Causally Consistent Normalizing Flow
Qingyang Zhou, Kangjie Lu, Meng Xu
Abstract
Causal inconsistency arises when the underlying causal graphs captured by generative models like Normalizing Flows are inconsistent with those specified in causal models like Struct Causal Models. This inconsistency can cause unwanted issues including unfairness. Prior works to achieve causal consistency inevitably compromise the expressiveness of their models by disallowing hidden layers. In this work, we introduce a new approach: Causally Consistent Normalizing Flow (CCNF). To the best of our knowledge, CCNF is the first causally consistent generative model that can approximate any distribution with multiple layers. CCNF relies on two novel constructs: a sequential representation of SCMs and partial causal transformations. These constructs allow CCNF to inherently maintain causal consistency without sacrificing expressiveness. CCNF can handle all forms of causal inference tasks, including interventions and counterfactuals. Through experiments, we show that CCNF outperforms current approaches in causal inference. We also empirically validate the practical utility of CCNF by applying it to real-world datasets and show how CCNF addresses challenges like unfairness effectively.
BibTeX
@article{Zhou_Lu_Xu_2025, title={Causally Consistent Normalizing Flow}, volume={39}, url={https://ojs.aaai.org/index.php/AAAI/article/view/34460}, DOI={10.1609/aaai.v39i21.34460}, abstractNote={Causal inconsistency arises when the underlying causal graphs captured by generative models like Normalizing Flows are inconsistent with those specified in causal models like Struct Causal Models.
This inconsistency can cause
unwanted issues including unfairness.
Prior works to achieve causal consistency inevitably compromise the expressiveness of their models
by disallowing hidden layers.
In this work, we introduce a new approach: Causally Consistent Normalizing Flow (CCNF).
To the best of our knowledge, CCNF is the first causally consistent generative model that can approximate any distribution with multiple layers.
CCNF relies on two novel constructs:
a sequential representation of SCMs and
partial causal transformations.
These constructs allow CCNF
to inherently maintain causal consistency
without sacrificing expressiveness.
CCNF can handle all forms of causal inference tasks, including interventions and counterfactuals.
Through experiments,
we show that CCNF outperforms current approaches in causal inference.
We also empirically validate the practical utility of CCNF
by applying it to real-world datasets
and show how CCNF
addresses challenges like unfairness effectively.}, number={21}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Zhou, Qingyang and Lu, Kangjie and Xu, Meng}, year={2025}, month={Apr.}, pages={22974-22981} }