Optimal Conservative Offline RL with General Function Approximation via Augmented Lagrangian

Open in new window