Skip to content
View GeminiLight's full-sized avatar
115 followers · 22 following

Achievements

Achievement: Pull Sharkx3Achievement: YOLOAchievement: Starstruckx3Achievement: Arctic Code Vault Contributor

Achievements

Achievement: Pull Sharkx3Achievement: YOLOAchievement: Starstruckx3Achievement: Arctic Code Vault Contributor

Highlights

  • Pro

Block or report GeminiLight

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
geminilight/README.md

Hi 👋 I'm Gemini Light / Tianfu Wang

Ph.D. student @ HKUST(GZ) · Product Builder · Human-centered Agentic AI

Home Page   Scholar   Zhihu   Email

🎓 Short Bio

My name is Tianfu Wang. I am a Ph.D. student at HKUST(GZ), supervised by Prof. Hui Xiong and Dr. Nicholas Jing Yuan. I received my M.S. from USTC and my B.E. from CQU. I have also worked as a research intern at Z.ai, Kimi, Tencent CodeBuddy / WorkBuddy, Microsoft AI / MSRA, and JD Explore Academy.

🔬 Research Interests

I seek to answer: How can we build socially aware AI agents that empower humans to thrive?

Recently, I've been working on:

  • Agent Foundations: shared memory and context, human–agent collaboration, and reusable agent workflows.
    Selected work: MindOS · Agentic Consensus · Kimi K3 (Technical Report · Contributor)

  • Agentic Applications: personalized tutoring, social skill development, and editable artifact creation.
    Selected work: GenMentor (WWW 2025 · Oral) · SocialCoach · EvoDesign (NeurIPS 2026).

Previously, I worked on:

  • Web3 & Financial Intelligence: NFT market modeling and risk-aware portfolio management.
    Selected work: COMET (KDD 2024) · MILLION (VLDB 2025).

  • AI for Combinatorial Optimization: reinforcement learning for network resource allocation.
    Selected work: Virne (ICLR 2026) · CONAL (ICMLW 2025 · Best Paper) · FlagVNE (IJCAI 2024).

I'm happy to discuss research and open-source collaborations in these areas.

1,767 stars       42 public repos       115 followers

Pinned Loading

  1. MindOS MindOS Public

    MindOS is a Human-AI Collaborative Mind System, where human thinks and agents act. Globally sync your mind for all agents: transparent, controllable, and evolving symbiotically.

    TypeScript 679 59

  2. virne virne Public

    [ICLR '26 - Virne] A simulator & benchmark for resource allocation (RA) problems in network function virtualization (NFV), i.e., NFV-RA, including virtual network embedding, service function chain …

    Python 160 39

  3. SocialCoach SocialCoach Public

    Your personal EQ coach. Build social skills and learn to handle conflict through realistic role-play and personalized feedback.

    TypeScript 211 29

  4. gen-mentor gen-mentor Public

    [WWW '25 Oral - GenMentor] Official code of our paper "LLM-powered Multi-agent Framework for Goal-oriented Learning in Intelligent Tutoring System", accepted by WWW 2025 (Industry Track) as an Oral…

    TypeScript 132 22

  5. awesome-ai-llm4education awesome-ai-llm4education Public

    Awesome artificial intelligence (AI) and large language model (LLM) for education papers.

    Jupyter Notebook 220 26

  6. FrameLark FrameLark Public

    帧好 · FrameLark — 摄影伙伴小帧：看图、找拍法、主题选片与精修。Web 工作台、摄影眼 Skill 与本地修片工具。

    JavaScript 1