01
Learning
I am currently studying reinforcement learning as a more serious foundation for my next stage of research and technical growth.
Current focus
Embodied AI, reinforcement learning, and agent-native workflows
This page is a lightweight snapshot of what I am actively studying, building, and trying to grow into at the moment.
Reinforcement Learning
Dexterous Manipulation
Embodied Intelligence
AI Agent Coding
02
Building
My current hands-on task is dexterous in-hand object rotation, which is pushing me to connect ideas from embodiment, control, and practical experimentation.
03
Writing Next
I am preparing to write about embodied intelligence, especially from the perspective of how to build understanding gradually instead of treating it as a buzzword.
04
Tracking
Lately I have been closely following progress in embodied AI and the broader large-model space, especially AI agent coding and how these systems can become real research partners.
Next 1-2 Months
Build the missing foundation, then turn it into real capability.
My next goal is to systematically strengthen the deep learning, control, robotics kinematics, and dynamics knowledge required for embodied intelligence, while also building a full working software and hardware stack. In parallel, I want to understand AI agents more deeply and push them much further in supporting my learning and research workflow.