---
title: "TFD — paper by Huang et al. (2025) — summary and ideas"
description: "The release gives Tibetan NLP a complete and reproducible training stack. 42 references, cited by 4 papers here."
authors: ["Cheng Huang", "Fan Gao", "Nyima Tashi", "Yutong Liu", "Yadi Liu", "Wenbin Wei", "Xiangxiang Wang", "Yongbin Yu"]
year: 2025
venue: "arXiv:2503.18288"
source: "https://arxiv.org/abs/2503.18288"
doi: "10.48550/arxiv.2503.18288"
arxiv: "2503.18288"
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/huang-2025-tfd-comprehensive-structured-tibetan-foundation/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/huang-2025-tfd-comprehensive-structured-tibetan-foundation/)

[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/12/) / TFD: A Comprehensive Structured Tibetan Foundation Dataset for Low-Resource Language Processing and Large-Scale Modeling

# TFD: A Comprehensive Structured Tibetan Foundation Dataset for Low-Resource Language Processing and Large-Scale Modeling

**Authors:** [Cheng Huang](https://md.research.elamur.ai/free/en/authors/cheng-huang/), [Fan Gao](https://md.research.elamur.ai/free/en/authors/fan-gao/), [Nyima Tashi](https://md.research.elamur.ai/free/en/authors/nyima-tashi/), [Yutong Liu](https://md.research.elamur.ai/free/en/authors/yutong-liu/), Yadi Liu, Wenbin Wei, [Xiangxiang Wang](https://md.research.elamur.ai/free/en/authors/xiangxiang-wang/) and Yongbin Yu · **Year:** 2025 · **Venue:** arXiv:2503.18288 · **Licence:** [CC BY 4.0 (arXiv)](http://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2503.18288) · [PDF](https://arxiv.org/pdf/2503.18288) · [DOI 10.48550/arxiv.2503.18288](https://doi.org/10.48550/arxiv.2503.18288) · [arXiv 2503.18288](https://arxiv.org/abs/2503.18288) · [OpenAlex](https://openalex.org/W6929269293)

**Citations:** 0 citations (OpenAlex, 2026-10-14) · [4](https://md.research.elamur.ai/free/en/papers/huang-2025-tfd-comprehensive-structured-tibetan-foundation/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary

Tibetan large language models trail far behind those for major languages. Raw text is scarce, but the bigger problem is that no Tibetan resource covers the full pipeline of modern model building, which spans pre-training and instruction tuning, alignment for safety, optimization of preferences, and supervision of reasoning. Earlier Tibetan datasets are small, noisy or tied to one domain, and a concurrent effort (Pan et al., 2025) covers only pre-training.

The authors assemble TFD from two parts. TIBSTC pairs a multi-domain pre-training Corpus of about 11B tokens with four alignment sets. These were adapted from Alpaca, Safety-Prompts, CValues and HH-RLHF using machine drafts from Google Translate and Claude-3.5-Sonnet, then refined through two rounds of expert review. TIBSTC-CoT adds chain-of-thought reasoning data in Tibetan, produced with LLM assistance and checked by experts.

To test the data, they train Sun-Shine models on Llama and Qwen backbones with pre-training, SFT and DPO. These models beat strong baselines on Tibetan understanding, safety, reasoning and generation, sometimes surpassing much larger general-purpose models. The release gives Tibetan NLP a complete and reproducible training stack.

## Key points

- **TIBSTC composition:**
  - Corpus: 122,756,184 instances, 11B tokens, 18 GB.
  - Alpaca-Ti: 42,676 instances, 76M tokens.
  - Safety-Prompts-Ti: 38,156 instances.
  - CValues-Ti: 2,000 instances.
  - hh-rlhf-Ti: 47,411 instances (23,822 harmless, 23,589 helpful).
- **Translation quality control:** two Tibetan specialists and five annotators chose the better of the two machine drafts. After expert adjudication, the approval rate rose from 28.74% to 100%.
- **Safety-Prompts-Ti structure:**
  - 15,000 instruction-attack samples across six subtypes (2,500 each).
  - Over 23,000 samples spanning eight typical safety scenarios.
- **Translation results (chrF++, EN→BO):** Sunshine-thinking-8B scores 25.15 versus 13.47 for Qwen3-8B. Sunshine-thinking-1.7B scores 23.34 versus 3.64 for Qwen3-1.7B.
- **Availability and limitations:** code and data are at github.com/Vicentvankor/sun-shine. Dialectal varieties are underrepresented, and evaluation is text-only.

*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 (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 (4): [🧠 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 · [🛡️ Bias, fairness & safety](https://md.research.elamur.ai/free/en/topics/bias-fairness-safety/) · 517 works · [🌱 Low-resource languages](https://md.research.elamur.ai/free/en/topics/low-resource-languages/) · 267 works

## Cite

Cite as: Cheng Huang, Fan Gao, Nyima Tashi, Yutong Liu, Yadi Liu, Wenbin Wei, Xiangxiang Wang and Yongbin Yu. 2025. TFD: A Comprehensive Structured Tibetan Foundation Dataset for Low-Resource Language Processing and Large-Scale Modeling. arXiv:2503.18288.

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

```bibtex
@misc{huang2025tfd,
  title = {{TFD: A Comprehensive Structured Tibetan Foundation Dataset for Low-Resource Language Processing and Large-Scale Modeling}},
  author = {Cheng Huang, Fan Gao, Nyima Tashi, Yutong Liu, Yadi Liu, Wenbin Wei, Xiangxiang Wang and Yongbin Yu},
  year = {2025},
  doi = {10.48550/arxiv.2503.18288},
  eprint = {2503.18288},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2503.18288}
}
```

- [← previous Subjectivity in the Annotation of Bridging Anaphora](https://md.research.elamur.ai/free/en/papers/levine-2025-subjectivity-annotation-bridging-anaphora/)
- [next → The Landscape of Arabic Large Language Models (ALLMs): A New Era for Arabic Language Technology](https://md.research.elamur.ai/free/en/papers/al-khalifa-2025-landscape-arabic-large-language-models/)
- [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/page/12/)

## References

42 references, 16 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, and 1 others. 2023. [Gpt-4 technical report](https://md.research.elamur.ai/free/en/papers/openai-2023-gpt-4-technical-report/). arXiv preprint arXiv:2303.08774.
2. Zhuome Gongqu, Peng Luo, Dongzhou Jiayang, Jia Cairang, Jiacuo Cizhen, and Dongzhu Renqing. 2024. A tibetan ancient uchen text line dataset for ocr. In 2024 International Conference on Image Processing, Computer Vision and Machine Learning, pages 796– 799. [find](https://scholar.google.com/scholar?q=Zhuome+Gongqu%2C+Peng+Luo%2C+Dongzhou+Jiayang%2C+Jia+Cairang%2C+Jiacuo+Cizhen%2C+and+Dongzhu+Renqing%2E+2024%2E+A+tibetan+ancient+uchen+text+line+dataset+for+ocr%2E+In+2024+International+Conference+on+Image+Processin)
3. Anthropic. 2024. Claude 3.5 models. [https://www.anthropic.com/news/claude-3-5](https://www.anthropic.com/news/claude-3-5). Accessed: 2025-01-06. [↗](https://www.anthropic.com/news/claude-3-5)
4. Yuntao Bai, Andy Jones, Kamal Ndousse, Amanda Askell, Anna Chen, Nova DasSarma, Dawn Drain, Stanislav Fort, Deep Ganguli, Tom Henighan, and 1 others. 2022. Training a helpful and harmless assistant with reinforcement learning from human feedback. arXiv preprint arXiv:2204.05862. [find](https://scholar.google.com/scholar?q=Yuntao+Bai%2C+Andy+Jones%2C+Kamal+Ndousse%2C+Amanda+Askell%2C+Anna+Chen%2C+Nova+DasSarma%2C+Dawn+Drain%2C+Stanislav+Fort%2C+Deep+Ganguli%2C+Tom+Henighan%2C+and+1+others%2E+2022%2E+Training+a+helpful+and+harmless+assistant+wi)
5. Google. Google translate. [https://translate.google.com/](https://translate.google.com/). Accessed: 2025-01-06. [↗](https://translate.google.com/)
6. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, and 1 others. 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.
7. Yuxiao Bai, Zhen Yang, Zehan Chen, Shansan Cao, Chenfei Dong, Zhengyan Zhang, Bowen Yu, Xiaozhe Wang, Xia Shen, and 1 others. 2024. Qwen2.5: Yanrong Hao, Bo Chen, and Xiaobing Zhao. 2024. Tilts: Tibetan long text summarization dataset. In CCF International Conference on Natural Language Processing a… [find](https://scholar.google.com/scholar?q=Yuxiao+Bai%2C+Zhen+Yang%2C+Zehan+Chen%2C+Shansan+Cao%2C+Chenfei+Dong%2C+Zhengyan+Zhang%2C+Bowen+Yu%2C+Xiaozhe+Wang%2C+Xia+Shen%2C+and+1+others%2E+2024%2E+Qwen2%2E5%3A+Yanrong+Hao%2C+Bo+Chen%2C+and+Xiaobing+Zhao%2E+2024%2E+Tilts%3A+Tibet)
8. Shuheng Liu and Michael Best. 2025. A survey of nlp progress in sino-tibetan low-resource languages. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics, pages 7804–7825. [find](https://scholar.google.com/scholar?q=Shuheng+Liu+and+Michael+Best%2E+2025%2E+A+survey+of+nlp+progress+in+sino%2Dtibetan+low%2Dresource+languages%2E+In+Proceedings+of+the+2025+Conference+of+the+Nations+of+the+Americas+Chapter+of+the+Association+for)
9. Edward J. Hu, Phil Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, Weizhu Chen, and et al. 2021. [Lora: Low-rank adaptation of large language models](https://md.research.elamur.ai/free/en/papers/hu-2021-lora-low-rank-adaptation-large/). In Proceedings of the International Conference on Learning Representations.
10. Yutong Liu, Ziyue Zhang, Cheng Huang, Yongbin Yu, Xiangxiang Wang, Yuqing Cai, and Nyima Tashi. 2025a. Listening, imagining & refining: A heuristic optimized asr correction framework with llms. arXiv preprint arXiv:2509.15095. — [arXiv:2509.15095](https://arxiv.org/abs/2509.15095) [find](https://scholar.google.com/scholar?q=Yutong+Liu%2C+Ziyue+Zhang%2C+Cheng+Huang%2C+Yongbin+Yu%2C+Xiangxiang+Wang%2C+Yuqing+Cai%2C+and+Nyima+Tashi%2E+2025a%2E+Listening%2C+imagining+%26+refining%3A+A+heuristic+optimized+asr+correction+framework+with+llms%2E+arXiv+)
11. Cheng Huang, Nyima Tashi, Fan Gao, Yutong Liu, Jiahao Li, Hao Tian, Siyang Jiang, Thupten Tsering, Ban Ma-bao, Renzeg Duojie, and 1 others. 2025. Tibetan language and ai: A comprehensive survey of resources, methods and challenges. arXiv preprint arXiv:2510.19144. — [arXiv:2510.19144](https://arxiv.org/abs/2510.19144) [find](https://scholar.google.com/scholar?q=Cheng+Huang%2C+Nyima+Tashi%2C+Fan+Gao%2C+Yutong+Liu%2C+Jiahao+Li%2C+Hao+Tian%2C+Siyang+Jiang%2C+Thupten+Tsering%2C+Ban+Ma%2Dbao%2C+Renzeg+Duojie%2C+and+1+others%2E+2025%2E+Tibetan+language+and+ai%3A+A+comprehensive+survey+of+res)
12. Yutong Liu, Ziyue Zhang, Ban Ma-bao, Renzeng Duojie, Yuqing Cai, Yongbin Yu, Xiangxiang Wang, Fan Gao, Cheng Huang, and Nyima Tashi. 2025b. Tmd-tts: A unified tibetan multi-dialect text-to-speech synthesis for\\" u-tsang, amdo and kham speech dataset generation. arXiv preprint arXiv:2509.18060. — [arXiv:2509.18060](https://arxiv.org/abs/2509.18060) [find](https://scholar.google.com/scholar?q=Yutong+Liu%2C+Ziyue+Zhang%2C+Ban+Ma%2Dbao%2C+Renzeng+Duojie%2C+Yuqing+Cai%2C+Yongbin+Yu%2C+Xiangxiang+Wang%2C+Fan+Gao%2C+Cheng+Huang%2C+and+Nyima+Tashi%2E+2025b%2E+Tmd%2Dtts%3A+A+unified+tibetan+multi%2Ddialect+text%2Dto%2Dspeech+synt)
13. Aaron Hurst, Adam Lerer, Adam P Goucher, Adam Perelman, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Welihinda, Alan Hayes, Alec Radford, and 1 others. 2024. [Gpt-4o system card](https://md.research.elamur.ai/free/en/papers/openai-2024-gpt-4o-system-card/). arXiv preprint arXiv:2410.21276.
14. Ilya Loshchilov and Frank Hutter. 2019. Decoupled weight decay regularization. In International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Ilya+Loshchilov+and+Frank+Hutter%2E+2019%2E+Decoupled+weight+decay+regularization%2E+In+International+Conference+on+Learning+Representations%2E)
15. Hui Lv, Chi Pu, La Duo, Yan Li, Qingguo Zhou, and Jun Shen. 2025. T-llama: a tibetan large language model based on llama2. Complex & Intelligent Systems, 11(1):72. [find](https://scholar.google.com/scholar?q=Hui+Lv%2C+Chi+Pu%2C+La+Duo%2C+Yan+Li%2C+Qingguo+Zhou%2C+and+Jun+Shen%2E+2025%2E+T%2Dllama%3A+a+tibetan+large+language+model+based+on+llama2%2E+Complex+%26+Intelligent+Systems%2C+11%281%29%3A72%2E)
16. Han Jiang, Xiaoyuan Yi, Zhihua Wei, Ziang Xiao, Shu Wang, and Xing Xie. 2024. Raising the bar: Investigating the values of large language models via generative evolving testing. arXiv preprint arXiv:2406.14230. — [arXiv:2406.14230](https://arxiv.org/abs/2406.14230) [find](https://scholar.google.com/scholar?q=Han+Jiang%2C+Xiaoyuan+Yi%2C+Zhihua+Wei%2C+Ziang+Xiao%2C+Shu+Wang%2C+and+Xing+Xie%2E+2024%2E+Raising+the+bar%3A+Investigating+the+values+of+large+language+models+via+generative+evolving+testing%2E+arXiv+preprint+arXiv%3A2)
17. Yanchun Ma, Yongjian Liu, Qing Xie, Shengwu Xiong, Lihua Bai, and Anshu Hu. 2021. A tibetan thangka data set and relative tasks. Image and Vision Computing, 108:104125. [find](https://scholar.google.com/scholar?q=Yanchun+Ma%2C+Yongjian+Liu%2C+Qing+Xie%2C+Shengwu+Xiong%2C+Lihua+Bai%2C+and+Anshu+Hu%2E+2021%2E+A+tibetan+thangka+data+set+and+relative+tasks%2E+Image+and+Vision+Computing%2C+108%3A104125%2E)
18. Hyeonwoo Kim, Dahyun Kim, Jihoo Kim, Sukyung Lee, Yungi Kim, and Chanjun Park. 2025. Open ko-llm leaderboard2: Bridging foundational and practical evaluation for korean llms. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics… [find](https://scholar.google.com/scholar?q=Hyeonwoo+Kim%2C+Dahyun+Kim%2C+Jihoo+Kim%2C+Sukyung+Lee%2C+Yungi+Kim%2C+and+Chanjun+Park%2E+2025%2E+Open+ko%2Dllm+leaderboard2%3A+Bridging+foundational+and+practical+evaluation+for+korean+llms%2E+In+Proceedings+of+the+202)
19. Arijit Maji, Raghvendra Kumar, Akash Ghosh, Sriparna Saha, and 1 others. 2025. Sanskriti: A comprehensive benchmark for evaluating language models’ knowledge of indian culture. Findings of the Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Arijit+Maji%2C+Raghvendra+Kumar%2C+Akash+Ghosh%2C+Sriparna+Saha%2C+and+1+others%2E+2025%2E+Sanskriti%3A+A+comprehensive+benchmark+for+evaluating+language+models%E2%80%99+knowledge+of+indian+culture%2E+Findings+of+the+Associa)
20. Robert Kirk, Ishita Mediratta, Christoforos Nalmpantis, Jelena Luketina, Eric Hambro, Edward Grefenstette, and Roberta Raileanu. 2023. [Understanding the effects of rlhf on llm generalisation and diversity](https://md.research.elamur.ai/free/en/papers/kirk-2023-understanding-effects-rlhf-llm-generalisation/). arXiv preprint arXiv:2310.06452.
21. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, and 1 others. 2022. [Training language models to follow instructions with human feedback](https://md.research.elamur.ai/free/en/papers/ouyang-2022-training-language-models-follow-instructions/). Advances in neural information processing systems, 35:27730–27744.
22. Haonan Li, Yixuan Zhang, Fajri Koto, Yifei Yang, Hai Zhao, Yeyun Gong, Nan Duan, and Timothy Baldwin. 2024. [Cmmlu: Measuring massive multitask language understanding in chinese](https://md.research.elamur.ai/free/en/papers/li-2023-cmmlu-measuring-massive-multitask-language/). In Findings of the Association for Computational Linguistics: ACL 2024, pages 11260–11285.
23. Leiyu Pan, Bojian Xiong, Lei Yang, Renren Jin, Shaowei Zhang, Yue Chen, Ling Shi, Jiang Zhou, Junru Wu, Zhen Wang, and 1 others. 2025. Advancing large language models for tibetan with curated data and continual pre-training. arXiv preprint arXiv:2507.09205. — [arXiv:2507.09205](https://arxiv.org/abs/2507.09205) [find](https://scholar.google.com/scholar?q=Leiyu+Pan%2C+Bojian+Xiong%2C+Lei+Yang%2C+Renren+Jin%2C+Shaowei+Zhang%2C+Yue+Chen%2C+Ling+Shi%2C+Jiang+Zhou%2C+Junru+Wu%2C+Zhen+Wang%2C+and+1+others%2E+2025%2E+Advancing+large+language+models+for+tibetan+with+curated+data+and)
24. Yatao Liang, Hui Lv, Yan Li, La Duo, Chuanyi Liu, and Qingguo Zhou. 2024. Tibetan-bert-wwm: A tibetan pretrained model with whole word masking for text classification. IEEE Transactions on Computational Social Systems, 11(5):6268–6277. [find](https://scholar.google.com/scholar?q=Yatao+Liang%2C+Hui+Lv%2C+Yan+Li%2C+La+Duo%2C+Chuanyi+Liu%2C+and+Qingguo+Zhou%2E+2024%2E+Tibetan%2Dbert%2Dwwm%3A+A+tibetan+pretrained+model+with+whole+word+masking+for+text+classification%2E+IEEE+Transactions+on+Computation)
25. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. [Bleu: a method for automatic evaluation of machine translation](https://md.research.elamur.ai/free/en/papers/papineni-2002-bleu-method-automatic-evaluation-machine/). In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 311–318.
26. Chin-Yew Lin. 2004. [Rouge: A package for automatic evaluation of summaries](https://md.research.elamur.ai/free/en/papers/lin-2004-rouge-package-automatic-evaluation-summaries/). In Text Summarization Branches Out, pages 74–81.
27. Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, and 1 others. 2024. Deepseek-v3 technical report. arXiv preprint arXiv:2412.19437. [find](https://scholar.google.com/scholar?q=Aixin+Liu%2C+Bei+Feng%2C+Bing+Xue%2C+Bingxuan+Wang%2C+Bochao+Wu%2C+Chengda+Lu%2C+Chenggang+Zhao%2C+Chengqi+Deng%2C+Chenyu+Zhang%2C+Chong+Ruan%2C+and+1+others%2E+2024%2E+Deepseek%2Dv3+technical+report%2E+arXiv+preprint+arXiv%3A2412)
28. Maja Popović. 2017. [chrf++: Words helping character n-grams](https://md.research.elamur.ai/free/en/papers/popovic-2017-chrf-words-helping-character-n/). In Proceedings of the Second Conference on Machine Translation, pages 612–618. Association for Computational Linguistics.
29. Rafael Rafailov, Aditya Sharma, Eric Mitchell, Christopher D. Manning, Stefano Ermon, and Chelsea Finn. 2024. [Direct preference optimization: Your language model is secretly a reward model](https://md.research.elamur.ai/free/en/papers/rafailov-2023-direct-preference-optimization-language-model/). Advances in Neural Information Processing Systems, 36.
30. Yue Zhao, Xiaona Xu, Jianjian Yue, Wei Song, Xiali Li, Licheng Wu, and Qiang Ji. 2020. An open speech resource for tibetan multi-dialect and multitask recognition. International Journal of Computational Science and Engineering, 22(2-3):297–304. [find](https://scholar.google.com/scholar?q=Yue+Zhao%2C+Xiaona+Xu%2C+Jianjian+Yue%2C+Wei+Song%2C+Xiali+Li%2C+Licheng+Wu%2C+and+Qiang+Ji%2E+2020%2E+An+open+speech+resource+for+tibetan+multi%2Ddialect+and+multitask+recognition%2E+International+Journal+of+Computation)
31. Hao Sun, Zhexin Zhang, Jiawen Deng, Jiale Cheng, and Minlie Huang. 2023. Safety assessment of chinese large language models. arXiv preprint arXiv:2304.10436. [find](https://scholar.google.com/scholar?q=Hao+Sun%2C+Zhexin+Zhang%2C+Jiawen+Deng%2C+Jiale+Cheng%2C+and+Minlie+Huang%2E+2023%2E+Safety+assessment+of+chinese+large+language+models%2E+arXiv+preprint+arXiv%3A2304%2E10436%2E)
32. Chen Zheng, Fan Yin, Hao Zhou, Fandong Meng, Jie Zhou, Kai-Wei Chang, Maosong Huang, and Nanyun Peng. 2024. On prompt-driven safeguarding for large language models. In Proceedings of the International Conference on Machine Learning. [find](https://scholar.google.com/scholar?q=Chen+Zheng%2C+Fan+Yin%2C+Hao+Zhou%2C+Fandong+Meng%2C+Jie+Zhou%2C+Kai%2DWei+Chang%2C+Maosong+Huang%2C+and+Nanyun+Peng%2E+2024%2E+On+prompt%2Ddriven+safeguarding+for+large+language+models%2E+In+Proceedings+of+the+International)
33. Rohan Taori, Ishaan Gulrajani, Tianyi Zhang, Yann Dubois, Xuechen Li, Carlos Guestrin, Percy Liang, and Tatsunori B. Hashimoto. 2023. Stanford alpaca: An instruction-following llama model. [https://github.com/tatsu-lab/stanford_alpaca](https://github.com/tatsu-lab/stanford_alpaca). [↗](https://github.com/tatsu-lab/stanford_alpaca)
34. Wenhao Zhuang and Yuan Sun. 2025. [Cute: A multilingual dataset for enhancing cross-lingual knowledge transfer in low-resource languages](https://md.research.elamur.ai/free/en/papers/zhuang-2025-cute-multilingual-dataset-enhancing-cross/). In Proceedings of the 31st International Conference on Computational Linguistics, pages 10037–10046.
35. Gemini Team, Petko Georgiev, Ving Ian Lei, Ryan Burnell, Libin Bai, Anmol Gulati, Garrett Tanzer, Damien Vincent, Zhufeng Pan, Shibo Wang, and 1 others. 2024. [Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context](https://md.research.elamur.ai/free/en/papers/gemini-team-2024-gemini-1-5-unlocking-multimodal/). arXiv preprint arXiv:2403.05530.
36. Andreas Wagner and Bettina Zeisler. 2004. A syntactically annotated corpus of tibetan. In Proceedings of the Fourth International Conference on Language Resources and Evaluation. [find](https://scholar.google.com/scholar?q=Andreas+Wagner+and+Bettina+Zeisler%2E+2004%2E+A+syntactically+annotated+corpus+of+tibetan%2E+In+Proceedings+of+the+Fourth+International+Conference+on+Language+Resources+and+Evaluation%2E)
37. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc Le, Ed Chi, and Denny Zhou. 2023. [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.
38. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, and 1 others. 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.
39. Guixian Xu, Zeli Su, Ziyin Zhang, Jianing Liu, XU Han, Ting Zhang, and Yushuang Dong. 2025. Cmhg: A dataset and benchmark for headline generation of minority languages in china. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 12350–12357. [find](https://scholar.google.com/scholar?q=Guixian+Xu%2C+Zeli+Su%2C+Ziyin+Zhang%2C+Jianing+Liu%2C+XU+Han%2C+Ting+Zhang%2C+and+Yushuang+Dong%2E+2025%2E+Cmhg%3A+A+dataset+and+benchmark+for+headline+generation+of+minority+languages+in+china%2E+In+Proceedings+of+the+)
40. An Yang and 1 others. 2025. [Qwen3 technical report](https://md.research.elamur.ai/free/en/papers/yang-2025-qwen3-technical-report/). Preprint, arXiv:2505.09388.
41. Yixin Yu, Shuohang Zuo, Haoming Jiang, Wenxuan Ren, Tongtao Zhao, and Chenguang Zhang. 2021. Fine-tuning pre-trained language model with weak supervision: A contrastive-regularized self-training approach. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Comp… [find](https://scholar.google.com/scholar?q=Yixin+Yu%2C+Shuohang+Zuo%2C+Haoming+Jiang%2C+Wenxuan+Ren%2C+Tongtao+Zhao%2C+and+Chenguang+Zhang%2E+2021%2E+Fine%2Dtuning+pre%2Dtrained+language+model+with+weak+supervision%3A+A+contrastive%2Dregularized+self%2Dtraining+appro)
42. Chen Zhang, Mingxu Tao, Zhiyuan Liao, and Yansong Feng. 2025. Milic-eval: Benchmarking multilingual llms for china’s minority languages. Findings of the Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Chen+Zhang%2C+Mingxu+Tao%2C+Zhiyuan+Liao%2C+and+Yansong+Feng%2E+2025%2E+Milic%2Deval%3A+Benchmarking+multilingual+llms+for+china%E2%80%99s+minority+languages%2E+Findings+of+the+Association+for+Computational+Linguistics%2E)

*Updated 31 October 2026*