RT DW A1 Wang, Shu. T1 Enhancing the Reasoning Capabilities of Large Language Models: From Game Playing to Embodied Planning