面向强化学习的状态空间建模：RSSM的介绍和PyTorch实现
循环状态空间模型（Recurrent state Space Models, RSSM）最初由 Danijar Hafer 等人在论文《Learning Latent Dynamics for Planning from Pixels》中提出。该模型在现代基于
循环状态空间模型（Recurrent state Space Models, RSSM）最初由 Danijar Hafer 等人在论文《Learning Latent Dynamics for Planning from Pixels》中提出。该模型在现代基于