Abstract
Neural network-based variational Monte Carlo (NN-VMC) has emerged as a promising cutting-edge technique of ab initio quantum chemistry. However, the high computational cost of existing approaches hinders their applications in realistic chemistry problems. Here we report a development of NN-VMC that achieves a remarkable speed-up rate, thereby greatly extending the applicability of NN-VMC to larger systems. Our key design is a computational framework named Forward Laplacian, which computes the Laplacian associated with neural networks, the bottleneck of NN-VMC, through an efficient forward propagation process. We then demonstrate that Forward Laplacian can further facilitate more developments of acceleration methods across various aspects, including optimization for sparse derivative matrix and efficient network design. Empirically, our approach enables NN-VMC to investigate a broader range of systems, providing valuable references to other ab initio methods. The results demonstrate a great potential in applying deep learning methods to solve general quantum mechanical problems.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 digital issues and online access to articles
$119.00 per year
only $9.92 per issue
Buy this article
- Purchase on Springer Link
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
Data availability
All data supporting the findings of this study are available within the Supplementary Information.
Code availability
We release the source code of Forward Laplacian (https://github.com/YWolfeee/lapjax) and LapNet (https://github.com/bytedance/LapNet) on GitHub. Code for producing the results in this work is available on CodeOcean, together with all the system configuration studied in this work: https://codeocean.com/capsule/1139728/tree/v1 (ref. 60).
References
Needs, R. J., Towler, M. D., Drummond, N. D., López Ríos, P. & Trail, J. R. Variational and diffusion quantum Monte Carlo calculations with the CASINO code. J. Chem. Phys. 152, 154106 (2020).
Foulkes, W. M. C., Mitas, L., Needs, R. J. & Rajagopal, G. Quantum Monte Carlo simulations of solids. Rev. Modern Phys. 73, 33–83 (2001).
Pfau, D., Spencer, J. S., G. Matthews, A. G. & Foulkes, W. M. C. Ab-initio solution of the many-electron Schrödinger equation with deep neural networks. Phys. Rev. Res. 2, 033429 (2020).
Spencer, J. S., Pfau, D., Botev, A. & Foulkes, W. M. C. Better, faster fermionic neural networks. Preprint at https://arXiv.org/2011.07125 (2020).
Hermann, J., Schätzle, Z. & Noé, F. Deep-neural-network solution of the electronic Schrödinger equation. Nat. Chem. 12, 891–897 (2020).
Glehn, I., Spencer, J.S. & Pfau, D. A self-attention ansatz for ab-initio quantum chemistry. in 11th International Conference on Learning Representations https://openreview.net/forum?id=xveTeHVlF7j (2023.
Schätzle, Z., Szabó, P., Mezera, M., Hermann, J. & Noé, F. DeepQMC: an open-source software suite for variational optimization of deep-learning molecular wave functions. J. Chem. Phys. 159, 094108 (2023).
Bradbury, J. et al. JAX: composable transformations of Python+NumPy programs. GitHub http://github.com/google/jax (2023).
Drummond, N. D., Towler, M. D. & Needs, R. J. Jastrow correlation factor for atoms, molecules, and solids. Phys. Rev. B 70, 235119 (2004).
Brown, T. et al. Language models are few-shot learners. in Advances in Neural Information Processing Systems Vol. 33 (eds Larochelle, H. et al.) 1877–1901 (Curran Associates, 2020).
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Ren, W., Fu, W., Wu, X. & Chen, J. Towards the ground state of molecules via diffusion Monte Carlo on neural networks. Nat. Commun. 14, 1860 (2023).
Entwistle, M. T., Schätzle, Z., Erdman, P. A., Hermann, J. & Noé, F. Electronic excited states in deep variational Monte Carlo. Nat. Commun. 14, 274 (2023).
Sugar, J. & Corliss, C. Atomic energy levels of the iron-period elements: potassium through nickel. J. Phys. Chem. Ref. Data 14, 1 (1985).
Page, R. H. & Gudeman, C. S. Completing the iron period: double-resonance, fluorescence-dip rydberg spectroscopy and ionization potentials of titanium, vanadium, iron, cobalt, and nickel. J. Opt. Soc. Am. B 7, 1761–1771 (1990).
Sohl, J. E., Zhu, Y. & Knight, R. D. Two-color laser photoionization spectroscopy of Ti i: multichannel quantum defect theory analysis and a new ionization potential. J. Opt. Soc. Am. B 7, 9–14 (1990).
James, A. M. et al. Resonant two photon ionization spectroscopy of the molecules V2, VNb, and Nb2. J. Chem. Phys. 101, 4485–4495 (1994).
Spiekermann, K., Pattanaik, L. & Green, W. H. High accuracy barrier heights, enthalpies, and rate coefficients for chemical reactions. Sci. Data 9, 417 (2022).
Goerigk, L. et al. A look at the density functional theory zoo with the advanced GMTKN55 database for general main group thermochemistry, kinetics and noncovalent interactions. Phys. Chem. Chem. Phys. 19, 32184–32215 (2017).
Karton, A. & Martin, J. M. L. Explicitly correlated Wn theory: W1-F12 and W2-F12. J. Chem. Phys. 136, 124114 (2012).
Jurečka, P., Šponer, J., Černy’, J. & Hobza, P. Benchmark database of accurate (MP2 and CCSD(T) complete basis set limit) interaction energies of small model complexes, DNA base pairs, and amino acid pairs. Phys. Chem. Chem. Phys. 8, 1985–1993 (2006).
Han, J., Zhang, L. & Weinan, E. Solving many-electron Schrödinger equation using deep neural networks. J. Comput. Phys. 399, 108929 (2019).
Lin, J., Goldshlager, G. & Lin, L. Explicitly antisymmetrized neural network layers for variational Monte Carlo simulation. J. Comput. Phys. 474, 111765 (2023).
Abrahamsen, N. & Lin, L. Taming the sign problem of explicitly antisymmetrized neural networks via rough activation functions. Preprint at https://arXiv.org/2205.12250 (2022).
Gerard, L., Scherbela, M., Marquetand, P. & Grohs, P. Gold-standard solutions to the Schrödinger equation using deep learning: how much physics do we need? in 36th Conference on Neural Information Processing Systems https://openreview.net/forum?id=nX-gReQ0OT (2022).
Li, X., Li, Z. & Chen, J. Ab initio calculation of real solids via neural network ansatz. Nat. Commun. 13, 7895 (2022).
Pescia, G., Han, J., Lovato, A., Lu, J. & Carleo, G. Neural-network quantum states for periodic systems in continuous space. Phys. Rev. Res. 4, 023138 (2022).
Wilson, M. et al. Neural network ansatz for periodic wave functions and the homogeneous electron gas. Phys. Rev. B 107, 235139 (2023).
Cassella, G. et al. Discovering quantum phase transitions with fermionic neural networks. Phys. Rev. Lett. 130, 036401 (2023).
Pescia, G., Nys, J., Kim, J., Lovato, A. & Carleo, G. Message-passing neural quantum states for the homogeneous electron gas. Preprint at https://arXiv.org/2305.07240 (2023).
Qian, Y., Fu, W., Ren, W. & Chen, J. Interatomic force from neural network based variational quantum Monte Carlo. J. Chem. Phys. 157, 164104 (2022).
Li, X., Qian, Y. & Chen, J. Electric polarization from many-body neural network ansatz. Preprint at https://arXiv.org/2307.02212 (2023).
Li, X., Fan, C., Ren, W. & Chen, J. Fermionic neural network with effective core potential. Phys. Rev. Res. 4, 013021 (2022).
Wilson, M., Gao, N., Wudarski, F., Rieffel, E. & Tubman, N. M. Simulations of state-of-the-art fermionic neural network wave functions with diffusion Monte Carlo. Preprint at https://arXiv.org/2103.12570 (2021).
Scherbela, M., Reisenhofer, R., Gerard, L., Marquetand, P. & Grohs, P. Solving the electronic schrödinger equation for multiple nuclear geometries with weight-sharing deep neural networks. Nat. Comput. Sci. 2, 331–341 (2022).
Gao, N. & Günnemann, S. Ab-initio potential energy surfaces by pairing GNNs with neural wave functions. In 10th International Conference on Learning Representations https://openreview.net/forum?id=apv504XsysP (2022).
Gao, N. & Günnemann, S. Sampling-free inference for ab-initio potential energy surface networks. In 11th International Conference on Learning Representations https://openreview.net/forum?id=Tuk3Pqaizx (2023).
Gao, N. & Günnemann, S. Generalizing neural wave functions. in 40th International Conference on Machine Learning (eds Andreas K. et al.) (2023).
Scherbela, M., Gerard, L. & Grohs, P. Towards a transferable fermionic neural wavefunction for molecules. Nat. Commun. 15, 120 (2024).
Hermann, J., Schätzle, Z., Szabó, P. B., Mezera, M. & DeepQMC contributers. DeepQMC. GitHub https://github.com/deepqmc/deepqmc (2023).
James, S., Spencer, D. P. & FermiNet contributors. FermiNet. GitHub http://github.com/deepmind/ferminet (2023).
Carleo, G. & Troyer, M. Solving the quantum many-body problem with artificial neural networks. Science 355, 602–606 (2017).
Adams, C., Carleo, G., Lovato, A. & Rocco, N. Variational monte carlo calculations of A≤4 nuclei with an artificial neural-network correlator ansatz. Phys. Rev. Lett. 127, 022502 (2021).
Astrakhantsev, N. et al. Broken-symmetry ground states of the heisenberg model on the pyrochlore lattice. Phys. Rev. X 11, 041021 (2021).
Nomura, Y., Darmawan, A. S., Yamaji, Y. & Imada, M. Restricted Boltzmann machine learning for solving strongly correlated quantum systems. Phys. Rev. B 96, 205152 (2017).
Saito, H. Solving the Bose-Hubbard model with machine learning. J. Phys. Soc. Japan 86, 093001 (2017).
Choo, K., Mezzacapo, A. & Carleo, G. Fermionic neural-network states for ab-initio electronic structure. Nat. Commun. 11, 2368 (2020).
Yang, P.-J., Sugiyama, M., Tsuda, K. & Yanai, T. Artificial neural networks applied as molecular wave function solvers. J. Chem. Theory Comput. 16, 3513–3529 (2020).
Barrett, T. D., Malyshev, A. & Lvovsky, A. I. Autoregressive neural-network wavefunctions for ab initio quantum chemistry. Nat. Mach. Intell. 4, 351–358 (2022).
Yoshioka, N., Mizukami, W. & Nori, F. Solving quasiparticle band spectra of real solids using neural-network quantum states. Commun. Phys. 4, 106 (2021).
Wang, C., Li, S., He, D. & Wang, L. Is L2 physics informed loss always suitable for training physics informed neural network? in 36th Conference on Neural Information Processing Systems https://openreview.net/forum?id=cy1TKLRAEML (2022).
Raissi, M., Perdikaris, P. & Karniadakis, G. E. Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J. Comput. Phys. 378, 686–707 (2019).
He, D. et al. Learning physics-informed neural networks without stacked back-propagation. in Proc. 26th International Conference on Artificial Intelligence and Statistics (eds Francisco, R.) 3034–3047 (PMLR, 2023).
Mulliken, R. S. Electronic population analysis on LCAO–MO molecular wave functions. i. J. Chem. Phys. 23, 1833–1840 (1955).
Paszke, A. et al. PyTorch: An imperative style, high-performance deep learning library. in Advances in Neural Information Processing Systems Vol. 32 (eds Wallach, H. et al.) 8024–8035 (Curran Associates, 2019).
Abadi, M. et al. TensorFlow, large-scale machine learning on heterogeneous systems. in Proc. 12th USENIX Symposium on Operating Systems Design and Implementation (eds Kimberly, K. & Timothy, R.) 265–283 (USENIX, 2016).
Botev, A. The Gauss-Newton Matrix for Deep Learning Models and Its Applications. PhD thesis (Univ. College London, 2020).
Martens, J., Sutskever, I. & Swersky, K. Estimating the Hessian by back-propagating curvature. in 29th International Conference on Machine Learning (2012).
Brown, M. D., Trail, J. R., López Ríos, P. & Needs, R. J. Energies of the first row atoms from quantum Monte Carlo. J. Chem. Phys. 126, 224110 (2007).
Li, R. et al. Supplementary code of ‘A new computational framework for neural network-based variational Monte Carlo with Forward Laplacian’. Code Ocean https://codeocean.com/capsule/1139728/tree/v1 (2023).
Balabanov, N. B. & Peterson, K. A. Basis set limit electronic excitation energies, ionization potentials, and electron affinities for the 3d transition metal atoms: Coupled cluster and multireference methods. J. Chem. Phys. 125, 074110 (2006).
Marshall, M. S., Burns, L. A. & Sherrill, C. D. Basis set convergence of the coupled-cluster correction, \({\delta }_{{{{\rm{MP2}}}}}^{{{{\rm{CCSD(T)}}}}}\): best practices for benchmarking non-covalent interactions and the attendant revision of the S22, NBC10, HBC6, and HSG databases. J. Chem. Phys. https://doi.org/10.1063/1.3659142 (2011).
Acknowledgements
We thank H. Li and ByteDance Research for support and inspiration. We thank D. Pfau for his valuable feedback. We thank B. Zhang and H. Wang for their helpful suggestions and discussion. L.W. is supported by the National Key R&D Program of China (grant no. 2022ZD0114900) and the National Science Foundation of China (grant no. NSFC62276005). J.C. acknowledges the National Natural Science Foundation of China for support under grant no. 92165101.
Author information
Authors and Affiliations
Contributions
R.L. and L.W. conceived the study. H.Y. implemented the main code with contributions from R.L., D.J. and Z.L. X.W., J.C. and W.R. suggested the experiments. R.L., D.J. and H.Y. performed the simulations and analyzed the results. X.W. and X.L. performed the chemical analyses. C.W. and R.L. developed the theoretical formalism. H.Y., R.L., D.H., W.R., D.J. and J.C. performed the figure designing. L.W. and W.R. organized the project. D.H., R.L., H.Y., W.R., J.C., D.J., C.W. and X.W. wrote the paper.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Machine Intelligence thanks the anonymous reviewer(s) for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Supplementary Information
Supplementary Figs. 1 and 2, Notes 1–9 and Tables 1–11.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Li, R., Ye, H., Jiang, D. et al. A computational framework for neural network-based variational Monte Carlo with Forward Laplacian. Nat Mach Intell 6, 209–219 (2024). https://doi.org/10.1038/s42256-024-00794-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s42256-024-00794-x
This article is cited by
-
A computational framework for neural network-based variational Monte Carlo with Forward Laplacian
Nature Machine Intelligence (2024)