Submitted by Timo K. 6 Understanding and Improving Hyperbolic Deep Reinforcement Learning University of Vienna 6 2