CutClean: Neural Network Pruning for Privacy-Preserving Inference

August 13, 2026 ยท Grace Period ยท ๐Ÿ› In: De Marsico, M., et al. Pattern Recognition. ICPR 2026. Lecture Notes in Computer Science, vol 16822. Springer, Cham

โณ Grace Period
This paper is less than 90 days old. We give authors time to release their code before passing judgment.
Authors Leonardo Magliolo, Vito Paolo Pastore, Giuseppe Valenzise, Enzo Tartaglione arXiv ID 2608.13773 Category cs.LG: Machine Learning Cross-listed cs.AI Citations 0 Venue In: De Marsico, M., et al. Pattern Recognition. ICPR 2026. Lecture Notes in Computer Science, vol 16822. Springer, Cham
Abstract
Neural networks are increasingly deployed in high-stakes applications with growing privacy leakage concerns. We show that this privacy leakage can occur even in the absence of representation imbalances that lead to traditional dataset biases. This poses significant privacy risks when deploying models that process sensitive attributes. In this context, we propose CutClean, a privacy-aware pruning method that allows to reduce privacy information flow through the network, while increasing its sparsity. Our approach employs auxiliary linear privacy heads placed at each network's block to quantify information leakage, and further applies increasing levels of sparsity to remove the private attribute leakage, measured in terms of the accuracy of the privacy head attached to the last block. Experiments on synthetic and real-world datasets demonstrate that our approach effectively minimizes private information flow while achieving high sparsity rates and preserving classification target accuracy.
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

๐Ÿ“œ Similar Papers

In the same crypt โ€” Machine Learning