Cautious optimism for deep parameterized quantum circuits
Marie Kempkes, Elies Gil-Fuster, Carlos Bravo-Prieto, Aroosa Ijaz, Alissa Wilms, Jens Eisert, Evert van Nieuwenburg, Vedran Dunjko
quant-ph, cs.LG, stat.ML
Submitted: 2026-07-23
Comments: 21 pages (6+15), 2 figures (1+1), comments welcome
License: http://creativecommons.org/licenses/by/4.0/
The gist: A central challenge in quantum machine learning is understanding the scaling behavior of parameterized quantum circuits (PQCs).
Terminology
Abstract
A central challenge in quantum machine learning is understanding the scaling behavior of parameterized quantum circuits (PQCs). In particular, it remains unclear how their performance on unseen data changes as the number of trainable parameters increases. Prior works have derived formal generalization guarantees for quantum models, but it is well-known that many such results do not fully characterize generalization behavior in practice. In this work, we show that gradient-based PQCs can exhibit improved performance on unseen data as model size increases, displaying the phenomenon of double descent. This contrasts with the traditional view that larger models lead to degraded generalization. We provide analytical results rigorously underpinning this behavior by leveraging add-one-in perturbation techniques and spectral properties of random matrices. We support these results with numerical experiments on re-uploading PQCs across several data sets and training set sizes, consistently observing the predicted double descent behavior. While other obstacles on the path toward practical quantum machine learning remain, our finding that deeper parameterized quantum circuits do not necessarily exhibit degraded performance provides reasons for cautious optimism.
Sources
- Quantum machine learning advantages beyond hardness of evaluation
- Prospects for quantum advantage in machine learning from the representability of functions
- Quantum Advantage in Learning Quantum Dynamics via Fourier coefficient extraction
- Provable learning separation for predicting time-evolution of quantum many-body systems
- Mind the gaps: The fraught road to quantum advantage
- A PAC-Bayesian approach to generalization for quantum models
- Grokking and epoch-wise double descent in quantum neural networks
- Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
- Spectral Born machines: classically trainable quantum generative models for discrete data
- Train on classical, deploy on quantum: scaling generative quantum machine learning to a thousand qubits
- Universality and kernel-adaptive training for classically trained, quantum-deployed generative models
- Qudit extension of parameterized IQP circuits: A generative quantum machine learning approach to integer data
- A Farewell to the Bias-Variance Tradeoff? An Overview of the Theory of Overparameterized Machine Learning
- Concentration of Measure and Large Random Matrices with an application to Sample Covariance Matrices
- Mildly Overparameterized ReLU Networks Have a Favorable Loss Landscape
- PennyLane: Automatic differentiation of hybrid quantum-classical computations
Related papers
- Reconquering Bell sampling on qudits: stabilizer learning and testing, quantum pseudorandomness bounds, and more
- Encrypted clones can leak: Classification of informative subsets in Quantum Encrypted Cloning
- Polynomial-time classical and quantum simulation of quantum impurity models
- A quasiparticle-protected superconducting qubit
- Theory of quantum-enhanced interferometry with general Markovian light sources
- A convergent hierarchy of spectral gap certificates for qubit Hamiltonians