Anthony W. Jung

I'm interested in various aspects of ML, especialy RL, to solve important problems in the world.

Apr 10, 2025 My failure log
It's hard to digest that I've failed. These are fragments from years past—a reflection I hope turns out to be meaningful progress.
Apr 10, 2025 Policy gradient
Don't ever forget about policy gradient.
Apr 10, 2025 Function approximation and representation learning in RL
The discrepancy between linear RL and deep RL, and intuitive understanding of representation in RL.
Apr 9, 2025 State vs Observation in RL
Apr 9, 2025 Can we trust math?
Apr 9, 2025 [Editing] From TRPO to PPO
Many people use PPO without understanding it, but that's why it's such a good algorithm.
Apr 9, 2025 [Editing] Overdetermined vs Overcomplete vs Overparameterized
Apr 9, 2025 [Editing] Most of the deep RL algorithms are TD