Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrapping (Searchformer)

Paper: https://arxiv.org/abs/2402.14083

Abstract:
While Transformers have enabled tremendous progress in various application settings, such architectures still lag behind traditional symbolic planners for solving complex decision making tasks. In this work, we demonstrate how to train Transformers to solve complex planning tasks and present Searchformer, a Transformer model that optimally solves previously unseen Sokoban puzzles 93.7% of the time, while using up to 26.8% fewer search steps than standard A∗ search. Searchformer is an encoder-decoder Transformer model trained to predict the …

abstract application architectures decision decision making making planning progress solve tasks train transformer transformer model transformers work