PeopleofSuperintelligence

Researcher

Kai Dong

Kai Dong is a DeepSeek-AI research contributor whose coauthored work spans open language models, code generation, vision-language systems and reasoning. In 2024, he coauthored DeepSeek LLM, DeepSeek-Coder, DeepSeek-V2, DeepSeek-V3 and DeepSeek-VL. The 2025 DeepSeek-R1 paper explored reinforcement-learning reasoning, and DeepSeek-V3.2 added sparse attention and a pipeline for agent training. As of October 2026, he contributes to DeepSeek-AI research as a coauthor of the V4.1-Flash report on multimodal, long-context models.

Now at
DeepSeek
Works
40

Publication numbers from OpenAlex, as of 22 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Timeline

  1. 2024

    DeepSeek LLM dataset

    In 2024, he coauthors DeepSeek LLM, a project that studies scaling laws for open-source language models and uses a pretraining dataset of two trillion tokens.

  2. 2024

    DeepSeek-Coder series

    In 2024, he coauthors DeepSeek-Coder, an open-source code-model series with model sizes from 1.3B to 33B, trained from scratch on two trillion tokens.

  3. March 2024

    DeepSeek-VL release

    In March 2024, he coauthors DeepSeek-VL, an open-source vision-language model for real-world understanding; the team releases base and chat models in 1.3B and 7B sizes.

  4. 2024

    DeepSeek-V2 language model

    In 2024, he coauthors DeepSeek-V2, a mixture-of-experts language model with 236B total parameters and 21B active per token. It supports a 128K-token context and uses sparse computation to make training more economical.

  5. 2024

    DeepSeek-V3 architecture

    In 2024, he coauthors the DeepSeek-V3 report, which describes a 671B-parameter mixture-of-experts model with 37B active per token. The model uses multi-head latent attention, DeepSeekMoE, auxiliary-loss-free load balancing and multi-token prediction.

  6. 2025

    DeepSeek-R1 reasoning

    In 2025, he coauthors the Nature paper on DeepSeek-R1, which shows that pure reinforcement learning can incentivize language-model reasoning without human-labelled reasoning trajectories. The framework develops self-reflection, verification and dynamic strategy adaptation.

  7. 2025

    DeepSeek-V3.2 and agents

    In 2025, he coauthors DeepSeek-V3.2, which combines DeepSeek Sparse Attention with a scalable reinforcement-learning framework and a synthesis pipeline that generates training data for tool-use tasks.

  8. As of October 2026

    DeepSeek-V4.1-Flash report

    As of October 2026, he contributes to DeepSeek-AI research as a coauthor of its V4.1-Flash report. The report describes a multimodal 552B-parameter mixture-of-experts model with a one-million-token context and compressed key-value caching for long-context agent workloads.

  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 Kai Dong yet. Send us a free-licensed photo.

  1. DeepSeek LLM: Scaling Open-Source Language Models with Longtermism arxiv.org. Checked 21 October 2026.
  2. DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence arxiv.org. Checked 21 October 2026.
  3. GitHub - deepseek-ai/DeepSeek-VL: DeepSeek-VL: Towards Real-World Vision-Language Understanding github.com. Checked 21 October 2026.
  4. DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model arxiv.org. Checked 21 October 2026.
  5. DeepSeek-V3 Technical Report arxiv.org. Checked 21 October 2026.
  6. DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning | Nature nature.com. Checked 21 October 2026.
  7. DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models arxiv.org. Checked 21 October 2026.
  8. DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression arxiv.org. Checked 21 October 2026.