Effect of barren plateaus on gradient-free optimization

Abstract

Barren plateau landscapes correspond to gradients that vanish exponentially in the number of qubits. Such landscapes have been demonstrated for variational quantum algorithms and quantum neural networks with either deep circuits or global cost functions. For obvious reasons, it is expected that gradient-based optimizers will be significantly affected by barren plateaus. However, whether or not gradient-free optimizers are impacted is a topic of debate, with some arguing that gradient-free approaches are unaffected by barren plateaus. Here we show that, indeed, gradient-free optimizers do not solve the barren plateau problem. Our main result proves that cost function differences, which are the basis for making decisions in a gradient-free optimization, are exponentially suppressed in a barren plateau. Hence, without exponential precision, gradient-free optimizers will not make progress in the optimization. We numerically confirm this by training in a barren plateau with several gradient-free optimizers (Nelder-Mead, Powell, and COBYLA algorithms), and show that the numbers of shots required in the optimization grows exponentially with the number of qubits.

arXiv comment12 pages, 3 figures, Updated to final publication version incorporating referee feedback

References in corpus (34)

• A variational eigenvalue solver on a quantum processor
• Variational Quantum Algorithms
• Barren plateaus in quantum neural network training landscapes
• Noisy intermediate-scale quantum (NISQ) algorithms
• Quantum Convolutional Neural Networks
• Cost Function Dependent Barren Plateaus in Shallow Parametrized Quantum Circuits
• The power of quantum neural networks
• Noise-Induced Barren Plateaus in Variational Quantum Algorithms
• Efficient Learning for Deep Quantum Neural Networks
• The quest for a Quantum Neural Network
• Connecting ansatz expressibility to gradient magnitudes and barren plateaus
• Quantum autoencoders for efficient compression of quantum data
• Quantum Natural Gradient
• Continuous-variable quantum neural networks
• An initialization strategy for addressing barren plateaus in parametrized quantum circuits
• Quantum-assisted quantum compiling
• Absence of Barren Plateaus in Quantum Convolutional Neural Networks
• Layerwise learning for quantum neural networks
• Random Quantum Circuits are Approximate 2-designs
• Local random quantum circuits are approximate polynomial-designs
• Variational Fast Forwarding for Quantum Simulation Beyond the Coherence Time
• Stochastic gradient descent for hybrid quantum-classical optimization
• Variational Quantum State Diagonalization
• Noise Resilience of Variational Quantum Compiling
• Equivalence of quantum barren plateaus to cost concentration and narrow gorges
• Entanglement Devised Barren Plateau Mitigation
• Barren plateaus preclude learning scramblers
• Higher Order Derivatives of Quantum Neural Networks with Barren Plateaus
• An Adaptive Optimizer for Measurement-Frugal Variational Algorithms
• On barren plateaus and cost function locality in variational quantum algorithms
• Variational Quantum Fidelity Estimation
• Variational Consistent Histories as a Hybrid Algorithm for Quantum Foundations
• Abrupt Transitions in Variational Quantum Circuit Training
• Natural Evolutionary Strategies for Variational Quantum Computation

Cited by in corpus (118)

• Variational Quantum Algorithms
• Cost Function Dependent Barren Plateaus in Shallow Parametrized Quantum Circuits
• Noise-Induced Barren Plateaus in Variational Quantum Algorithms
• Challenges and Opportunities in Quantum Machine Learning
• Generalization in quantum machine learning from few training data
• Beyond Barren Plateaus: Quantum Variational Algorithms Are Swamped With Traps
• Diagnosing Barren Plateaus with Tools from Quantum Optimal Control
• Barren Plateaus in Variational Quantum Computing
• Quantum Computing for High-Energy Physics: State of the Art and Challenges. Summary of the QC4HEP Working Group
• Equivalence of quantum barren plateaus to cost concentration and narrow gorges
• Hybrid quantum-classical algorithms in the noisy intermediate-scale quantum era and beyond
• Avoiding barren plateaus using classical shadows
• Quantum Krylov subspace algorithms for ground and excited state energy estimation
• Quantum Machine Learning: from physics to software engineering
• Theory for Equivariant Quantum Neural Networks
• A comprehensive review of Quantum Machine Learning: from NISQ to Fault Tolerance
• Understanding quantum machine learning also requires rethinking generalization
• Quantum computing of the ${}^6$Li nucleus via ordered unitary coupled clusters
• Subtleties in the trainability of quantum machine learning models
• The Presence and Absence of Barren Plateaus in Tensor-network Based Machine Learning
• Building spatial symmetries into parameterized quantum circuits for faster training
• Avoiding Barren Plateaus with Classical Deep Neural Networks
• Quantum approximate optimization via learning-based adaptive optimization
• Detecting and quantifying entanglement on near-term quantum devices
• Can Error Mitigation Improve Trainability of Noisy Variational Quantum Algorithms?
• Trainability barriers and opportunities in quantum generative modeling
• Mitigating Barren Plateaus with Transfer-learning-inspired Parameter Initializations
• Absence of barren plateaus in finite local-depth circuits with long-range entanglement
• Variational Quantum Anomaly Detection: Unsupervised mapping of phase diagrams on a physical quantum computer
• Investigating the effect of circuit cutting in QAOA for the MaxCut problem on NISQ devices
• Dynamical simulation via quantum machine learning with provable generalization
• Exponential data encoding for quantum supervised learning
• A semi-agnostic ansatz with variable structure for quantum machine learning
• On Physics-Informed Neural Networks for Quantum Computers
• Generalized Quantum Assisted Simulator
• Fock State-enhanced Expressivity of Quantum Machine Learning Models
• Quantum Capsule Networks
• Entanglement entropy production in Quantum Neural Networks
• On fundamental aspects of quantum extreme learning machines
• Retrieving information from a black hole using quantum machine learning
• Engineered dissipation to mitigate barren plateaus
• Representation of binary classification trees with binary features by quantum circuits
• Training variational quantum algorithms with random gate activation
• Reduction of finite sampling noise in quantum neural networks
• Efficient estimation of trainability for variational quantum circuits
• Classical Splitting of Parametrized Quantum Circuits
• A General Approach to Dropout in Quantum Neural Networks
• Using Differential Evolution to avoid local minima in Variational Quantum Algorithms
• The Unified Effect of Data Encoding, Ansatz Expressibility and Entanglement on the Trainability of HQNNs
• The quantum cost function concentration dependency on the parametrization expressivity
• Quantum Optimization for Training Quantum Neural Networks
• Solution of SAT Problems with the Adaptive-Bias Quantum Approximate Optimization Algorithm
• Quantum Advantage Seeker with Kernels (QuASK): a software framework to speed up the research in quantum machine learning
• Variational Quantum Cloning: Improving Practicality for Quantum Cryptanalysis
• Beyond unital noise in variational quantum algorithms: noise-induced barren plateaus and limit sets
• Evolution strategies: Application in hybrid quantum-classical neural networks
• Improving the speed of variational quantum algorithms for quantum error correction
• Beyond MP2 initialization for unitary coupled cluster quantum circuits
• The battle of clean and dirty qubits in the era of partial error correction
• Quantum variational learning for entanglement witnessing
• Generalization Error Bound for Quantum Machine Learning in NISQ Era -- A Survey
• F-Divergences and Cost Function Locality in Generative Modelling with Quantum Circuits
• Sequential optimal selection of a single-qubit gate and its relation to barren plateau in parameterized quantum circuits
• Emergence of noise-induced barren plateaus in arbitrary layered noise models
• Fast-forwarding quantum simulation with real-time quantum Krylov subspace algorithms
• Layering and subpool exploration for adaptive Variational Quantum Eigensolvers: Reducing circuit depth, runtime, and susceptibility to noise
• Laziness, Barren Plateau, and Noise in Machine Learning
• Resource frugal optimizer for quantum machine learning
• Image Classification with Rotation-Invariant Variational Quantum Circuits
• Equivalence of cost concentration and gradient vanishing for quantum circuits: An elementary proof in the Riemannian formulation
• Efficient Strategies for Reducing Sampling Error in Quantum Krylov Subspace Diagonalization
• Variational quantum computing for quantum simulation: principles, implementations, and challenges
• Exponential Qubit Reduction in Optimization for Financial Transaction Settlement
• Optimisation-free Classification and Density Estimation with Quantum Circuits
• Classification and reconstruction for single-pixel imaging with classical and quantum neural networks
• Quantum Computing for Data Centric Engineering and Science
• QSlack: A slack-variable approach for variational quantum semi-definite programming
• Enhancing Variational Quantum Circuit Training: An Improved Neural Network Approach for Barren Plateau Mitigation
• Image Classification by Throwing Quantum Kitchen Sinks at Tensor Networks
• Quantum Extremal Learning
• Calculation of the moscovium ground-state energy by quantum algorithms
• Restricting to the chip architecture maintains the quantum neural network accuracy
• Variational Quantum Simulation of Valence-Bond Solids
• Universal compiling and (No-)Free-Lunch theorems for continuous variable quantum learning
• Deep-Circuit QAOA
• Pitfalls when tackling the exponential concentration of parameterized quantum models
• Optimized numerical gradient and Hessian estimation for variational quantum algorithms
• A Monte Carlo Tree Search approach to QAOA: finding a needle in the haystack
• Data-Dependent Generalization Bounds for Parameterized Quantum Models Under Noise
• Learning to learn with an evolutionary strategy applied to variational quantum algorithms
• Variational quantum compiling for three-qubit gates design in quantum dots
• Exploring new variational quantum circuit ansatzes for solving $SU(2)$ matrix models
• The Lie Algebra of XY-mixer Topologies and Warm Starting QAOA for Constrained Optimization
• Generalized Probabilistic Approximate Optimization Algorithm
• An Efficient Gradient Sensitive Alternate Framework for VQE with Variable Ansatz
• Syndrome decoding by quantum approximate optimization
• Revisiting semiconductor bulk hamiltonians using quantum computers
• Barren plateaus are amplified by the dimension of qudits
• Photonic variational quantum eigensolver for NISQ-compatible quantum technology
• Policy Gradient Approach to Compilation of Variational Quantum Circuits
• Reinforcement learning to learn quantum states for Heisenberg scaling accuracy
• The Dual Role of Low-Weight Pauli Propagation: A Flawed Simulator but a Powerful Initializer for Variational Quantum Algorithms
• Robustness of optimized numerical estimation schemes for noisy variational quantum algorithms
• Batched Line Search Strategy for Navigating through Barren Plateaus in Quantum Circuit Training
• Exploiting many-body localization for scalable variational quantum simulation
• Variational simulation of higher-spin systems on qubit-based quantum simulators
• Scaling of symmetry-restricted quantum circuits
• A general approach to quantum integration of cross sections in high-energy physics
• A variational quantum algorithm for entanglement quantification
• Encoding strongly-correlated many-boson wavefunctions on a photonic quantum computer: application to the attractive Bose-Hubbard model
• Quantum circuit evolutionary framework applied on set partitioning problem
• Trainability issues in quantum policy gradients
• Equivalence between exponential concentration in quantum machine learning kernels and barren plateaus in variational algorithms
• Slice-Wise Initial State Optimization to Improve Cost and Accuracy of the VQE on Lattice Models
• Quantum approximate optimization of finite-state bosonic systems
• Improving the trainability of VQE on NISQ computers for solving portfolio optimization using convex interpolation
• Quantifying the Expressive Capacity of Quantum Systems: Fundamental Limits and Eigentasks
• Hamiltonian Expressibility for Ansatz Selection in Variational Quantum Algorithms