A Numerical Gradient Inversion Attack in Variational Quantum Neural-Networks
Not on the map. Why?
Abstract (arXiv)
The loss landscape of Variational Quantum Neural Networks (VQNNs) is characterized by local minima that grow exponentially with increasing qubits. Because of this, it is more challenging to recover information from model gradients during training compared to classical Neural Networks (NNs). In this paper we present a numerical scheme that successfully reconstructs input training, real-world, practical data from trainable VQNNs' gradients. Our scheme is based on gradient inversion that works by combining gradients estimation with the finite difference method and adaptive low-pass filtering. The scheme is further optimized with Kalman filter to obtain efficient convergence. Our experiments show that our algorithm can invert even batch-trained data, given the VQNN model is sufficiently over-parameterized.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 43 of its 91 references that are in the map. First posted on arXiv 17 Apr 2025.
Builds on
43 of 91 references.
References · 43
- 2014Adam: A Method for Stochastic Optimizationcited by 37,240
- 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
- 2016Towards Evaluating the Robustness of Neural Networkscited by 2,998
- 2019Deep Leakage from Gradientscited by 1,021
- 2020Inverting Gradients -- How easy is it to break privacy in federated learning?cited by 712
- 2020Variational Quantum Algorithmscited by 692
- 2018Barren plateaus in quantum neural network training landscapescited by 630
- 2019Federated Learning with Differential Privacy: Algorithms and Performance Analysiscited by 536
- 2019A Survey on Federated Learning Systems: Vision, Hype and Reality for Data Privacy and Protectioncited by 374
- 2021See through Gradients: Image Batch Recovery via GradInversioncited by 326
- 2019Differential Privacy Has Disparate Impact on Model Accuracycited by 316
- 2019The Secret Revealer: Generative Model-Inversion Attacks Against Deep Neural Networkscited by 246
31 more
- 2023Challenges and Opportunities in Quantum Machine Learningcited by 203
- 2021Supervised quantum machine learning models are kernel methodscited by 200
- 2020Threats to Federated Learning: A Surveycited by 178
- 2019Efficient Learning for Deep Quantum Neural Networkscited by 161
- 2024Barren Plateaus in Variational Quantum Computingcited by 133
- 2020InstaHide: Instance-hiding Schemes for Private Distributed Learningcited by 90
- 2020Quantum circuit architecture search for variational quantum algorithmscited by 82
- 2021Federated Quantum Machine Learningcited by 82
- 2020Quantum noise protects quantum classifiers against adversariescited by 68
- 2020Decentralizing Feature Extraction with Quantum Convolutional Neural Network for Automatic Speech Recognitioncited by 57
- 2021Privacy-preserving Federated Learning based on Multi-key Homomorphic Encryptioncited by 48
- 2021Opportunities of Federated Learning in Connected, Cooperative and Automated Industrial Systemscited by 45
- 2021Quantum federated learning through blind quantum computingcited by 44
- 2022Quantum Differential Privacy: An Information Theory Perspectivecited by 32
- 2021Quantum machine learning with differential privacycited by 31
- 2022Model Inversion Attacks against Graph Neural Networkscited by 26
- 2020Optimal Provable Robustness of Quantum Classification via Quantum Hypothesis Testingcited by 25
- 2022Enhancing Quantum Adversarial Robustness by Randomized Encodingscited by 17
- 2023Federated Quantum Machine Learning with Differential Privacycited by 17
- 2021Real-time End-to-End Federated Learning: An Automotive Case Studycited by 15
- 2023Privacy-preserving quantum federated learning via gradient hidingcited by 15
- 2023Reduction of finite sampling noise in quantum neural networkscited by 13
- 2024Quantum Mixed-State Self-Attention Networkcited by 11
- 2024Benchmarking quantum machine learning kernel training for classification taskscited by 10
- 2024Predominant Aspects on Security for Quantum Machine Learning: Literature Reviewcited by 9
- 2023Expressive variational quantum circuits provide inherent privacy in federated learningcited by 8
- 2020Minimal Model Structure Analysis for Input Reconstruction in Federated Learningcited by 5
- 2023Variational Quantum Neural Networks (VQNNS) in Image Classificationcited by 4
- 2024Bridging Quantum Computing and Differential Privacy: Insights into Quantum Computing Privacycited by 4
- 2024Prospects of Privacy Advantage in Quantum Machine Learningcited by 3
- 2025Certified randomness using a trapped-ion quantum processorcited by 3
Built on by
Citing papers per year: 2025: 1; 2026: 1.
2 papers.