This repository contains the official implementation for the paper:

Machine Unlearning in the Era of Quantum Machine Learning: An Empirical Study

We present the first comprehensive empirical study of machine unlearning (MU) in hybrid quantum-classical neural networks. While MU has been extensively explored in classical deep learning, its behavior within variational quantum circuits (VQCs) and quantum-augmented architectures remains largely unexplored.

In this work, we:

• Adapt a broad suite of machine unlearning methods to quantum settings, including:
  • gradient-based methods
  • distillation-based methods
  • regularization-based methods
  • certified unlearning techniques
• Propose two novel unlearning strategies specifically tailored for hybrid quantum–classical models.
• Evaluate unlearning under both subset removal and full-class deletion scenarios.

Experiments are conducted on Iris, MNIST, and Fashion-MNIST datasets using hybrid quantum-classical neural networks. Results show that quantum models can support effective unlearning, but outcomes depend strongly on:

• circuit depth
• entanglement structure
• task complexity

Shallow VQCs exhibit high intrinsic stability with limited memorization, while deeper hybrid models reveal stronger trade-offs between utility, forgetting strength, and alignment with oracle retraining. Across settings, methods such as EU-k, LCA, and Certified Unlearning provide the most consistent balance across metrics.

This repository is intended to serve as a baseline and benchmark for future research in quantum machine unlearning.