Portrait de Zahra Dehghani Tafti

Zahra Dehghani Tafti

zahra.dehghani-tafti@mila.quebec
Doctorat - École de technologie suprérieure
Superviseur⋅e principal⋅e
Pablo Piantanida
Sujets de recherche
Apprentissage profond
Désapprentissage
Théorie de l'apprentissage automatique
Vie privée
Vision par ordinateur
Google Scholar
LinkedIn
GitHub

Publications

A Universal Source-Free Class Unlearning Framework via Synthetic Embeddings
Zahra Dehghani Tafti
Pablo Piantanida
Mohammadhadi Shateri
Class unlearning in neural classifiers refers to selectively removing the model’s ability to recognize a target (forget) class by reshapin… (voir plus)g the decision boundaries. This is essential when taxonomies change, labels are corrected, or legal or ethical requirements mandate class removal. The objective is to preserve performance on the remaining (retain) classes while avoiding costly full retraining. Existing methods generally require access to the source, i.e., forget/retain data or a relevant surrogate dataset. This dependency limits their applicability in scenarios where access to source data is restricted or unavailable. Even the recent source-free class unlearning methods rely on generating samples in the data space, which is computationally expensive and not even essential for doing class unlearning. In this work, we propose a novel source-free class unlearning framework that enables existing unlearning methods to operate using only the deployed model. We show that, under assumptions on the forget loss with respect to logits, class unlearning can be performed source-free for any given neural classifier by utilizing randomly generated samples within the classifier’s intermediate space. Specifically, randomly generated embeddings pseudo-labeled by the model as belonging to the forget or retain classes can support effective source-free unlearning. Our analysis further shows that, under conditions on the forget loss and synthetic forget embeddings, minimizing the forget loss induces expected logit shifts consistent with class unlearning, without requiring a specific parametric form of the embedding distribution. We validate our framework on four backbone architectures, ResNet-18, ResNet-50, ViT-B/16, and Swin-T, across three benchmark datasets, CIFAR-10, CIFAR-100, and TinyImageNet. Our experimental results show that existing class unlearning methods can operate within our source-free framework, with minimal impact on their forgetting efficacy and retain class accuracy. The code is available at https://github.com/Yasaman-dt/Source_Free_Class_Unlearning.
2026-05-18
Transactions on Machine Learning Research (accepté)
openreview.net
openreview.net