The paper generated discussion among the reviewers. On the positive side, there are strong empirical results, and the ablation study included in the rebuttal allowed the reviewers to see the merits of the proposed approach on top of SPACE. On the other hand, the novelty of the approach was considered low by a couple of the reviewers. However, the use of the approach in MDPs was judged interesting by the other two reviewers. On the balance, in my opinion, the paper does a good job of proposing an extension to existing work that is sufficiently novel and works very well, and the experiments are well done and convincing, so I am recommending acceptance. Please take into account the reviewers' comments in revising the paper.