Avoiding barren plateaus using classical shadows

Abstract

Variational quantum algorithms are promising algorithms for achieving quantum advantage on near-term devices. The quantum hardware is used to implement a variational wave function and measure observables, whereas the classical computer is used to store and update the variational parameters. The optimization landscape of expressive variational ansätze is however dominated by large regions in parameter space, known as barren plateaus, with vanishing gradients which prevents efficient optimization. In this work we propose a general algorithm to avoid barren plateaus in the initialization and throughout the optimization. To this end we define a notion of weak barren plateaus (WBP) based on the entropies of local reduced density matrices. The presence of WBPs can be efficiently quantified using recently introduced shadow tomography of the quantum state with a classical computer. We demonstrate that avoidance of WBPs suffices to ensure sizable gradients in the initialization. In addition, we demonstrate that decreasing the gradient step size, guided by the entropies allows to avoid WBPs during the optimization process. This paves the way for efficient barren plateau free optimization on near-term devices.

9 pages, 4 figures, comments are welcome; v2: improved readability, added new section and data, added more citations

References in corpus (20)

• The density-matrix renormalization group in the age of matrix product states
• Black holes as mirrors: quantum information in random subsystems
• Direct Fidelity Estimation from Few Pauli Measurements
• Training variational quantum algorithms is NP-hard
• Evenly distributed unitaries: on the structure of unitary designs
• Effect of barren plateaus on gradient-free optimization
• Real- and imaginary-time evolution with compressed quantum circuits
• Symmetry-resolved entanglement detection using partial transpose moments
• Efficient estimation of Pauli observables by derandomization
• Quantum annealing initialization of the quantum approximate optimization algorithm
• Entanglement Devised Barren Plateau Mitigation
• Robust shadow estimation
• Capacity and quantum geometry of parametrized quantum circuits
• Fisher Information in Noisy Intermediate-Scale Quantum Applications
• An efficient quantum algorithm for the time evolution of parameterized circuits
• Classical simulation versus universality in measurement based quantum computation
• Simultaneous Perturbation Stochastic Approximation of the Quantum Fisher Information
• Quantum Fisher information from randomized measurements
• Emergence of typical entanglement in two-party random processes
• Sharp Continuity Bounds for Entropy and Conditional Entropy

Cited by in corpus (29)

• The Variational Quantum Eigensolver: a review of methods and best practices
• Classical Shadows With Noise
• Near-Term Quantum Computing Techniques: Variational Quantum Algorithms, Error Mitigation, Circuit Compilation, Benchmarking and Classical Simulation
• Theoretical Guarantees for Permutation-Equivariant Quantum Neural Networks
• Barren plateaus in quantum tensor network optimization
• Avoiding barren plateaus via transferability of smooth solutions in Hamiltonian Variational Ansatz
• Scalable and Flexible Classical Shadow Tomography with Tensor Networks
• Shallow shadows: Expectation estimation using low-depth random Clifford circuits
• High Dimensional Quantum Machine Learning With Small Quantum Computers
• A semi-agnostic ansatz with variable structure for quantum machine learning
• Variational thermal quantum simulation of the lattice Schwinger model
• An Alternative Approach to Quantum Imaginary Time Evolution
• Entanglement entropy production in Quantum Neural Networks
• A Performance Study of Variational Quantum Algorithms for Solving the Poisson Equation on a Quantum Computer
• Performance analysis of multi-shot shadow estimation
• Quantum variational learning for quantum error-correcting codes
• Classical Splitting of Parametrized Quantum Circuits
• Quantum Quantile Mechanics: Solving Stochastic Differential Equations for Generating Time-Series
• Solution of SAT Problems with the Adaptive-Bias Quantum Approximate Optimization Algorithm
• Quantum-assisted Monte Carlo algorithms for fermions
• Coarse grained intermolecular interactions on quantum processors
• Bayesian Learning of Parameterised Quantum Circuits
• Quantum-enhanced neural networks in the neural tangent kernel framework
• Measuring energy by measuring any other observable
• Optimisation-free Classification and Density Estimation with Quantum Circuits
• Twisted hybrid algorithms for combinatorial optimization
• Perturbative gadgets for gate-based quantum computing: Non-recursive constructions without subspace restrictions
• Quantum Circuit Cutting for Classical Shadows
• Symmetry-based quantum algorithms for open-shop scheduling with hard constraints