Efficient Robust Conformal Prediction via Lipschitz-Bounded Networks

Thomas Massena, Léo Andéol, Thibaut Boissin, Franck Mamalet, Corentin Friedrich, Mathieu Serrurier, Sébastien Gerchinovitz
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:43225-43247, 2025.

Abstract

Conformal Prediction (CP) has proven to be an effective post-hoc method for improving the trustworthiness of neural networks by providing prediction sets with finite-sample guarantees. However, under adversarial attacks, classical conformal guarantees do not hold anymore: this problem is addressed in the field of Robust Conformal Prediction. Several methods have been proposed to provide robust CP sets with guarantees under adversarial perturbations, but, for large scale problems, these sets are either too large or the methods are too computationally demanding to be deployed in real life scenarios. In this work, we propose a new method that leverages Lipschitz-bounded networks to precisely and efficiently estimate robust CP sets. When combined with a 1-Lipschitz robust network, we demonstrate that our lip-rcp method outperforms state-of-the-art results in both the size of the robust CP sets and computational efficiency in medium and large-scale scenarios such as ImageNet. Taking a different angle, we also study vanilla CP under attack, and derive new worst-case coverage bounds of vanilla CP sets, which are valid simultaneously for all adversarial attack levels. Our lip-rcp method makes this second approach as efficient as vanilla CP while also allowing robustness guarantees.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-massena25a, title = {Efficient Robust Conformal Prediction via {L}ipschitz-Bounded Networks}, author = {Massena, Thomas and And\'{e}ol, L\'{e}o and Boissin, Thibaut and Mamalet, Franck and Friedrich, Corentin and Serrurier, Mathieu and Gerchinovitz, S\'{e}bastien}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {43225--43247}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/massena25a/massena25a.pdf}, url = {https://proceedings.mlr.press/v267/massena25a.html}, abstract = {Conformal Prediction (CP) has proven to be an effective post-hoc method for improving the trustworthiness of neural networks by providing prediction sets with finite-sample guarantees. However, under adversarial attacks, classical conformal guarantees do not hold anymore: this problem is addressed in the field of Robust Conformal Prediction. Several methods have been proposed to provide robust CP sets with guarantees under adversarial perturbations, but, for large scale problems, these sets are either too large or the methods are too computationally demanding to be deployed in real life scenarios. In this work, we propose a new method that leverages Lipschitz-bounded networks to precisely and efficiently estimate robust CP sets. When combined with a 1-Lipschitz robust network, we demonstrate that our lip-rcp method outperforms state-of-the-art results in both the size of the robust CP sets and computational efficiency in medium and large-scale scenarios such as ImageNet. Taking a different angle, we also study vanilla CP under attack, and derive new worst-case coverage bounds of vanilla CP sets, which are valid simultaneously for all adversarial attack levels. Our lip-rcp method makes this second approach as efficient as vanilla CP while also allowing robustness guarantees.} }
Endnote
%0 Conference Paper %T Efficient Robust Conformal Prediction via Lipschitz-Bounded Networks %A Thomas Massena %A Léo Andéol %A Thibaut Boissin %A Franck Mamalet %A Corentin Friedrich %A Mathieu Serrurier %A Sébastien Gerchinovitz %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-massena25a %I PMLR %P 43225--43247 %U https://proceedings.mlr.press/v267/massena25a.html %V 267 %X Conformal Prediction (CP) has proven to be an effective post-hoc method for improving the trustworthiness of neural networks by providing prediction sets with finite-sample guarantees. However, under adversarial attacks, classical conformal guarantees do not hold anymore: this problem is addressed in the field of Robust Conformal Prediction. Several methods have been proposed to provide robust CP sets with guarantees under adversarial perturbations, but, for large scale problems, these sets are either too large or the methods are too computationally demanding to be deployed in real life scenarios. In this work, we propose a new method that leverages Lipschitz-bounded networks to precisely and efficiently estimate robust CP sets. When combined with a 1-Lipschitz robust network, we demonstrate that our lip-rcp method outperforms state-of-the-art results in both the size of the robust CP sets and computational efficiency in medium and large-scale scenarios such as ImageNet. Taking a different angle, we also study vanilla CP under attack, and derive new worst-case coverage bounds of vanilla CP sets, which are valid simultaneously for all adversarial attack levels. Our lip-rcp method makes this second approach as efficient as vanilla CP while also allowing robustness guarantees.
APA
Massena, T., Andéol, L., Boissin, T., Mamalet, F., Friedrich, C., Serrurier, M. & Gerchinovitz, S.. (2025). Efficient Robust Conformal Prediction via Lipschitz-Bounded Networks. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:43225-43247 Available from https://proceedings.mlr.press/v267/massena25a.html.

Related Material