What is causal about causal models and representations?

2025-01-31

AI Summary

Formalizing “actions” as interventions in causal Bayesian networks is essential to avoid circular reasoning, ensure falsifiability, and support causal understanding, discovery, and abstraction. Method: The authors introduce the first formal framework for intervention semantics, articulating rigorous rationality conditions. They prove that any intervention interpretation satisfying intuitive naturalness necessarily induces logical circularity, and that no interpretation can simultaneously satisfy causality, observability, and acyclicity. Contribution/Results: The work establishes novel falsifiability criteria for causal models, delineates fundamental theoretical limits of causal representation learning, provides a formal benchmark for causal abstraction, and exposes structural deficiencies in existing methods regarding empirical validation. It demonstrates that prevailing intervention-based approaches inherently compromise either causal fidelity, empirical testability, or semantic coherence—revealing an irreconcilable trade-off among core desiderata in causal modeling.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Causal Bayesian networks are 'causal' models since they make predictions about interventional distributions. To connect such causal model predictions to real-world outcomes, we must determine which actions in the world correspond to which interventions in the model. For example, to interpret an action as an intervention on a treatment variable, the action will presumably have to a) change the distribution of treatment in a way that corresponds to the intervention, and b) not change other aspects, such as how the outcome depends on the treatment; while the marginal distributions of some variables may change as an effect. We introduce a formal framework to make such requirements for different interpretations of actions as interventions precise. We prove that the seemingly natural interpretation of actions as interventions is circular: Under this interpretation, every causal Bayesian network that correctly models the observational distribution is trivially also interventionally valid, and no action yields empirical data that could possibly falsify such a model. We prove an impossibility result: No interpretation exists that is non-circular and simultaneously satisfies a set of natural desiderata. Instead, we examine non-circular interpretations that may violate some desiderata and show how this may in turn enable the falsification of causal models. By rigorously examining how a causal Bayesian network could be a 'causal' model of the world instead of merely a mathematical object, our formal framework contributes to the conceptual foundations of causal representation learning, causal discovery, and causal abstraction, while also highlighting some limitations of existing approaches.

Problem

• Causal Bayesian Networks
• Action Representation
• Avoiding Circular Logic

Innovation

• Causal Bayesian Networks
• Intervention Mapping
• Model Validation