Relational Structural Causal Models

Adiba Ejaz, Elias Bareinboim
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:27741-27793, 2026.

Abstract

An artificial intelligence must have a model of its environment that is causal, supporting reasoning about interventions and counterfactuals, and also combinatorial, supporting generalization to unseen combinations of objects. In this work, we formally study when and how such a model can be learned. We develop relational structural causal models, extending structural causal models (Pearl 2009) to settings where objects and their relations vary. First, we show how answers to not only causal but also observational queries about unseen combinations of objects can not be identified without further assumptions. To enable such identification—including in the presence of unobserved confounding—we define relational causal graphs and derive symbolic identification criteria. Finally, we propose relational neural causal models, a provably correct approach that outperforms non-relational baselines on simulated traffic scenes with varying cars, signals, and pedestrians.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-ejaz26a, title = {Relational Structural Causal Models}, author = {Ejaz, Adiba and Bareinboim, Elias}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {27741--27793}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/ejaz26a/ejaz26a.pdf}, url = {https://proceedings.mlr.press/v306/ejaz26a.html}, abstract = {An artificial intelligence must have a model of its environment that is causal, supporting reasoning about interventions and counterfactuals, and also combinatorial, supporting generalization to unseen combinations of objects. In this work, we formally study when and how such a model can be learned. We develop relational structural causal models, extending structural causal models (Pearl 2009) to settings where objects and their relations vary. First, we show how answers to not only causal but also observational queries about unseen combinations of objects can not be identified without further assumptions. To enable such identification—including in the presence of unobserved confounding—we define relational causal graphs and derive symbolic identification criteria. Finally, we propose relational neural causal models, a provably correct approach that outperforms non-relational baselines on simulated traffic scenes with varying cars, signals, and pedestrians.} }
Endnote
%0 Conference Paper %T Relational Structural Causal Models %A Adiba Ejaz %A Elias Bareinboim %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-ejaz26a %I PMLR %P 27741--27793 %U https://proceedings.mlr.press/v306/ejaz26a.html %V 306 %X An artificial intelligence must have a model of its environment that is causal, supporting reasoning about interventions and counterfactuals, and also combinatorial, supporting generalization to unseen combinations of objects. In this work, we formally study when and how such a model can be learned. We develop relational structural causal models, extending structural causal models (Pearl 2009) to settings where objects and their relations vary. First, we show how answers to not only causal but also observational queries about unseen combinations of objects can not be identified without further assumptions. To enable such identification—including in the presence of unobserved confounding—we define relational causal graphs and derive symbolic identification criteria. Finally, we propose relational neural causal models, a provably correct approach that outperforms non-relational baselines on simulated traffic scenes with varying cars, signals, and pedestrians.
APA
Ejaz, A. & Bareinboim, E.. (2026). Relational Structural Causal Models. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:27741-27793 Available from https://proceedings.mlr.press/v306/ejaz26a.html.

Related Material