Full models of the world require complex knowledge of immense detail. While pre-trained large models have been hypothesized to contain similar knowledge due to extensive pre-training on vast amounts of internet scale data, using them directly in a search procedure is inefficient and inaccurate. Conversely, partial models focus on making high quality predictions for a subset of state and actions: those linked through affordances that achieve user intents~\citep{khetarpal2020can}. Can we posit large models as partial world models? We provide a formal answer to this question, proving that agents achieving task-agnostic, language-conditioned intents necessarily possess predictive partial-world models informed by affordances. In the multi-task setting, we introduce distribution-robust affordances and show that partial models can be extracted to significantly improve search efficiency. Empirical evaluations in tabletop robotics tasks demonstrate that our affordance-aware partial models reduce the search branching factor and achieve higher rewards compared to full world models.
@article{arxiv.2602.10390,
title = {Affordances Enable Partial World Modeling with LLMs},
author = {Khimya Khetarpal and Gheorghe Comanici and Jonathan Richens and Jeremy Shar and Fei Xia and Laurent Orseau and Aleksandra Faust and Doina Precup},
journal= {arXiv preprint arXiv:2602.10390},
year = {2026}
}