f-Policy Gradients: A General Framework for Goal Conditioned RL using f-Divergences

Open in new window