AI Researcher / Software Engineer / Builder

I study how AI systems learn from data, improve through feedback, and act in interactive environments.

I'm pursuing a B.S. in Computer Science at the University of Southern California ✌️

My current research interests are post-training, data-centric ML, and model evaluation. Previously, I was a technical staff intern at Kimi (Moonshot AI), where I worked on agent evaluation infrastructure and harnesses. At USC, I work with Yue Zhao and Jieyu Zhao.

Reach out to me at siuw [at] usc [dot] edu

News

  • Jul 27, 2026The Kimi K3 tech report is out! Our 2.8T open frontier model with fully open weights is now available on Hugging Face. Check out the paper.
  • Jun 25, 2026OpenThoughts-Agent is out: open datasets and recipes for training agentic models. Check out the paper and our OpenThinkerAgent-32B model.

Research Publications

* Equal contribution

Blog