2606.19594

Total: 1

#1 Unsupervised Causal Abstractions Discovery [PDF] [Copy] [Kimi] [REL]

Authors: Théo Saulus, Simon Lacoste-Julien, Dhanya Sridhar

Causal abstractions formalize when a high-level structural causal model (SCM) captures the interventional behavior of a lower-level SCM. Existing applications of this notion largely follow a hypothesis-testing paradigm: an expert proposes a candidate high-level model and then evaluates if the low-level system implements it. We study the complementary problem of learning a high-level model directly from low-level measurements. Our contributions leverage hypotheses from low-rank causal discovery, and can be summarized as follows: (1) we show that observations generated by a low-rank graph induce latents that form a causal abstraction, (2) we provide identifiability results about these latents, and (3) we propose a practical objective to learn this high-level SCM.

Subject: Machine Learning

Publish: 2026-06-17 20:56:44 UTC