Curiosity-based topological reinforcement learning | IEEE Conference Publication | IEEE Xplore