Improved Optimization for the Neural-network Quantum States and Tests on the Chromium Dimer (2404.09280v3)
Abstract: The advent of Neural-network Quantum States (NQS) has significantly advanced wave function ansatz research, sparking a resurgence in orbital space variational Monte Carlo (VMC) exploration. This work introduces three algorithmic enhancements to reduce computational demands of VMC optimization using NQS: an adaptive learning rate algorithm, constrained optimization, and block optimization. We evaluate the refined algorithm on complex multireference bond stretches of $\rm H_2O$ and $\rm N_2$ within the cc-pVDZ basis set and calculate the ground-state energy of the strongly correlated chromium dimer ($\rm Cr_2$) in the Ahlrichs SV basis set. Our results achieve superior accuracy compared to coupled cluster theory at a relatively modest CPU cost. This work demonstrates how to enhance optimization efficiency and robustness using these strategies, opening a new path to optimize large-scale Restricted Boltzmann Machine (RBM)-based NQS more effectively and marking a substantial advancement in NQS's practical quantum chemistry applications.
- G. D. Purvis and R. J. Bartlett, J. Chem. Phys. 76, 1910–1918 (1982).
- R. F. Fink, Chem. Phys. 356, 39–46 (2009).
- W. Liu and M. R. Hoffmann, Theor. Chem. Acc. 133, 1481 (2014).
- F. Chen and Z. Fan, J. Comput. Chem. 35, 121–129 (2014).
- H.-J. Werner and P. J. Knowles, J. Chem. Phys. 89, 5803–5814 (1988).
- F. A. Evangelista, J. Chem. Phys. 149, 030901 (2018).
- U. Schollwöck, Ann. Phys. 326, 96–192 (2011).
- A. Baiardi and M. Reiher, J. Chem. Phys. 152, 040903 (2020).
- S. D. S. George H. Booth and A. Alavi, Mol. Phys. 112, 1855–1869 (2014).
- W. Liu and M. R. Hoffmann, J. Chem. Theory Comput. 12, 1169–1178 (2016).
- K. R. Brorsen, J. Chem. Theory Comput. 16, 2379–2388 (2020).
- Y. Yao and C. J. Umrigar, J. Chem. Theory Comput. 17, 4183–4194 (2021).
- G. Carleo and M. Troyer, Science 355, 602–606 (2017).
- A.-J. Liu and B. K. Clark, “Neural network backflow for ab-initio quantum chemistry,” arXiv:2403.03286 (2024).
- I. von Glehn, J. S. Spencer, and D. Pfau, “A self-attention ansatz for ab-initio quantum chemistry,” arXiv:2211.13672 (2022).
- H. Wei and E. Neuscamman, J. Chem. Phys. 149, 184106 (2018).
- I. Sabzevari and S. Sharma, J. Chem. Theory Comput. 14, 6276–6286 (2018).
- S. Sorella, Phys. Rev. B 64, 024512 (2001).
- A. Mahajan and S. Sharma, J. Chem. Theory Comput. 123, 3911–3921 (2019).
- D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv:1412.6980 (2014).
- L. Otis and E. Neuscamman, Phys. Chem. Chem. Phys. 21, 14491–14510 (2019).
- J. Toulouse and C. J. Umrigar, J. Chem. Phys. 126, 084102 (2007).
- J. Martens and R. Grosse, in Proceedings of the 32nd International Conference on International Conference on Machine Learning - Volume 37, ICML’15 (JMLR.org, 2015) p. 2408–2417.
- J. P. Coe, J. Chem. Theory Comput. 15, 6179–6189 (2019).
- Y. Kurashige and T. Yanai, J. Chem. Phys. 130, 234114 (2009).
- H. Shang, C. Guo, Y. Wu, Z. Li, and J. Yang, “Solving schrödinger equation with a language model,” arXiv:2307.09343 (2023).
- S. Sharma, “Stochastic perturbation theory to correct non-linearly parametrized wavefunctions,” arXiv:1803.04341 (2018).
- V. Kvasnička, Chem. Phys. Lett. 43, 377–381 (1976).