A Knowledge-Informed Pretrained Model for Causal Discovery

TL;DR

This paper introduces a knowledge-informed pretrained model for causal discovery that effectively integrates weak prior knowledge, improving robustness and applicability across diverse datasets and scenarios.

Contribution

It proposes a novel dual encoder-decoder architecture with a curriculum learning strategy to incorporate coarse domain knowledge into causal discovery.

Findings

Consistent improvements over baseline methods in various datasets.

Robust performance in out-of-distribution scenarios.

Effective integration of weak prior knowledge enhances practical deployment.

Abstract

Causal discovery has been widely studied, yet many existing methods rely on strong assumptions or fall into two extremes: either depending on costly interventional signals or partial ground truth as strong priors, or adopting purely data driven paradigms with limited guidance, which hinders practical deployment. Motivated by real-world scenarios where only coarse domain knowledge is available, we propose a knowledge-informed pretrained model for causal discovery that integrates weak prior knowledge as a principled middle ground. Our model adopts a dual source encoder-decoder architecture to process observational data in a knowledge-informed way. We design a diverse pretraining dataset and a curriculum learning strategy that smoothly adapts the model to varying prior strengths across mechanisms, graph densities, and variable scales. Extensive experiments on in-distribution, out-of distribution, and real-world datasets demonstrate consistent improvements over existing baselines, with strong robustness and practical applicability.