The Wheel of Intelligence (WoI): A State-Transition Framework for Building Artificial Minds

This paper presents the Wheel of Intelligence (WoI), an architectural framework that models cognition as a recurrent state-transition process over a shared state contract. WoI defines seven operators-Perception (P), Knowledge Representation (KR), Memory (M), Reasoning (R), Learning (L), Self-Reflection (SR), and Actuation (A)-and treats intelligence as the ability to execute this loop stably, adaptively, and safely across tasks. A core thesis is KR centrality: cross-operator coordination is mediated through a shared KR substrate rather than hidden pairwise couplings. We contribute a unified state contract, operator definitions with explicit artifact routing, a compositional formal model, and a competence view that links local operator quality to system-level performance. We also describe governance patterns centered on SR-gated uncertainty handling and provide appendices with a mathematical addendum, implementation pathways, and an evaluation plan. The aim is to provide a rigorous, implementation-ready preprint that can guide both software agents and embodied systems.