Quantitative convergence of trained quantum neural networks to a Gaussian process

Abstract

We study quantum neural networks where the generated function is the expectation value of the sum of single-qubit observables across all qubits. In [Girardi \emph{et al.}, arXiv:2402.08726], it is proven that the probability distributions of such generated functions converge in distribution to a Gaussian process in the limit of infinite width for both untrained networks with randomly initialized parameters and trained networks. In this paper, we provide a quantitative proof of this convergence in terms of the Wasserstein distance of order $1$. First, we establish an upper bound on the distance between the probability distribution of the function generated by any untrained network with finite width and the Gaussian process with the same covariance. This proof utilizes Stein's method to estimate the Wasserstein distance of order $1$. Next, we analyze the training dynamics of the network via gradient flow, proving an upper bound on the distance between the probability distribution of the function generated by the trained network and the corresponding Gaussian process. This proof is based on a quantitative upper bound on the maximum variation of a parameter during training. This bound implies that for sufficiently large widths, training occurs in the lazy regime, \emph{i.e.}, each parameter changes only by a small amount. While the convergence result of [Girardi \emph{et al.}, arXiv:2402.08726] holds at a fixed training time, our upper bounds are uniform in time and hold even as $t\to \infty$.

References in corpus (21)

• Quantum Machine Learning
• Supervised learning with quantum enhanced feature spaces
• Barren plateaus in quantum neural network training landscapes
• An introduction to quantum machine learning
• Cost Function Dependent Barren Plateaus in Shallow Parametrized Quantum Circuits
• The effect of data encoding on the expressive power of variational quantum machine learning models
• A rigorous and robust quantum speed-up in supervised machine learning
• Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient Descent
• Wasserstein Distance based Deep Adversarial Transfer Learning for Intelligent Fault Diagnosis
• The Principles of Deep Learning Theory
• Barren Plateaus in Variational Quantum Computing
• A Lie Algebraic Theory of Barren Plateaus for Deep Parameterized Quantum Circuits
• Barren plateaus in quantum tensor network optimization
• Representation Learning via Quantum Neural Tangent Kernels
• Hamiltonian variational ansatz without barren plateaus
• Learning quantum data with the quantum Earth Mover's distance
• Quantum optimal transport with quantum channels
• Absence of barren plateaus in finite local-depth circuits with long-range entanglement
• A Hyperparameter Study for Quantum Kernel Methods
• Quantum-enhanced neural networks in the neural tangent kernel framework
• Trained quantum neural networks are Gaussian processes

Cited by in corpus (2)

• Efficient classical computation of the neural tangent kernel of quantum neural networks
• Mean-field limit from general mixtures of experts to quantum neural networks