1 paper
Yuhan Guo, Jinming Liu, Liang Xu +8
Large language models (LLMs) are increasingly deployed as agents for multi-step decision-making, yet transfer poorly to unseen environments. World-model methods address this by tra…