Learning to Theorize the World from Observation

What does it mean for an AI system to understand the world? Contemporary world models often operationalize understanding as accurate prediction in latent or observation space. Developmental cognitive science, however, suggests a different view: human understanding emerges through the construction of internal theories of how the world works, even before mature language is acquired. Inspired by this theory-building view of cognition, we introduce Learning-to-Theorize (L2T), a learning paradigm for inferring explicit explanatory theories of the world from raw, non-textual observations. In L2T, a theory is represented as an executable, compositional program whose learned primitives can be systematically recombined to explain novel phenomena. We instantiate this paradigm with the Neural Theorizer, a probabilistic neural model that induces latent programs as a learned Language of Thought and executes them through a shared transition model. In experiments, we show that this formulation enables explanation-driven generalization, allowing observations to be understood in terms of the programs that generate.