Hi — I write research notes about AI agents, multimodal systems, computer use, interpretability, and machine learning.
This site is intentionally simple: long-form technical writing, equations, diagrams, code, references, and experiment notes.
Research interests
- Multimodal and computer-use agents
- Long-horizon GUI/CLI interaction
- Agent harnesses and self-improvement
- Mechanistic interpretability
- Reinforcement learning and post-training
Contact
Replace this section with your email, Google Scholar, GitHub, and other links.