PeopleofSuperintelligence

Researcher

Bingxuan Wang

Also known as 王炳宣

Bingxuan Wang is a researcher whose work spans computer vision, language models, code and multimodal systems. As an undergraduate at Peking University's Yuanpei College, he joined PKU Team Zero for a 2019 text-recognition challenge. He coauthored a 2020 study of feature aggregation and, in 2024 and 2025, papers on language, code, vision-language and reasoning models. He joined Tencent's Hunyuan team by June 2026.

Now at
DeepSeek
Works
18
Citations
2,290
h-index
7
Badges

Publication numbers from OpenAlex, as of 19 October 2026.

Timeline

  1. 2019

    PKU Team Zero at ICDAR

    In 2019, he is a junior undergraduate at Peking University's Yuanpei College on PKU Team Zero; the team's algorithm ranks first in the ICDAR 2019 Robust Reading Challenge for arbitrary-shaped text.

  2. 2020

    Feature aggregation for distillation

    In 2020, he coauthors “Differentiable Feature Aggregation Search for Knowledge Distillation,” which introduces a two-stage search for feature aggregations that can transfer supervision from multiple teacher feature maps.

  3. January 2024

    DeepSeek LLM

    In January 2024, he coauthors DeepSeek LLM, a study of scaling open-source language models in 7B and 67B configurations.

  4. March 2024

    DeepSeek-VL vision models

    In March 2024, he coauthors DeepSeek-VL, a vision-language chatbot family with 1.3B and 7B models for real-world applications.

  5. May–June 2024

    DeepSeek-V2 and Coder-V2

    In May 2024, he coauthors DeepSeek-V2, an MoE language model; in June 2024, he coauthors DeepSeek-Coder-V2, an open-source code model for code-specific tasks.

  6. December 2024

    Two DeepSeek models

    In December 2024, he coauthors DeepSeek-VL2 for multimodal understanding and DeepSeek-V3, a 671-billion-parameter MoE language model.

  7. January 2025

    DeepSeek-R1 reasoning

    In January 2025, he coauthors DeepSeek-R1, a report that describes how pure reinforcement learning can incentivize reasoning abilities in large language models.

  8. By June 2026

    Tencent Hunyuan

    By June 2026, he joins Tencent's Hunyuan team after contributing as a core author to DeepSeek's first-generation large language model.

Career

  1. 2024–2026DeepSeek

Education

  1. Peking UniversityUndergraduate

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

Most-cited works

  1. DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning2025 · Nature · 1,112 citationsWith Bochao Wu, Chenggang Zhao, Daya Guo and Zhibin Gou
  2. Can Open Large Language Models Catch Vulnerabilities?2025 · 538 citationsWith Bochao Wu, Chenggang Zhao and Zhibin Gou
  3. DeepSeek-V3 Technical Report2024 · 292 citationsWith Bochao Wu, Chenggang Zhao and Daya Guo
  4. DeepSeek LLM: Scaling Open-Source Language Models with Longtermism2024 · 98 citationsWith Chenggang Zhao and Daya Guo
  5. DeepSeek-VL: Towards Real-World Vision-Language Understanding2024 · 60 citationsWith Zhenda Xie

From OpenAlex, as of 19 October 2026.

  1. The median h-index is 54 at universities and 19 at frontier AI labsIncludes Bingxuan Wang. The median h-index is 54 at universities and 19 at frontier AI labs. We compare publishing careers, citations and moves to estimate how universities and frontier labs contribute to AI research.
  2. The company share of AI researchers’ top papers rose from 2% to 32%Includes DeepSeek. We estimate where leading AI researchers worked when their most-cited papers appeared. The company share rose from 2% in 1975–1979 to 32% in 2020–2024.
  3. Google alumni founded or led 81 other organizations, the most among the labs we compareIncludes DeepSeek. We trace 194 organizations founded or led by AI lab alumni, and 202 career moves between the labs.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes Peking 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 Bingxuan Wang yet. Send us a free-licensed photo.

  1. GitHub - longshangbang/ICDAR2019-ArT-Recognition-Alchemy: PKU Team Zero's code for participation in ICDAR2019 ArT Recognition track (Champion) github.com. Checked 18 October 2026.
  2. Differentiable Feature Aggregation Search for Knowledge Distillation arxiv.org. Checked 18 October 2026.
  3. DeepSeek LLM: Scaling Open-Source Language Models with Longtermism arxiv.org. Checked 18 October 2026.
  4. DeepSeek-VL: Towards Real-World Vision-Language Understanding arxiv.org. Checked 18 October 2026.
  5. DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model arxiv.org. Checked 18 October 2026.
  6. DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence arxiv.org. Checked 18 October 2026.
  7. DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding arxiv.org. Checked 18 October 2026.
  8. DeepSeek-V3 Technical Report arxiv.org. Checked 18 October 2026.
  9. DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning arxiv.org. Checked 18 October 2026.
  10. DeepSeek Launches Large-Scale Recruitment, Aims to At Least Double the Size of All Departments eu.36kr.com. Checked 18 October 2026.