evaluating the evaluation of diversity in natural language generation](https://md.research.elamur.ai/free/en/papers/tevet-2021-evaluating-evaluation-diversity-natural-language/). In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, EACL 2021, Online, April 19 - 23, 2021, pages 326–346. Association… 46. Junyi Ye, Jingyi Gu, Xinyun Zhao, Wenpeng Yin, and Grace Guiling Wang. 2025. [Assessing the creativity of llms in proposing novel solutions to mathematical problems](https://md.research.elamur.ai/free/en/papers/ye-2024-assessing-creativity-llms-proposing-novel/). In AAAI-25, Sponsored by the Association for the Advancement of Artificial Intelligence, February 25 - March 4, 2025, Philadelphia, PA,… 47. Yufei Tian, Abhilasha Ravichander, Lianhui Qin, Ronan Le Bras, Raja Marjieh, Nanyun Peng, Yejin Choi, Thomas L. Griffiths, and Faeze Brahman. 2024. Macgyver: Are large language models creative problem solvers? In Proceedings of the 2024 Conference of the North American Chapter of the Association for… [find](https://scholar.google.com/scholar?q=Yufei+Tian%2C+Abhilasha+Ravichander%2C+Lianhui+Qin%2C+Ronan+Le+Bras%2C+Raja+Marjieh%2C+Nanyun+Peng%2C+Yejin+Choi%2C+Thomas+L%2E+Griffiths%2C+and+Faeze+Brahman%2E+2024%2E+Macgyver%3A+Are+large+language+models+creative+problem) 48. Jiahao Ying, Wei Tang, Yiran Zhao, Yixin Cao, Yu Rong, and Wenxuan Zhang. 2025. [Disentangling language and culture for evaluating multilingual large language models](https://md.research.elamur.ai/free/en/papers/ying-2025-disentangling-language-culture-evaluating-multilingual/). In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2025, Vienna,… 49. Qihan Wang, Shidong Pan, Tal Linzen, and Emily Black. 2025a. Multilingual prompting for improving LLM generation diversity. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, EMNLP 2025, Suzhou, China, November 4-9, 2025, pages 6367–6389. Association for Compu… [find](https://scholar.google.com/scholar?q=Qihan+Wang%2C+Shidong+Pan%2C+Tal+Linzen%2C+and+Emily+Black%2E+2025a%2E+Multilingual+prompting+for+improving+LLM+generation+diversity%2E+In+Proceedings+of+the+2025+Conference+on+Empirical+Methods+in+Natural+Langua) 50. Wenxuan Wang, Wenxiang Jiao, Jingyuan Huang, Ruyi Dai, Jen-tse Huang, Zhaopeng Tu, and Michael R. Lyu. 2024. [Not all countries celebrate thanksgiving: On the cultural dominance in large language models](https://md.research.elamur.ai/free/en/papers/wang-2024-not-all-countries-celebrate-thanksgiving/). In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volum… 51. Yiming Wang, Pei Zhang, Jialong Tang, Haoran Wei, Baosong Yang, Rui Wang, Chenshu Sun, Feitong Sun, Jiran Zhang, Junxuan Wu, Qiqian Cang, Yichang Zhang, Fei Huang, Junyang Lin, Fei Huang, and Jingren Zhou. 2025b. Polymath: Evaluating mathematical reasoning in multilingual contexts. CoRR, abs/2504.18… — [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+Qiqian+Cang%2C+Yichang+Zhang%2C+Fei+Huang%2C+Junyang+Lin%2C+Fei+Huang%2C+and+Jingren+) 52. Benjamin Lee Whorf. 2012. Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf. MIT Press. [find](https://scholar.google.com/scholar?q=Benjamin+Lee+Whorf%2E+2012%2E+Language%2C+Thought%2C+and+Reality%3A+Selected+Writings+of+Benjamin+Lee+Whorf%2E+MIT+Press%2E) 53. Shaoyang Xu, Yongqi Leng, Linhao Yu, and Deyi Xiong. 2025. Self-pluralising culture alignment for large language models. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL 2025 - Volume 1:… [find](https://scholar.google.com/scholar?q=Shaoyang+Xu%2C+Yongqi+Leng%2C+Linhao+Yu%2C+and+Deyi+Xiong%2E+2025%2E+Self%2Dpluralising+culture+alignment+for+large+language+models%2E+In+Proceedings+of+the+2025+Conference+of+the+Nations+of+the+Americas+Chapter+of) 54. Zheng-Xin Yong, Muhammad 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. CoRR, abs/2505.05408. — [arXiv:2505.05408](https://arxiv.org/abs/2505.05408) [find](https://scholar.google.com/scholar?q=Zheng%2DXin+Yong%2C+Muhammad+Farid+Adilazuarda%2C+Jonibek+Mansurov%2C+Ruochen+Zhang%2C+Niklas+Muennighoff%2C+Carsten+Eickhoff%2C+Genta+Indra+Winata%2C+Julia+Kreutzer%2C+Stephen+H%2E+Bach%2C+and+Alham+Fikri+Aji%2E+2025%2E+Cross) 55. Zhiyuan Zeng, Qinyuan Cheng, Zhangyue Yin, Yunhua Zhou, and Xipeng Qiu. 2025. [Revisiting the test-time scaling of o1-like models: Do they truly possess test-time scaling capabilities?](https://md.research.elamur.ai/free/en/papers/zeng-2025-revisiting-test-time-scaling-o1/) In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers),… 56. Wenxuan Zhang. 2026. Towards inclusive AI: advancing multilingual large language models. In Fortieth AAAI Conference on Artificial Intelligence, Thirty-Eighth Conference on Innovative Applications of Artificial Intelligence, Sixteenth Symposium on Educational Advances in Artificial Intelligence, AAA… [find](https://scholar.google.com/scholar?q=Wenxuan+Zhang%2E+2026%2E+Towards+inclusive+AI%3A+advancing+multilingual+large+language+models%2E+In+Fortieth+AAAI+Conference+on+Artificial+Intelligence%2C+Thirty%2DEighth+Conference+on+Innovative+Applications+of+) 57. Wenxuan Zhang, Hou Pong Chan, Yiran Zhao, Mahani Aljunied, Jianyu Wang, Chaoqun Liu, Yue Deng, Zhiqiang Hu, Weiwen Xu, Yew Ken Chia, Xin Li, and Lidong Bing. 2025a. Seallms 3: Open foundation and chat multilingual large language models for southeast asian languages. In Proceedings of the 2025 Confer… [find](https://scholar.google.com/scholar?q=Wenxuan+Zhang%2C+Hou+Pong+Chan%2C+Yiran+Zhao%2C+Mahani+Aljunied%2C+Jianyu+Wang%2C+Chaoqun+Liu%2C+Yue+Deng%2C+Zhiqiang+Hu%2C+Weiwen+Xu%2C+Yew+Ken+Chia%2C+Xin+Li%2C+and+Lidong+Bing%2E+2025a%2E+Seallms+3%3A+Open+foundation+and+chat) 58. Yiming Zhang, Harshita Diddee, Susan Holm, Hanchen Liu, Xinyue Liu, Vinay Samuel, Barry Wang, and Lang Think-Target (%) Output-EN (%) Qwen3-8B en non-en 98.29 98.28 ± 1.31 Qwen3-14B en non-en 100.00 99.88 ± 0.25 Qwen3-32B en non-en 100.00 98.61 ± 0.69 DeepSeek-14B en non-en 100.00 99.54 ± 1.47 100.0… — [arXiv:2504.05228](https://arxiv.org/abs/2504.05228) [find](https://scholar.google.com/scholar?q=Yiming+Zhang%2C+Harshita+Diddee%2C+Susan+Holm%2C+Hanchen+Liu%2C+Xinyue+Liu%2C+Vinay+Samuel%2C+Barry+Wang%2C+and+Lang+Think%2DTarget+%28%25%29+Output%2DEN+%28%25%29+Qwen3%2D8B+en+non%2Den+98%2E29+98%2E28+%C2%B1+1%2E31+Qwen3%2D14B+en+non%2Den+100%2E00+9) 59. Weixiang Zhao, Jiahe Guo, Yang Deng, Tongtong Wu, Wenxuan Zhang, Yulin Hu, Xingyu Sui, Yanyan Zhao, Wanxiang Che, Bing Qin, Tat-Seng Chua, and Ting Liu. 2025a. When less language is more: Language-reasoning disentanglement makes llms better multilingual reasoners. CoRR, abs/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+Tat%2DSeng+Chua%2C+and+Ting+Liu%2E+2025a%2E+When+less+language+is+more%3A+Language%2Drea) 60. Yiran Zhao, Chaoqun Liu, Yue Deng, Jiahao Ying, Mahani Aljunied, Zhaodonghui Li, Lidong Bing, Hou Pong Chan, Yu Rong, Deli Zhao, and Wenxuan Zhang. 2025b. [Babel: Open multilingual large language models serving over 90% of global speakers](https://md.research.elamur.ai/free/en/papers/zhao-2025-babel-open-multilingual-large-language/). CoRR, abs/2503.00865. 61. Yiran Zhao, Wenxuan Zhang, Guizhen Chen, Kenji Kawaguchi, and Lidong Bing. 2024. [How do large language models handle multilingualism?](https://md.research.elamur.ai/free/en/papers/zhao-2024-large-language-models-handle-multilingualism/) In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2024, NeurIPS 2024, Vancouver, BC, Canada, Decembe… 62. Yaoming Zhu, Sidi Lu, Lei Zheng, Jiaxian Guo, Weinan Zhang, Jun Wang, and Yong Yu. 2018. [Texygen: A benchmarking platform for text generation models](https://md.research.elamur.ai/free/en/papers/zhu-2018-texygen-benchmarking-platform-text-generation/). In The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval, SIGIR 2018, Ann Arbor, MI, USA, July 08-12, 2018, p… *Updated 1 November 2026*