Spectral Normalisation for Deep Reinforcement Learning:
                An Optimisation Perspective

      Florin Gogianu * 1 2 Tudor Berariu * 3 Mihaela Rosca 4 5 Claudia Clopath 3 4 Lucian Busoniu 2
                          Razvan Pascanu 4

              Abstract
    Most of the recent deep reinforcement learning
    advances take an RL-centric perspective and fo-
    cus on refinements of the training objective. We
    diverge from this view and show we can recover
    the performance of th ...