PeopleofSuperintelligence

Researcher

Zhewen Hao

Also known as Hao Zhewen

Zhewen Hao works at DeepSeek-AI on data and pre-training for large language models. Educated in the Turing Class of 2019 at Peking University's School of EECS, he captained the PKU Supercomputing Team that won a silver medal at the SC20 Student Cluster Competition in 2020. He was an author of DeepSeek LLM, DeepSeek-V2 and DeepSeek-Coder-V2 in 2024, then of DeepSeek-V3 in December 2024 and DeepSeek-R1 in January 2025, which Nature published in September 2025. He was also an author of DeepSeek-V3.2 in December 2025 and the DeepSeek-V4 preview in 2026, whose two models support a context length of one million tokens.

Now at
DeepSeek

Timeline

  1. By 2020

    Turing Class at Peking University

    He studies for a bachelor's degree in the School of EECS at Peking University, as part of the Turing Class of 2019. He joins the PKU Supercomputing Team in his sophomore fall.

  2. 2020

    Silver medal at SC20

    He captains the PKU Supercomputing Team at SC20 in 2020. The six-member team, advised by EECS associate professor Liang Yun, wins the silver medal in the finals, a virtual contest among 19 teams.

  3. January 2024

    The DeepSeek LLM paper

    He is among the DeepSeek-AI authors of DeepSeek LLM, posted in January 2024. The paper presents open-source 7B and 67B language models, guided by scaling laws and pre-trained on a dataset of 2 trillion tokens.

  4. May 2024–June 2024

    DeepSeek-V2 and Coder-V2

    In May 2024, he is among the authors of DeepSeek-V2, a Mixture-of-Experts model with 236B total parameters and 21B activated for each token. In June 2024, he is also an author of DeepSeek-Coder-V2, an open-source code model further pre-trained from a DeepSeek-V2 checkpoint on 6 trillion additional tokens.

  5. December 2024–January 2025

    DeepSeek-V3 and DeepSeek-R1

    He is an author of the DeepSeek-V3 Technical Report, posted in December 2024, which describes a Mixture-of-Experts model with 671B total parameters pre-trained on 14.8 trillion tokens. In January 2025, he is also an author of DeepSeek-R1, which shows that the reasoning abilities of language models can be incentivized through reinforcement learning.

  6. September 2025

    DeepSeek-R1 appears in Nature

    Nature publishes the DeepSeek-R1 paper in September 2025 under the title "DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning", with him among its authors.

  7. December 2025

    The DeepSeek-V3.2 report

    He is an author of DeepSeek-V3.2, posted in December 2025. The report introduces DeepSeek Sparse Attention, an attention mechanism that reduces computational complexity in long-context scenarios.

  8. As of October 2026

    Data and pre-training at DeepSeek

    He works at DeepSeek-AI, focused on data and pre-training for large language models. In 2026, he is an author of the DeepSeek-V4 preview report, whose two Mixture-of-Experts models support a context length of one million tokens.

Career

  1. NowDeepSeek

Education

  1. Peking UniversityBachelor's degree

From the biography.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes DeepSeek. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  2. Companies supply the most prior-work hires at 7 of the 9 frontier AI labsIncludes DeepSeek. Our estimates of the organizations where 550 people worked directly before joining 9 frontier AI labs across 620 career transitions, comparing the main feeder organizations per lab.
  3. 5,000 frontier users would be 38.2% of the core labs' 13,100 reported staffIncludes DeepSeek. 5,000 frontier users would be 38.2% of the core labs' 13,100 reported staff. We test the estimate against 3,913 named model contributors and dated all-staff records across Google DeepMind, OpenAI and Anthropic, plus six additional frontier-grade labs.
  4. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes DeepSeek. We estimate that alumni of 17 major AI labs went on to found or lead 350 other organizations, including 226 they founded, and we trace 472 career moves directly between 23 labs since 2010.

Sources and credits

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

  1. Hao Zhewen | PKU Supercomputing Team | Peking University Student Cluster Competition Team hpc.pku.edu.cn. Checked 20 October 2026.
  2. PKU student supercomputing team wins silver medal in SC20 Student Cluster Competition english.pku.edu.cn. Checked 20 October 2026.
  3. DeepSeek LLM: Scaling Open-Source Language Models with Longtermism arxiv.org. Checked 20 October 2026.
  4. DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model arxiv.org. Checked 20 October 2026.
  5. DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence arxiv.org. Checked 20 October 2026.
  6. DeepSeek-V3 Technical Report arxiv.org. Checked 20 October 2026.
  7. DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning arxiv.org. Checked 20 October 2026.
  8. DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning | Nature nature.com. Checked 20 October 2026.
  9. DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models arxiv.org. Checked 20 October 2026.
  10. DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence arxiv.org. Checked 20 October 2026.