Equivalence of quantum barren plateaus to cost concentration and narrow gorges

被引:73
作者
Arrasmith, Andrew [1 ]
Holmes, Zoe [2 ]
Cerezo, M. [1 ,3 ]
Coles, Patrick J. [1 ]
机构
[1] Los Alamos Natl Lab, Theoret Div, MS B213, Los Alamos, NM 87545 USA
[2] Los Alamos Natl Lab, Informat Sci, Los Alamos, NM 87544 USA
[3] Los Alamos Natl Lab, Ctr Nonlinear Studies, Los Alamos, NM USA
关键词
barren plateau; quantum computing; quantum machine learning; narrow gorge; variational algorithm;
D O I
10.1088/2058-9565/ac7d06
中图分类号
O4 [物理学];
学科分类号
0702 ;
摘要
Optimizing parameterized quantum circuits (PQCs) is the leading approach to make use of near-term quantum computers. However, very little is known about the cost function landscape for PQCs, which hinders progress towards quantum-aware optimizers. In this work, we investigate the connection between three different landscape features that have been observed for PQCs: (1) exponentially vanishing gradients (called barren plateaus (BPs)), (2) exponential cost concentration about the mean, and (3) the exponential narrowness of minima (called narrow gorges). We analytically prove that these three phenomena occur together, i.e., when one occurs then so do the other two. A key implication of this result is that one can numerically diagnose BPs via cost differences rather than via the computationally more expensive gradients. More broadly, our work shows that quantum mechanics rules out certain cost landscapes (which otherwise would be mathematically possible), and hence our results could be interesting from a quantum foundations perspective.
引用
收藏
页数:14
相关论文
共 66 条
[1]  
Low RA, 2010, Arxiv, DOI [arXiv:1006.5227, 10.48550/arXiv.1006.5227]
[2]   The power of quantum neural networks [J].
Abbas, Amira ;
Sutter, David ;
Zoufal, Christa ;
Lucchi, Aurelien ;
Figalli, Alessio ;
Woerner, Stefan .
NATURE COMPUTATIONAL SCIENCE, 2021, 1 (06) :403-409
[3]  
Arrasmith A, 2020, Arxiv, DOI arXiv:2004.06252
[4]   Effect of barren plateaus on gradient-free optimization [J].
Arrasmith, Andrew ;
Cerezo, M. ;
Czarnik, Piotr ;
Cincio, Lukasz ;
Coles, Patrick J. .
QUANTUM, 2021, 5
[5]   Variational consistent histories as a hybrid algorithm for quantum foundations [J].
Arrasmith, Andrew ;
Cincio, Lukasz ;
Sornborger, Andrew T. ;
Zurek, Wojciech H. ;
Coles, Patrick J. .
NATURE COMMUNICATIONS, 2019, 10 (1)
[6]   Training deep quantum neural networks [J].
Beer, Kerstin ;
Bondarenko, Dmytro ;
Farrelly, Terry ;
Osborne, Tobias J. ;
Salzmann, Robert ;
Scheiermann, Daniel ;
Wolf, Ramona .
NATURE COMMUNICATIONS, 2020, 11 (01)
[7]   Noisy intermediate-scale quantum algorithms [J].
Bharti, Kishor ;
Cervera-Lierta, Alba ;
Kyaw, Thi Ha ;
Haug, Tobias ;
Alperin-Lea, Sumner ;
Anand, Abhinav ;
Degroote, Matthias ;
Heimonen, Hermanni ;
Kottmann, Jakob S. ;
Menke, Tim ;
Mok, Wai-Keong ;
Sim, Sukin ;
Kwek, Leong-Chuan ;
Aspuru-Guzik, Alan .
REVIEWS OF MODERN PHYSICS, 2022, 94 (01)
[8]   Quantum-assisted simulator [J].
Bharti, Kishor ;
Haug, Tobias .
PHYSICAL REVIEW A, 2021, 104 (04)
[9]   Quantum machine learning [J].
Biamonte, Jacob ;
Wittek, Peter ;
Pancotti, Nicola ;
Rebentrost, Patrick ;
Wiebe, Nathan ;
Lloyd, Seth .
NATURE, 2017, 549 (7671) :195-202
[10]   Local Random Quantum Circuits are Approximate Polynomial-Designs [J].
Brandao, Fernando G. S. L. ;
Harrow, Aram W. ;
Horodecki, Michal .
COMMUNICATIONS IN MATHEMATICAL PHYSICS, 2016, 346 (02) :397-434