Dyna-Style Safety Augmented Reinforcement Learning: Staying Safe in the Face of Uncertainty — Artur Eisele, Bernd Frauenknecht, Friedrich Solowjow, Sebastian Trimpe | Kutubxona