LATIDIA · Investigación

RoleWorld: Transferencia de conocimiento en el modelado del mundo con abstracciones de roles funcionales

arXiv: 2610.13172v1Tipo de anuncio: nuevo Resumen: Los modelos mundiales permiten a los agentes del modelo de lenguaje grande (LLM) anticiparse a los efectos de la acción y tomar decisiones informadas. Para la generalización, es deseable que tales modelos mundiales

La noticia

arXiv:2610.13172v1 Announce Type: new Abstract: 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 transfer efficiently to new environments with similar interaction structure and underlying dynamics. Recent approaches encode world dynamics as symbolic rules or executable programs, whose compositional structure in principle affords transfer, but still organize such knowledge around environment-specific objects. To address this limitation, we introduce RoleWorld, a symbolic world modeling approach which uses reusable prototypes and environment-specific parameter values as functional role abstractions to represent how objects behave in interactions. From interaction trajectories, a teacher LLM extracts role prototypes and