florentdelgrange/wae_mdp

Wasserstein Auto-encoded Markov Decision Processes: a framework to distill Reinforcement Learning policies into verifiable controllers with bisimulation guarantees

Jupyter Notebook