Writing
Technical notes and updates on reinforcement learning, reward design, simulation workflows, and evaluation.
A measurement plan for locating where useful variation disappears between reward, advantage estimation, and policy updates.
Our paper on surgical assistance agents in simulation was published in the