Study reinforcement learning for coding agents, graders, evals and training systems, then run a mock research deep-dive in the Research Scientist capstone.