Adaptive Mixing of Policies from Searching and Policies from Learning

Excerpt

arXiv:2608.15700v1 Announce Type: new Abstract: Background: Distillation of training targets generated thru search/planning has proven useful in reinforcement learning, but search can take exceedingly long. Objectives: Rather than perform search to the same depth every time (typically at a fixed period of steps), reduce the search depth proportionally to the quality of the policy network priors. Methods: We describe Flexer, an architecture that, for each step, mixes the policy from a neural netw