Deep Reinforcement Learning
Self-study of UC Berkeley CS285, with core deep RL algorithms implemented from scratch.
Self-study, Jul 2026 – present.
I am working through UC Berkeley’s CS285: Deep Reinforcement Learning and implementing its core algorithms from scratch:
- Behavioral cloning with a flow-matching policy
- Policy gradients
- Actor-critic methods
- Deep Q-learning
I trained and benchmarked the agents on environments such as PushT, InvertedPendulum, and LunarLander.