PeopleofSuperintelligence

Builder

Xiaodong Liu

Also known as 刘晓东

Xiaodong Liu is an artificial intelligence researcher employed by DeepSeek. In January 2024 he was an author of the DeepSeek LLM report, which introduced DeepSeek LLM in 7B and 67B configurations and a dataset of 2 trillion tokens. DeepSeek-V2, in May 2024, named him as an author and described a mixture-of-experts model with 236B total parameters, of which 21B are activated for each token. The December 2024 DeepSeek-V3 technical report also listed him and presented a mixture-of-experts model with 671B total parameters, with 37B activated for each token. Nature published DeepSeek-R1 in September 2025 with him among the authors.

Now at
DeepSeek

Timeline

  1. January 2024

    DeepSeek LLM author

    In January 2024 he is an author of the DeepSeek LLM report from DeepSeek-AI. The report introduces DeepSeek LLM in 7B and 67B configurations and a dataset of 2 trillion tokens.

  2. May 2024

    DeepSeek-V2 engineering credits

    In May 2024, DeepSeek-V2 names him as an author of a mixture-of-experts language model with 236B total parameters, of which 21B are activated for each token. The report lists his name in the research and engineering section.

  3. August 2024

    Fire-Flyer AI-HPC

    In August 2024 he is an author of the Fire-Flyer AI-HPC report. It introduces the Fire-Flyer AI-HPC architecture, a hardware-software co-design for deep learning.

  4. December 2024

    The DeepSeek-V3 report

    In December 2024 he is an author of the DeepSeek-V3 technical report. That report presents a mixture-of-experts language model with 671B total parameters, with 37B activated for each token.

  5. September 2025

    DeepSeek-R1 in Nature

    Nature publishes DeepSeek-R1 in September 2025 and lists Xiaodong Liu among the authors. The paper shows that reinforcement learning can incentivize reasoning in large language models.

  6. As of October 2026

    Researcher at DeepSeek

    He is an artificial intelligence researcher employed by DeepSeek, with an affiliation at DeepSeek-AI in Hangzhou. In September 2026 the DeepSeek Elastic Compute report lists him as an author affiliated with DeepSeek-AI.

Sources and credits

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

  1. DeepSeek LLM Scaling Open-Source Language Models with Longtermism arxiv.org. Checked 16 October 2026.
  2. [2405.04434] DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model arxiv.org. Checked 16 October 2026.
  3. [2405.04434] DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model ar5iv.labs.arxiv.org. Checked 16 October 2026.
  4. [2408.14158] Fire-Flyer AI-HPC: A Cost-Effective Software-Hardware Co-Design for Deep Learning arxiv.org. Checked 16 October 2026.
  5. [2412.19437] DeepSeek-V3 Technical Report arxiv.org. Checked 16 October 2026.
  6. DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning | Nature nature.com. Checked 16 October 2026.
  7. Xiaodong Liu - Wikidata wikidata.org. Checked 16 October 2026.
  8. Online bibliography of Xiaodong Liu csauthors.net. Checked 16 October 2026.
  9. DeepSeek Elastic Compute (DSec): A Sandbox Infrastructure for Effective Agentic Training at Scale arxiv.org. Checked 16 October 2026.