Exponential Speedups by Rerooting Levin Tree Search

Levin Tree Search (LTS) (Orseau et al., 2018) is a search algorithm for deterministic environments that uses a user-specified policy to guide the search. It comes with a formal guarantee on the number of search steps (node visits) for finding a solution node that depends on the quality of the policy. In this paper, we introduce a new algorithm, called √LTS (pronounce root-LTS), which implicitly starts an LTS search rooted at every node of the search tree. Each LTS search is assigned a rerooting weight by a (user-defined or learnt) rerooter, and the search effort is shared between all LTS searches proportionally to their weights. The rerooting mechanism implicitly decomposes the search space into subtasks, leading to significant speedups. We prove that the number of node visits that √LTS takes is competitive with the best decomposition into subtasks, at the price of a factor that relates to the uncertainty of the rerooter. If LTS takes time T, in the best case with q rerooting points, √LTS only takes time O(q√[q]T). Like the policy, the rerooter can be learnt from data, and we expect √LTS to be applicable to a wide range of domains.