Barren plateaus in quantum neural network training landscapes
Explore this paper's citation graph
Summary
It is shown that for a wide class of reasonable parameterized quantum circuits, the probability that the gradient along any reasonable direction is non-zero to some fixed precision is exponentially small as a function of the number of qubits.
- Type
- article
- Published
- 2018-03-29
- Cited by
- 2,956
- References
- 56
- Access
- Open access
- OpenAlex
- https://openalex.org/W2794444783
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:4465524
Keywords
Computer science, Qubit, Parameterized complexity, Quantum circuit, Quantum algorithm
References
- Learning in modular systems
- Quantum implementation of the unitary coupled cluster for simulating molecular electronic structure
- Gradient Flow in Recurrent Nets: the Difficulty of Learning Long-Term Dependencies
- Understanding the difficulty of training deep feedforward neural networks
- A Quantum Approximate Optimization Algorithm
- The concentration of measure phenomenon
- Products of independent Gaussian random matrices
- Exact and approximate unitary 2-designs and their application to fidelity estimation
- Most quantum States are too entangled to be useful as computational resources.
- Are random pure States useful for quantum computation?
- Optimal quantum measurements of expectation values of observables
- Random Quantum Circuits are Approximate 2-designs
- Corrigendum: RecG and UvsW catalyse robust DNA rewinding critical for stalled DNA replication fork rescue
- Symmetric informationally complete quantum measurements
- The foundations of statistical mechanics from entanglement: Individual states vs. averages
- Exploiting Locality in Quantum Computation for Quantum Chemistry.
- A Fast Learning Algorithm for Deep Belief Nets
- From transistor to trapped-ion computers for quantum chemistry
- A variational eigenvalue solver on a photonic quantum processor
- Deep Residual Learning for Image Recognition
Cited by
- From the Quantum Approximate Optimization Algorithm to a Quantum Alternating Operator Ansatz
- Quantum Chemistry Calculations on a Trapped-Ion Quantum Simulator
- Towards quantum machine learning with tensor networks
- Encoding Electronic Spectra in Quantum Circuits with Linear T Complexity
- Universal discriminative quantum neural networks
- Classification of the MNIST data set with quantum slow feature analysis
- Adversarial quantum circuit learning for pure state approximation
- Quantum-assisted quantum compiling
- Error-Mitigated Digital Quantum Simulation.
- Quantum computational chemistry
- Approximate Unitary t-Designs by Short Random Quantum Circuits Using Nearest-Neighbor and Long-Range Gates
- Generalized Unitary Coupled Cluster Wave functions for Quantum Computation.
- Variational quantum state diagonalization
- A quantum alternating operator ansatz with hard and soft constraints for lattice protein folding
- Finding the ground state of the Hubbard model by variational methods on a quantum computer with gate errors
- Quantum Approximate Optimization Algorithm: Performance, Mechanism, and Implementation on Near-Term Devices
- Potential of quantum computing for drug discovery
- Quantum Chemistry in the Age of Quantum Computing.
- Variational Quantum Generators: Generative Adversarial Quantum Machine Learning for Continuous Distributions
- Quantum Computation of Electronic Transitions Using a Variational Quantum Eigensolver.
Related papers
- Performance Comparison of Quantum Phase Estimation Algorithm with Different Number of Register Qubits on Noisy Quantum Processor
- Decomposition Algorithm of an Arbitrary Pauli Exponential Through a Quantum Circuit
- Grover algorithm circuit optimization and noise analysis on integer factorization
- Some Quantum States Prepared with Polynomial Quantum Circuit
- Universal quantum circuit for N-qubit quantum gate: a programmable quantum gate
- Demonstrating quantum algorithm acceleration with NMR quantum computer