AI & Causality
Towards Error-Centric Intelligence II: Energy-Structured Causal Models
Causal models often describe variables and dependencies without giving a deep account of why some structures remain stable under intervention. This work develops Energy-Structured Causal Models as a way to connect causal explanation with constraints, locality, autonomy, and mechanism-level editability.
This paper develops Energy-Structured Causal Models as a formal language for causal explanation in systems designed to support intervention, mechanism-level editing, and criticism.
It extends the error-centric intelligence program by connecting causal semantics, energy functions, and structural principles such as locality, autonomy, and independent causal mechanisms.