1 paper · 1 filter
Boyuan Xiao, Pu Wang, Ji Feng +1
World models enable large language model (LLM) agents to anticipate action effects and make informed decisions. For generalization, it is desirable for such world models to transfe…