Costate-focused models for reinforcement learning

OffRL
Re-assign community