Dissecting Causal Mechanism Shifts via FANS: Function And Noise Separation

Gyeongdeok Seo, Jaeyoon Shim, Mingyu Kim, Hoyoon Byun, Yonghan Jung, Kyungwoo Song
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:109074-109104, 2026.

Abstract

Identifying the drivers of causal mechanism shifts, distinguishing functional changes from noise alterations, termed dissection, is a critical yet under-explored problem in data science (e.g., biomedical science and manufacturing). This paper introduces a more general and unified framework, the function and noise separation (FANS) framework, that detects and dissects shifts in non-additive, non-linear Structural Causal Models (SCMs) beyond existing additive noise models. Our approach is grounded in a theoretical independence criterion, where function shifts induce a statistical dependence between a node’s parents and residual noise. Building on this foundation, we develop a practical two-stage algorithm to efficiently detect and dissect these shifts without retraining. Furthermore, we address the complex challenge of simultaneous function and noise shifts, introducing a formal assumption to resolve their inherent non-identifiability. Our results are corroborated by simulations. Our code is available at https://github.com/MLAI-Yonsei/FANS/.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-seo26d, title = {Dissecting Causal Mechanism Shifts via {FANS}: Function And Noise Separation}, author = {Seo, Gyeongdeok and Shim, Jaeyoon and Kim, Mingyu and Byun, Hoyoon and Jung, Yonghan and Song, Kyungwoo}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {109074--109104}, 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/seo26d/seo26d.pdf}, url = {https://proceedings.mlr.press/v306/seo26d.html}, abstract = {Identifying the drivers of causal mechanism shifts, distinguishing functional changes from noise alterations, termed dissection, is a critical yet under-explored problem in data science (e.g., biomedical science and manufacturing). This paper introduces a more general and unified framework, the function and noise separation (FANS) framework, that detects and dissects shifts in non-additive, non-linear Structural Causal Models (SCMs) beyond existing additive noise models. Our approach is grounded in a theoretical independence criterion, where function shifts induce a statistical dependence between a node’s parents and residual noise. Building on this foundation, we develop a practical two-stage algorithm to efficiently detect and dissect these shifts without retraining. Furthermore, we address the complex challenge of simultaneous function and noise shifts, introducing a formal assumption to resolve their inherent non-identifiability. Our results are corroborated by simulations. Our code is available at https://github.com/MLAI-Yonsei/FANS/.} }
Endnote
%0 Conference Paper %T Dissecting Causal Mechanism Shifts via FANS: Function And Noise Separation %A Gyeongdeok Seo %A Jaeyoon Shim %A Mingyu Kim %A Hoyoon Byun %A Yonghan Jung %A Kyungwoo Song %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-seo26d %I PMLR %P 109074--109104 %U https://proceedings.mlr.press/v306/seo26d.html %V 306 %X Identifying the drivers of causal mechanism shifts, distinguishing functional changes from noise alterations, termed dissection, is a critical yet under-explored problem in data science (e.g., biomedical science and manufacturing). This paper introduces a more general and unified framework, the function and noise separation (FANS) framework, that detects and dissects shifts in non-additive, non-linear Structural Causal Models (SCMs) beyond existing additive noise models. Our approach is grounded in a theoretical independence criterion, where function shifts induce a statistical dependence between a node’s parents and residual noise. Building on this foundation, we develop a practical two-stage algorithm to efficiently detect and dissect these shifts without retraining. Furthermore, we address the complex challenge of simultaneous function and noise shifts, introducing a formal assumption to resolve their inherent non-identifiability. Our results are corroborated by simulations. Our code is available at https://github.com/MLAI-Yonsei/FANS/.
APA
Seo, G., Shim, J., Kim, M., Byun, H., Jung, Y. & Song, K.. (2026). Dissecting Causal Mechanism Shifts via FANS: Function And Noise Separation. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:109074-109104 Available from https://proceedings.mlr.press/v306/seo26d.html.

Related Material