PeopleofSuperintelligence

Researcher

Baosong Yang

Baosong Yang is an AI researcher whose work connects machine translation with multilingual large language models. He earned degrees at Sichuan University, Waseda University and the University of Macau, and began publishing on attention during a 2018 internship at Tencent AI Lab. His work moved from attention methods for neural machine translation to multilingual models such as PolyLM, Qwen3 and Qwen3-Omni. As of October 2026, he supervises a research scientist on Alibaba's Qwen Multilingual team at Tongyi Lab.

Now at
Alibaba
Works
122
Citations
2,163
h-index
25
Badges

Publication numbers from OpenAlex, as of 18 October 2026.

Timeline

  1. 2014–2015

    Degrees at Sichuan and Waseda

    He earns a bachelor's degree at Sichuan University in 2014 and a master's degree at Waseda University in 2015 under Professor Yves Lepage.

  2. 2018

    Modeling localness for self-attention networks

    In 2018, he interns at Tencent AI Lab under Zhaopeng Tu and coauthors "Modeling Localness for Self-Attention Networks," an EMNLP paper that uses a learnable Gaussian bias to model local context.

  3. 2019

    Context-aware self-attention networks

    He completes a Ph.D. in the NLP2CT Lab at the University of Macau in 2019 under Derek F. Wong and Lidia S. Chao. His 2019 AAAI paper explores adding context to self-attention for machine translation.

  4. 2023–2024

    PolyLM and MoNMT

    He coauthors PolyLM in 2023, a multilingual LLM trained on 640 billion tokens. His 2024 MoNMT paper proposes a modular strategy to combine monolingual and bilingual knowledge in neural machine translation.

  5. 2025

    Qwen3 and Qwen3-Omni

    In 2025, he coauthors the Qwen3 technical report and is a core contributor to Qwen3-Omni, a multimodal model for text, image, audio and video.

  6. 2025

    Leading multilingual Qwen research

    In 2025, he is a scientist at Alibaba's Tongyi Lab and leads efforts to improve multilingual capabilities across the Qwen series.

  7. As of October 2026

    Supervising Qwen multilingual research

    As of October 2026, he supervises a research scientist on Alibaba's Qwen Multilingual team at Tongyi Lab.

Career

  1. Since 2025AlibabaScientist
  2. 2018TencentIntern

Education

  1. 2019University of MacauPhD
  2. 2015Waseda UniversityMaster's degree
  3. 2014Sichuan UniversityBachelor's degree

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. Modeling Localness for Self-Attention Networks2018 · Conference on Empirical Methods in Natural Language Processing (EMNLP) · 198 citations
  2. Multi-Head Attention with Disagreement Regularization2018 · Conference on Empirical Methods in Natural Language Processing (EMNLP) · 194 citations
  3. Qwen3 Technical Report2025 · 147 citationsWith Beichen Zhang, Fei Huang, Le Yu, Rui Men and Zihan Qiu
  4. Convolutional Self-Attention Networks2019 · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · 128 citations
  5. Qwen2.5 Technical Report2024 · 122 citationsWith Beichen Zhang, Fei Huang, Le Yu, Rui Men and Zihan Qiu

From OpenAlex, as of 18 October 2026.

  1. Google alumni founded or led 82 other organizations, the most among the labs we compareIncludes Baosong Yang. We trace 189 organizations founded or led by AI lab alumni, and 196 career moves between the labs.
  2. The company share of AI researchers’ top papers rose from 3% to 32%Includes Alibaba. We estimate where leading AI researchers worked when their most-cited papers appeared. The company share rose from 3% in 1975–1979 to 32% in 2020–2024.
  3. Google has the highest combined h-index among AI organizationsIncludes Alibaba. Our estimates of combined and median h-index at AI organizations, using current affiliations and OpenAlex records.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes Sichuan University. Our estimates of where the people shaping AI studied and where their careers led, from education records in cited biographies.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Baosong Yang yet. Send us a free-licensed photo.

  1. Baosong Yang | Alibaba baosongyang.site. Checked 18 October 2026.
  2. Modeling Localness for Self-Attention Networks - ACL Anthology aclanthology.org. Checked 18 October 2026.
  3. Context-Aware Self-Attention Networks | Proceedings of the AAAI Conference on Artificial Intelligence ojs.aaai.org. Checked 18 October 2026.
  4. Baosong Yang – GOSIM AI Paris 2025 paris2025.gosim.org. Checked 18 October 2026.
  5. PolyLM: An Open Source Polyglot Large Language Model arxiv.org. Checked 18 October 2026.
  6. MoNMT: Modularly Leveraging Monolingual and Bilingual Knowledge for Neural Machine Translation - ACL Anthology aclanthology.org. Checked 18 October 2026.
  7. Qwen3 Technical Report arxiv.org. Checked 18 October 2026.
  8. Qwen3-Omni Technical Report arxiv.org. Checked 18 October 2026.
  9. Minghao Wu (吴明昊) - Alibaba Qwen - Research Scientist minghao-wu.github.io. Checked 18 October 2026.