Enhancing the Hierarchical Environment Design via Generative Trajectory Modeling

arXiv:2310.00301v2 Announce Type: replace
Abstract: Unsupervised Environment Design (UED) is a paradigm for automatically generating a curriculum of training environments, enabling agents trained in these environments to develop general capabilities, i.e., achieving good zero-shot transfer performance. However, existing UED approaches focus primarily on the random generation of environments for open-ended agent training. This is impractical in scenarios with limited resources, such as the constraints on the number of generated environments. In this paper, we introduce a hierarchical MDP framework for …

abstract agent agents arxiv capabilities cs.ai cs.lg curriculum design enabling environment environments focus general generative good hierarchical modeling paradigm performance random training trajectory transfer type unsupervised via zero-shot