1 paper
Yinhao Li, Weixin Mao, Zihan Lan +21
Vision-language-action (VLA) models, world-action models (WAMs), and offline reinforcement learning methods are rapidly expanding the design space of embodied policies, yet turning…