Pre-trained deep image representations are useful for post-training task...
Bilevel planning, in which a high-level search over an abstraction of an...
Effective and efficient planning in continuous state and action spaces i...
Sparse rewards and long time horizons remain challenging for reinforceme...