kkk's Notes · Machine Learning

Between experiments and papers,
thinking about how machines understand the world.

Notes on world models, embodied AI, and physical AI — what I ran, what I read, and what went wrong along the way.

Focus areas

Agents that learn in imagination — environment modeling, planning, and model-based RL.

Agents with bodies — perceiving, deciding and acting in real or simulated environments, and bringing models onto robots.

Learning the rules of the physical world — from simulation and dynamics modeling to AI that works reliably in reality.

First post — why this blog exists and how it was built.