Skip to content
View dragonbra's full-sized avatar
🎵
🎵

Block or report dragonbra

Report abuse

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

Report abuse
dragonbra/README.md

Hey, I'm dragonbra 👋

I'm a Quant Researcher working with deep learning, model architecture, and experimentation.

🏁 Kaggle · 💼 LinkedIn · ✍️ Blog · ⚔️ Codeforces

  • 👤 Background

    • 💼 Work: Quant Researcher.
    • 🔬 Earlier research: AI infrastructure and AutoML.
    • 🎓 Education: M.S. in Computer Science at Xiamen University.
    • 🟠 Codeforces: Master with a rating of 2174.
    • 🥈 Kaggle: Silver Medal, 91st out of 2,665 teams, in the Kaggle - LLM Science Exam (team solution).
    • 🏆 Earlier: Champion of Meta-learning from Learning Curves, 2nd Round.
  • 🔍 Interests

    • 📈 Quantitative research with deep learning models and careful experimentation.
    • 🧠 Model architecture, reinforcement learning, agent systems, and research engineering.
    • 🃏 Treating Kaggle competitions as compact research environments—not just leaderboards.
  • 🚧 Recently working on

    • 🌾 Kaggriculture: exploring an agent world through execution, planning, economic reasoning, strategy, and learning.
    • 🤖🛠️ Working with Codex and Claude Code as research collaborators—and turning their output into artifacts I can inspect, correct, and verify.

Pinned Loading

  1. MetaLC-2nd-Round MetaLC-2nd-Round Public

    Winning method (1st place) in Meta-learning from Learning Curves - 2ND ROUND competition.

    Python 6 2

  2. dongzelian/SSF dongzelian/SSF Public

    [NeurIPS'22] This is an official implementation for "Scaling & Shifting Your Features: A New Baseline for Efficient Model Tuning".

    Python 200 16

  3. 490CAD/LLM4Science 490CAD/LLM4Science Public

    kaggle LLM contest Silver solution

    Python 6 3

  4. zhanglei1172/XBBO zhanglei1172/XBBO Public

    Black Box Optimization

    Python 7