Towards a Research Community in Interpretable Reinforcement Learning: the InterpPol Workshop