This week, we’ll be discussing Discovering Agents (Zachary et al. 2022), a recent paper from the DeepMind Safety team. Abstract:
Causal models of agents have been used to analyse the safety aspects of machine learning systems. But identifying agents is non-trivial—often the causal model is just assumed by the modeler without much justification—and modelling failures can lead to mistakes in the safety analysis. This paper proposes the first formal causal definition of agents—roughly that agents are systems that would adapt their policy if their actions influenced the world in a different way. From this we derive the first causal discovery algorithm for discovering agents from empirical data, and give algorithms for translating between causal models and game-theoretic influence diagrams. We demonstrate our approach by resolving some previous confusions caused by incorrect causal modelling of agents.