Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
GPT-4o
Gemini 2.5 Pro Pro
o3 Pro
GPT-4.1 Pro
DeepSeek R1 via Azure Pro
2000 character limit reached

Improved Optimization for the Neural-network Quantum States and Tests on the Chromium Dimer (2404.09280v3)

Published 14 Apr 2024 in physics.chem-ph and quant-ph

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.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (28)
  1. G. D. Purvis and R. J. Bartlett, J. Chem. Phys. 76, 1910–1918 (1982).
  2. R. F. Fink, Chem. Phys. 356, 39–46 (2009).
  3. W. Liu and M. R. Hoffmann, Theor. Chem. Acc. 133, 1481 (2014).
  4. F. Chen and Z. Fan, J. Comput. Chem. 35, 121–129 (2014).
  5. H.-J. Werner and P. J. Knowles, J. Chem. Phys. 89, 5803–5814 (1988).
  6. F. A. Evangelista, J. Chem. Phys. 149, 030901 (2018).
  7. U. Schollwöck, Ann. Phys. 326, 96–192 (2011).
  8. A. Baiardi and M. Reiher, J. Chem. Phys. 152, 040903 (2020).
  9. S. D. S. George H. Booth and A. Alavi, Mol. Phys. 112, 1855–1869 (2014).
  10. W. Liu and M. R. Hoffmann, J. Chem. Theory Comput. 12, 1169–1178 (2016).
  11. K. R. Brorsen, J. Chem. Theory Comput. 16, 2379–2388 (2020).
  12. Y. Yao and C. J. Umrigar, J. Chem. Theory Comput. 17, 4183–4194 (2021).
  13. G. Carleo and M. Troyer, Science 355, 602–606 (2017).
  14. A.-J. Liu and B. K. Clark, “Neural network backflow for ab-initio quantum chemistry,” arXiv:2403.03286  (2024).
  15. I. von Glehn, J. S. Spencer,  and D. Pfau, “A self-attention ansatz for ab-initio quantum chemistry,” arXiv:2211.13672  (2022).
  16. H. Wei and E. Neuscamman, J. Chem. Phys. 149, 184106 (2018).
  17. I. Sabzevari and S. Sharma, J. Chem. Theory Comput. 14, 6276–6286 (2018).
  18. S. Sorella, Phys. Rev. B 64, 024512 (2001).
  19. A. Mahajan and S. Sharma, J. Chem. Theory Comput. 123, 3911–3921 (2019).
  20. D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv:1412.6980  (2014).
  21. L. Otis and E. Neuscamman, Phys. Chem. Chem. Phys. 21, 14491–14510 (2019).
  22. J. Toulouse and C. J. Umrigar, J. Chem. Phys. 126, 084102 (2007).
  23. 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.
  24. J. P. Coe, J. Chem. Theory Comput. 15, 6179–6189 (2019).
  25. Y. Kurashige and T. Yanai, J. Chem. Phys. 130, 234114 (2009).
  26. H. Shang, C. Guo, Y. Wu, Z. Li,  and J. Yang, “Solving schrödinger equation with a language model,” arXiv:2307.09343  (2023).
  27. S. Sharma, “Stochastic perturbation theory to correct non-linearly parametrized wavefunctions,” arXiv:1803.04341  (2018).
  28. V. Kvasnička, Chem. Phys. Lett. 43, 377–381 (1976).
Citations (3)

Summary

We haven't generated a summary for this paper yet.

Dice Question Streamline Icon: https://streamlinehq.com

Follow-up Questions

We haven't generated follow-up questions for this paper yet.