WorldFoundry offers open-source infrastructure for reproducible inference and evaluation of generative world models across video, 3D/4D, embodied action, and interactive systems, with traceable evidence and benchmarks.
Visual generative models increasingly blur the line between synthesizing observations and simulating environments. Yet visual plausibility alone does not constitute successful world modeling: compelling outputs may lack persistent state, physical dynamics, or causally consistent responses to interactions. We define a generative world model as a predictive dynamical system that maintains a persistent, history-conditioned representation of an environment, updates that representation in response to actions or interventions, and generates temporally coherent trajectories that reflect outcomes under observed and counterfactual interventions. From this definition, we derive an…
visual generative modelsgenerative world modelpredictive dynamical systempersistent history-conditioned representationtemporally coherent trajectoriescounterfactual interventionsperceptual groundingpersistent state modelingcontrollable transformationdynamics-aware prediction
- Upvotes
- 17
- Authors
- 16
- Published
- Nov 7, 2026
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.