Junying Chen   陈俊颖

Hello!

I am a Ph.D. candidate in Computer Science at The Chinese University of Hong Kong, Shenzhen, advised by Prof. Benyou Wang. Prior to CUHKSZ, I earned my master's from Harbin Institute of Technology (Shenzhen), supervised by Prof. Qingcai Chen, and my bachelor's from Jinan University.

Find me on Google Scholar (2.5k+ citations), and GitHub (6K stars)!

Email: junying.chen.cs@gmail.com

✦ I expect to graduate in 2027 and am exploring opportunities in both industry and academia.

Education

  • 2023 - Present, Ph.D., The Chinese University of Hong Kong, Shenzhen. (GPA Rank: 1st / 24)
  • 2020 - 2023, Master, Harbin Institute of Technology, Shenzhen. (GPA Rank: 1st / 111)
  • 2016 - 2020, Bachelor, Jinan University. (GPA Rank: 1st / 51)

    • Research

  • Google Scholar Citations: 2.5K (1.5K first-author citations)
  • GitHub Stars: 6K
  • Open-Source Model Downloads: 487K
  • Open-Source Dataset Downloads: 274K
  • Research Interests: Medical LLMs, Multimodal LLMs, and Medical AI Agents

    • Publications

    Selected (co-)first-author papers are listed. Full list: Google Scholar.

    Junying Chen*, Xinyuan Xie*, et al.
    OnePO: Direct One-stage Policy Optimization for SFT-free Domain Adaptation
    ICML 2026 (HuatuoGPT-3)

    Junying Chen, Zhenyang Cai, et al.
    Incentivizing Medical Vision Capabilities from Large-Scale Multimodal Pre-training
    Under Review 2026 (HuatuoGPT-Vision2)

    Junying Chen, Zhenyang Cai, et al.
    HuatuoGPT-o1, Towards Medical Complex Reasoning with LLMs
    ACL Findings 2025 [pdf] [code]

    Junying Chen, Chi Gui, et al.
    HuatuoGPT-Vision, Towards Injecting Medical Visual Knowledge into Multimodal LLMs at Scale
    EMNLP 2024 [pdf] [code]

    Junying Chen, Xidong Wang, et al.
    HuatuoGPT-II, One-stage Training for Medical Adaption of LLMs
    COLM 2024 [pdf] [code]

    Hongbo Zhang*, Junying Chen*, Feng Jiang*, et al.
    HuatuoGPT, Towards Taming Language Model to Be a Doctor
    EMNLP Findings 2023 [pdf] [code]

    Junying Chen*, Chi Gui*, et al.
    CoD, Towards an Interpretable Medical Agent using Chain of Diagnosis
    ACL Findings 2025 [pdf] [code]

    Zhenyang Cai*, Junying Chen*, Rongsheng Wang*, et al.
    On the Compositional Generalization of Multimodal LLMs for Medical Imaging
    ACL 2025 [pdf] [code]

    Wanlong Liu*, Junying Chen*, Yunjin Yang*, et al.
    Benchmarking LLMs on Authentic Cases from Medical Journals
    ACL Findings 2026

    Ke Ji*, Junying Chen*, et al.
    LLMs Could Autonomously Learn Without External Supervision
    ACL Findings 2025 [pdf] [code]

    Wanlong Liu*, Junying Chen*, et al.
    RAG-Instruct: Boosting LLMs with Diverse Retrieval-Augmented Instructions
    EMNLP 2025 [pdf] [code]

    Junying Chen*, Dongfang Li*, et al.
    Diaformer: Automatic Diagnosis via Symptoms Sequence Generation
    AAAI 2022 [pdf] [code]

    Junying Chen*, Zhenyang Cai*, et al.
    ShizhenGPT: Towards Multimodal LLMs for Traditional Chinese Medicine
    Preprint 2025 [pdf] [code]

    Junying Chen*, Zhenyang Cai*, Pengcheng Chen*, Shunian Chen*, et al.
    ShareGPT-4o-Image: Aligning Multimodal Models with GPT-4o-Level Image Generation
    Preprint 2025 [pdf] [code]

    Rongsheng Wang*, Junying Chen*, et al.
    MedGen: Unlocking Medical Video Generation by Scaling Granularly-annotated Medical Videos
    Preprint 2025 [pdf] [code]

    Xidong Wang*, Nuo Chen*, Junying Chen*, et al.
    Apollo: A Lightweight Multilingual Medical LLM towards Democratizing Medical AI to 6B People
    Preprint 2024 [pdf] [code]

    Junying Chen, Qingcai Chen, et al.
    SeDR: Segment Representation Learning for Long Documents Dense Retrieval
    Preprint 2022 [pdf] [code]

    Research Philosophy

    I am motivated by the potential of LLMs to support broad applications and create real-world value. My research focuses on making LLMs more usable in real-world practice:

  • Domain adaptation: specializing LLMs for high-stakes domains (e.g., healthcare).
    (HuatuoGPT, HuatuoGPT-II, HuatuoGPT-3, HuatuoGPT-o1, AL, Apollo)
  • Multimodal adaptation: enabling LLMs to perceive, reason over, and align with the visual world.
    (HuatuoGPT-Vision, HuatuoGPT-Vision2, ShizhenGPT, ShareGPT-4o-Image, MedGen, CG)
  • Trustworthy agents: building interpretable, controllable, and reliable AI systems.
    (HuatuoGPT-R [Coming], CoD, RAG-Instruct, Diaformer)

    • Toward useful and trustworthy AI on the path to AGI.

    Awards

  • National Scholarship for Undergraduate (Top 0.2%)
  • National Scholarship for Graduate (Top 0.2%)
  • Merit Fellowship for CUHKSZ PhD

    • This website is adapted from Gregory Gunderson and Tianyu Gao.