Year: 2025
Type: preprint
Abstract: World models have been recently proposed as sandbox environments in which AI agents can be trained and evaluated before deployment. While realistic world models often have high computational demands, this can often be alleviated by exploiting the fact that real-world scenarios tend to involve subcomponents that interact in a modular manner. In this paper, we explore this idea by developing a framework for decomposing complex world models represented by transducers, a class of models generalising POMDPs. Whereas the composition of transducers is well understood, our results clarify how to invert this process by deriving sub-transducers operating on distinct input-output subspaces, enabling parallelizable and interpretable alternatives to monolithic world modelling that can support distributed inference. Overall, these results lay groundwork for bridging the computational efficiency required for real-world inference and the structural transparency demanded by AI safety.
Source: arXiv (Cornell University)
Institutions: University of Sussex, University of Oxford, Araya Inc. (Japan)
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: G4511761941, G903899529
OpenAlex ID: W4417001667
Cites:
Cited by:
Related to: