RECAST: Model Reconstruction via Counterfactual-Aware Wasserstein Geometry under Limited Data

Xuan Zhao, Lena Krieger, Zhuo Cao, Arya Bangun, Hanno Scharr, Ira Assent
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:161567-161591, 2026.

Abstract

Counterfactual explanations (CFs) help understand machine learning models by identifying minimal input changes that would lead to alternative model outcomes. Recent work demonstrates their utility for reconstructing black-box models, enabling third-party auditing of opaque decision systems for fairness and accountability. Still, CF-based reconstruction may suffer from decision boundary shifts, overfitting, and restrictive assumptions requiring online query access to target platforms. We propose REconstruction via Counterfactual-Aware waSserstein opTimization (RECAST) under limited data and restricted access, a behavioral surrogate model based on Wasserstein barycenteric prototypes. Our approach addresses decision boundary shifts by incorporating CFs as informative, though less representative, samples for both classes, maintaining high surrogate fidelity in low-sample regimes without requiring online access during reconstruction. To enhance fairness auditing, our method enables systematic group fairness diagnostics. Experiments on real-world datasets and various setups show that RECAST effectively achieves high fidelity and query efficiency, as well as stable results even when the access is limited and noisy.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-zhao26u, title = {{RECAST}: Model Reconstruction via Counterfactual-Aware {W}asserstein Geometry under Limited Data}, author = {Zhao, Xuan and Krieger, Lena and Cao, Zhuo and Bangun, Arya and Scharr, Hanno and Assent, Ira}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {161567--161591}, 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/zhao26u/zhao26u.pdf}, url = {https://proceedings.mlr.press/v306/zhao26u.html}, abstract = {Counterfactual explanations (CFs) help understand machine learning models by identifying minimal input changes that would lead to alternative model outcomes. Recent work demonstrates their utility for reconstructing black-box models, enabling third-party auditing of opaque decision systems for fairness and accountability. Still, CF-based reconstruction may suffer from decision boundary shifts, overfitting, and restrictive assumptions requiring online query access to target platforms. We propose REconstruction via Counterfactual-Aware waSserstein opTimization (RECAST) under limited data and restricted access, a behavioral surrogate model based on Wasserstein barycenteric prototypes. Our approach addresses decision boundary shifts by incorporating CFs as informative, though less representative, samples for both classes, maintaining high surrogate fidelity in low-sample regimes without requiring online access during reconstruction. To enhance fairness auditing, our method enables systematic group fairness diagnostics. Experiments on real-world datasets and various setups show that RECAST effectively achieves high fidelity and query efficiency, as well as stable results even when the access is limited and noisy.} }
Endnote
%0 Conference Paper %T RECAST: Model Reconstruction via Counterfactual-Aware Wasserstein Geometry under Limited Data %A Xuan Zhao %A Lena Krieger %A Zhuo Cao %A Arya Bangun %A Hanno Scharr %A Ira Assent %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-zhao26u %I PMLR %P 161567--161591 %U https://proceedings.mlr.press/v306/zhao26u.html %V 306 %X Counterfactual explanations (CFs) help understand machine learning models by identifying minimal input changes that would lead to alternative model outcomes. Recent work demonstrates their utility for reconstructing black-box models, enabling third-party auditing of opaque decision systems for fairness and accountability. Still, CF-based reconstruction may suffer from decision boundary shifts, overfitting, and restrictive assumptions requiring online query access to target platforms. We propose REconstruction via Counterfactual-Aware waSserstein opTimization (RECAST) under limited data and restricted access, a behavioral surrogate model based on Wasserstein barycenteric prototypes. Our approach addresses decision boundary shifts by incorporating CFs as informative, though less representative, samples for both classes, maintaining high surrogate fidelity in low-sample regimes without requiring online access during reconstruction. To enhance fairness auditing, our method enables systematic group fairness diagnostics. Experiments on real-world datasets and various setups show that RECAST effectively achieves high fidelity and query efficiency, as well as stable results even when the access is limited and noisy.
APA
Zhao, X., Krieger, L., Cao, Z., Bangun, A., Scharr, H. & Assent, I.. (2026). RECAST: Model Reconstruction via Counterfactual-Aware Wasserstein Geometry under Limited Data. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:161567-161591 Available from https://proceedings.mlr.press/v306/zhao26u.html.

Related Material