On barren plateaus and cost function locality in variational quantum algorithms

Variational quantum algorithms rely on gradient based optimization to iteratively minimize a cost function evaluated by measuring output(s) of a quantum processor. A barren plateau is the phenomenon of exponentially vanishing gradients in sufficiently expressive parametrized quantum circuits. It has been established that the onset of a barren plateau regime depends on the cost function, although the particular behavior has been demonstrated only for certain classes of cost functions. Here we derive a lower bound on the variance of the gradient, which depends mainly on the width of the circuit causal cone of each term in the Pauli decomposition of the cost function. Our result further clarifies the conditions under which barren plateaus can occur.

[1]  Ryan Babbush,et al.  The theory of variational hybrid quantum-classical algorithms , 2015, 1509.04279.

[2]  Alán Aspuru-Guzik,et al.  A variational eigenvalue solver on a photonic quantum processor , 2013, Nature Communications.

[3]  Ryan Babbush,et al.  Barren plateaus in quantum neural network training landscapes , 2018, Nature Communications.

[4]  Alán Aspuru-Guzik,et al.  The theory of variational hybrid quantum-classical algorithms , 2015, 1509.04279.

[5]  K. Birgitta Whaley,et al.  Towards quantum machine learning with tensor networks , 2018, Quantum Science and Technology.

[6]  A. Harrow,et al.  Approximate Unitary t-Designs by Short Random Quantum Circuits Using Nearest-Neighbor and Long-Range Gates , 2018, Communications in Mathematical Physics.

[7]  Kristan Temme,et al.  Supervised learning with quantum-enhanced feature spaces , 2018, Nature.

[8]  J. McClean,et al.  Strategies for quantum computing molecular energies using the unitary coupled cluster ansatz , 2017, Quantum Science and Technology.

[9]  Masoud Mohseni,et al.  Layerwise learning for quantum neural networks , 2020, Quantum Machine Intelligence.

[10]  Matthew P. A. Fisher,et al.  Measurement-driven entanglement transition in hybrid quantum circuits , 2019, Physical Review B.

[11]  E. Farhi,et al.  A Quantum Approximate Optimization Algorithm , 2014, 1411.4028.

[12]  C. Gogolin,et al.  Evaluating analytic gradients on quantum hardware , 2018, Physical Review A.

[13]  Andrew G. Taube,et al.  New perspectives on unitary coupled‐cluster theory , 2006 .

[14]  Ivano Tavernelli,et al.  Quantum algorithms for electronic structure calculations: Particle-hole Hamiltonian and optimized wave-function expansions , 2018, Physical Review A.

[15]  V. Akshay,et al.  Reachability Deficits in Quantum Approximate Optimization , 2019, Physical review letters.

[16]  Kunal Sharma,et al.  Trainability of Dissipative Perceptron-Based Quantum Neural Networks , 2020, ArXiv.

[17]  Ryan LaRose,et al.  Quantum-assisted quantum compiling , 2018, Quantum.

[18]  M. Hastings,et al.  Progress towards practical quantum variational algorithms , 2015, 1507.08969.

[19]  Tobias J. Osborne,et al.  No Free Lunch for Quantum Machine Learning , 2020, 2003.14103.

[20]  F. Jin,et al.  Benchmarking the quantum approximate optimization algorithm , 2019, Quantum Inf. Process..

[21]  Andrey Kardashin,et al.  Machine Learning Phase Transitions with a Quantum Processor , 2019, Physical Review A.

[22]  Johannes Jakob Meyer,et al.  Stochastic gradient descent for hybrid quantum-classical optimization , 2019, Quantum.

[23]  B. Collins,et al.  Integration with Respect to the Haar Measure on Unitary, Orthogonal and Symplectic Group , 2004, math-ph/0402073.

[24]  F. Brandão,et al.  Local random quantum circuits are approximate polynomial-designs: numerical results , 2012, 1208.0692.

[25]  Alán Aspuru-Guzik,et al.  Quantum autoencoders for efficient compression of quantum data , 2016, 1612.02806.

[26]  M. Cerezo,et al.  A semi-agnostic ansatz with variable structure for quantum machine learning , 2021, arXiv.org.

[27]  Akira Sone,et al.  Cost-Function-Dependent Barren Plateaus in Shallow Quantum Neural Networks , 2020, ArXiv.

[28]  A. Garcia-Saez,et al.  Addressing hard classical problems with Adiabatically Assisted Variational Quantum Eigensolvers , 2018, 1806.02287.

[29]  J. Gambetta,et al.  Hardware-efficient variational quantum eigensolver for small molecules and quantum magnets , 2017, Nature.

[30]  Keisuke Fujii,et al.  Quantum circuit learning , 2018, Physical Review A.

[31]  P. Zoller,et al.  Self-verifying variational quantum simulation of lattice models , 2018, Nature.

[32]  J. Biamonte Universal variational quantum computation , 2019, Physical Review A.

[33]  Naoki Yamamoto,et al.  Expressibility of the alternating layered ansatz for quantum computation. , 2020 .

[34]  Richard Andrew Low,et al.  Pseudo-randonmess and Learning in Quantum Computation , 2010, 1006.5227.

[35]  R. Brockett,et al.  Time optimal control in spin systems , 2000, quant-ph/0006114.

[36]  Harper R. Grimsley,et al.  An adaptive variational algorithm for exact molecular simulations on a quantum computer , 2018, Nature Communications.

[37]  M. Schuld,et al.  Circuit-centric quantum classifiers , 2018, Physical Review A.

[38]  Seth Lloyd,et al.  Convergence conditions for random quantum circuits , 2005, quant-ph/0503210.

[39]  John Watrous,et al.  The Theory of Quantum Information , 2018 .

[40]  Artur F. Izmaylov,et al.  Iterative Qubit Coupled Cluster approach with efficient screening of generators. , 2019, Journal of chemical theory and computation.

[41]  Edward Grant,et al.  An initialization strategy for addressing barren plateaus in parametrized quantum circuits , 2019, Quantum.

[42]  Vladyslav Verteletskyi,et al.  Measurement optimization in the variational quantum eigensolver using a minimum clique cover. , 2019, The Journal of chemical physics.

[43]  Stuart Samuel,et al.  U(N) Integrals, 1/N, and the De Wit–’t Hooft anomalies , 1980 .

[44]  John Napp,et al.  Low-Depth Gradient Measurements Can Improve Convergence in Variational Hybrid Quantum-Classical Algorithms. , 2019, Physical review letters.

[45]  Jacob Biamonte,et al.  Variational learning of Grover's quantum search algorithm , 2018, Physical Review A.

[46]  S. Lloyd Quantum approximate optimization is computationally universal , 2018, 1812.11075.

[47]  Navin Khaneja,et al.  Cartan Decomposition of SU (2 n ) , Constructive Controllability of Spin Systems and Universal Quantum Computing , 2001 .