What I am drawn to
AI, robotics, and computer systems. Why an algorithm works, why a program fails, and how a system behaves under real constraints are questions worth taking apart carefully.
What you will find here
Research ideas, engineering attempts, course projects, and technical notes. Some can already be explained through a small experiment. Others are questions that still need evidence.
I want each project to make four things clear: the problem, the work, the evidence, and what remains unknown.
Currently exploring
- Feedback, failure analysis, and repair in agent and robot execution.
- The mechanisms inside language models, alongside fundamental algorithms I can implement myself.
- Visual ways to learn compilers and computer systems.
Leave room for curiosity
Learning rarely follows a straight line. Keeping the intermediate steps helps my future self, and may give someone else a starting point for the same question.