Jointly Learning Consistent Causal Abstractions Over Multiple Interventional Distributions

TL;DR

This paper introduces a framework for learning causal abstractions between structural causal models that ensures consistency across interventional distributions, enabling joint reasoning at multiple levels of system resolution.

Contribution

It presents the first formal framework for causal abstraction learning based on Rischel's formalization, along with a differentiable programming approach for joint learning.

Findings

The proposed method outperforms independent approaches on synthetic data.

It effectively handles a real-world electric vehicle battery manufacturing problem.

Joint learning improves consistency across multiple interventional distributions.

Abstract

An abstraction can be used to relate two structural causal models representing the same system at different levels of resolution. Learning abstractions which guarantee consistency with respect to interventional distributions would allow one to jointly reason about evidence across multiple levels of granularity while respecting the underlying cause-effect relationships. In this paper, we introduce a first framework for causal abstraction learning between SCMs based on the formalization of abstraction recently proposed by Rischel (2020). Based on that, we propose a differentiable programming solution that jointly solves a number of combinatorial sub-problems, and we study its performance and benefits against independent and sequential approaches on synthetic settings and on a challenging real-world problem related to electric vehicle battery manufacturing.