Scalable Contrastive Causal Discovery under Unknown Soft Interventions

TL;DR

This paper introduces a scalable method for causal discovery that effectively integrates observational and interventional data with unknown soft interventions, improving structure recovery and generalization.

Contribution

It proposes a novel contrastive causal discovery model that handles unknown soft interventions and shared causal structures, with theoretical guarantees and improved empirical performance.

Findings

Enhanced causal structure recovery on synthetic data

Better generalization to unseen graphs and causal mechanisms

Scalability to larger graphs demonstrated

Abstract

Observational causal discovery is only identifiable up to the Markov equivalence class. While interventions can reduce this ambiguity, in practice interventions are often soft with multiple unknown targets. In many realistic scenarios, only a single intervention regime is observed. We propose a scalable causal discovery model for paired observational and interventional settings with shared underlying causal structure and unknown soft interventions. The model aggregates subset-level PDAGs and applies contrastive cross-regime orientation rules to construct a globally consistent maximal PDAG under Meek closure, enabling generalization to both in-distribution and out-of-distribution settings. Theoretically, we prove that our model is sound with respect to a restricted $\Psi$ equivalence class induced solely by the information available in the subset-restricted setting. We further show that the model asymptotically recovers the corresponding identifiable PDAG and can orient additional edges compared to non-contrastive subset-restricted methods. Experiments on synthetic data demonstrate improved causal structure recovery, generalization to unseen graphs with held-out causal mechanisms, and scalability to larger graphs, with ablations supporting the theoretical results.