Teleological Inference in Structural Causal Models via Intentional Interventions

This paper extends structural causal models (SCMs)—a tool typically used to understand cause-and-effect relationships—to answer questions about agent intentions and goals. The researchers introduce a new concept called 'intentional interventions' that can detect when a goal-directed agent has deliberately changed a system and infer what that agent was trying to achieve by comparing what actually happened to what would have happened without intervention.