Variational ab initio methods in quantum chemistry stand out among other methods in providing direct access to the wave function. This allows, in principle, straightforward extraction of any other observable of interest, besides the energy, but, in practice, this extraction is often technically difficult and computationally impractical. Here, we consider the electron density as a central observable in quantum chemistry and introduce a novel method to obtain accurate densities from real-space many-electron wave functions by representing the density with a neural network that captures known asymptotic properties and is trained from the wave function by score matching and noise-contrastive estimation. We use variational quantum Monte Carlo with deep-learning Ansätze to obtain highly accurate wave functions free of basis set errors and from them, using our novel method, correspondingly accurate electron densities, which we demonstrate by calculating dipole moments, nuclear forces, contact densities, and other density-based properties.
REFERENCES
1.
2.
R.
Assaraf
, M.
Caffarel
, and A.
Scemama
, “Improved Monte Carlo estimators for the one-body density
,” Phys. Rev. E
75
, 035701
(2007
).3.
D.
Varsano
, M.
Barborini
, and L.
Guidoni
, “Kohn–Sham orbitals and potentials from quantum Monte Carlo molecular densities
,” J. Chem. Phys.
140
, 054102
(2014
).4.
P.
Håkansson
and M.
Mella
, “Efficient and robust quantum Monte Carlo estimate of the total and spin electron densities at nuclei
,” J. Chem. Phys.
129
, 124101
(2008
).5.
M. C.
Per
, I. K.
Snook
, and S. P.
Russo
, “Zero-variance zero-bias quantum Monte Carlo estimators for the electron density at a nucleus
,” J. Chem. Phys.
135
, 134112
(2011
).6.
D. G.
Truhlar
, “Basis-set extrapolation
,” Chem. Phys. Lett.
294
, 45
–48
(1998
).7.
A.
Karton
, N.
Sylvetsky
, and J. M. L.
Martin
, “W4-17: A diverse and high-confidence dataset of atomization energies for benchmarking high-level electronic structure methods
,” J. Comput. Chem.
38
, 2063
–2075
(2017
).8.
S. A.
Kucharski
and R. J.
Bartlett
, “Noniterative energy corrections through fifth-order to the coupled cluster singles and doubles method
,” J. Chem. Phys.
108
, 5243
–5254
(1998
).9.
B.
Silvi
and A.
Savin
, “Classification of chemical bonds based on topological analysis of electron localization functions
,” Nature
371
, 683
–686
(1994
).10.
A.
Hyvärinen
, “Estimation of non-normalized statistical models by score matching
,” J. Mach. Learn. Res.
6
, 695
(2005
).11.
P.
Vincent
, “A connection between score matching and denoising autoencoders
,” Neural Comput.
23
, 1661
–1674
(2011
).12.
Y.
Song
and S.
Ermon
, “Generative modeling by estimating gradients of the data distribution
,” in Advances in Neural Information Processing Systems
, edited by H.
Wallach
, H.
Larochelle
, A.
Beygelzimer
, F.
d’ Alché-Buc
, E.
Fox
, and R.
Garnett
(Curran Associates, Inc.
, 2019
), Vol. 32
.13.
M.
Gutmann
and A.
Hyvärinen
, “Noise-contrastive estimation: A new estimation principle for unnormalized statistical models
,” in Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics
(JMLR Workshop and Conference Proceedings
, 2010
), pp. 297
–304
.14.
J.
Han
, L.
Zhang
, and W.
E
, “Solving many-electron Schrödinger equation using deep neural networks
,” J. Comput. Phys.
399
, 108929
(2019
).15.
J.
Hermann
, Z.
Schätzle
, and F.
Noé
, “Deep-neural-network solution of the electronic Schrödinger equation
,” Nat. Chem.
12
, 891
–897
(2020
).16.
D.
Pfau
, J. S.
Spencer
, A. G.
de G. Matthews
, and W. M. C.
Foulkes
, “Ab initio solution of the many-electron Schrödinger equation with deep neural networks
,” Phys. Rev. Res.
2
, 033429
(2020
).17.
J.
Hermann
, J.
Spencer
, K.
Choo
, A.
Mezzacapo
, W. M. C.
Foulkes
, D.
Pfau
, G.
Carleo
, and F.
Noé
, “Ab initio quantum chemistry with neural-network wavefunctions
,” Nat. Rev. Chem.
7
, 692
–709
(2023
).18.
R.
Assaraf
and M.
Caffarel
, “Zero-variance zero-bias principle for observables in quantum Monte Carlo: Application to forces
,” J. Chem. Phys.
119
, 10536
–10552
(2003
).19.
G.
Carleo
and M.
Troyer
, “Solving the quantum many-body problem with artificial neural networks
,” Science
355
, 602
–606
(2017
).20.
J. S.
Spencer
, D.
Pfau
, A.
Botev
, and W. M. C.
Foulkes
, “Better, faster fermionic neural networks
,” arXiv:2011.07125 [physics.comp-ph] (2020
).21.
Z.
Schätzle
, P. B.
Szabó
, M.
Mezera
, J.
Hermann
, and F.
Noé
, “DeepQMC: An open-source software suite for variational optimization of deep-learning molecular wave functions
,” J. Chem. Phys.
159
, 094108
(2023
).22.
N.
Gao
and S.
Günnemann
, “Neural Pfaffians: Solving many many-electron Schrödinger equations
,” arXiv:2405.14762 (2024
).23.
R.
Li
, H.
Ye
, D.
Jiang
, X.
Wen
, C.
Wang
, Z.
Li
, X.
Li
, D.
He
, J.
Chen
, W.
Ren
, and L.
Wang
, “A computational framework for neural network-based variational Monte Carlo with forward Laplacian
,” Nat. Mach. Intell.
6
, 209
–219
(2024
).24.
L.
Gerard
, M.
Scherbela
, P.
Marquetand
, and P.
Grohs
, “Gold-standard solutions to the Schrödinger equation using deep learning: How much physics do we need?
,” in Advances in Neural Information Processing Systems
, edited by S.
Koyejo
, S.
Mohamed
, A.
Agarwal
, D.
Belgrave
, K.
Cho
, and A.
Oh
(Curran Associates, Inc.
, 2022
), Vol. 35
, pp. 10282
–10294
.25.
J.
Lin
, G.
Goldshlager
, and L.
Lin
, “Explicitly antisymmetrized neural network layers for variational Monte Carlo simulation
,” J. Comput. Phys.
474
, 111765
(2023
).26.
J.
Kim
, G.
Pescia
, B.
Fore
, J.
Nys
, G.
Carleo
, S.
Gandolfi
, M.
Hjorth-Jensen
, and A.
Lovato
, “Neural-network quantum states for ultra-cold Fermi gases
,” Commun. Phys.
7
, 148
(2024
).27.
X.
Li
, C.
Fan
, W.
Ren
, and J.
Chen
, “Fermionic neural network with effective core potential
,” Phys. Rev. Res.
4
, 013021
(2022
).28.
I.
von Glehn
, J. S.
Spencer
, and D.
Pfau
, “A self-attention ansatz for ab-initio quantum chemistry
,” arXiv:2211.13672 (2022
).29.
R. P.
Feynman
and M.
Cohen
, “Energy spectrum of the excitations in liquid helium
,” Phys. Rev.
102
, 1189
–1204
(1956
).30.
D. M.
Ceperley
, “Fermion nodes
,” J. Stat. Phys.
63
, 1237
–1267
(1991
).31.
T.
Kato
, “On the eigenfunctions of many-particle systems in quantum mechanics
,” Commun. Pure Appl. Math.
10
, 151
–177
(1957
).32.
We express these conditions in terms of ∇ log ρ as it gives the cleanest comparison with the density models that we introduce in Sec. III.
33.
S.
Fournais
, M.
Hoffmann-Ostenhof
, T.
Hoffmann-Ostenhof
, and T.
Østergaard Sørensen
, “The electron density is smooth away from the nuclei
,” Commun. Math. Phys.
228
, 401
–415
(2002
).34.
S.
Fournais
, M.
Hoffmann-Ostenhof
, T.
Hoffmann-Ostenhof
, and T.
Østergaard Sørensen
, “Analyticity of the density of electronic wavefunctions
,” Ark. Mat.
42
, 87
–106
(2004
).35.
M.
Hoffmann-Ostenhof
and T.
Hoffmann-Ostenhof
, “‘Schrödinger inequalities’ and asymptotic behavior of the electron density of atoms and molecules
,” Phys. Rev. A
16
, 1782
–1785
(1977
).36.
R.
Ahlrichs
, M.
Hoffmann-Ostenhof
, T.
Hoffmann-Ostenhof
, and J. D.
Morgan
, “Bounds on the decay of electron densities with screening
,” Phys. Rev. A
23
, 2106
–2117
(1981
).37.
The term “energy” in EBMs is not connected to the energy in quantum chemistry that arises as an eigenvalue of the Hamiltonian. The “energy” in our setting is equivalent to −log ρ and thus we use a notation based on Uϕ ≈ −log ρ + C.
38.
39.
G. E.
Hinton
, “Training products of experts by minimizing contrastive divergence
,” Neural Comput.
14
, 1771
–1800
(2002
).40.
41.
L.
Wenliang
, D. J.
Sutherland
, H.
Strathmann
, and A.
Gretton
, “Learning deep kernels for exponential family densities
,” in International Conference on Machine Learning
(PMLR
, 2019
), pp. 6737
–6746
.42.
O.
Chehab
, A.
Gramfort
, and A.
Hyvärinen
, “The optimal noise in noise-contrastive learning is not what you think
,” in Uncertainty in Artificial Intelligence
(PMLR
, 2022
), pp. 307
–316
.43.
NIST
, Atomic spectra database—Ionization energies form—Physics.nist.gov, https://physics.nist.gov/PhysRefData/ASD/ionEnergy.html (accessed 15 May 2024).44.
M.
Filatov
, W.
Zou
, and D.
Cremer
, “Analytic calculation of contact densities and Mössbauer isomer shifts using the normalized elimination of the small-component formalism
,” J. Chem. Theory Comput.
8
, 875
–882
(2012
).45.
E. D.
Hedegard
, J.
Kongsted
, and S. P.
Sauer
, “Validating and analyzing EPR hyperfine coupling constants with density functional theory
,” J. Chem. Theory Comput.
9
, 2380
–2388
(2013
).46.
C.
Filippi
, X.
Gonze
, and C.
Umrigar
, in Recent Developments and Applications of Modern Density Functional Theory
, edited by J. M. Seminario (Elsevier, Amsterdam
, 1996
).47.
G.
Hunter
, “Conditional probability amplitudes in wave mechanics
,” Int. J. Quantum Chem.
9
, 237
–242
(1975
).48.
O. V.
Gritsenko
and E. J.
Baerends
, “Effect of molecular dissociation on the exchange-correlation Kohn–Sham potential
,” Phys. Rev. A
54
, 1957
(1996
).49.
E. J.
Baerends
and O. V.
Gritsenko
, “A quantum chemical view of density functional theory
,” J. Phys. Chem. A
101
, 5383
–5403
(1997
).50.
O. V.
Gritsenko
, R.
van Leeuwen
, and E. J.
Baerends
, “Molecular exchange-correlation Kohn–Sham potential and energy density from ab initio first- and second-order density matrices: Examples for XH (X = Li, B, F)
,” J. Chem. Phys.
104
, 8535
–8545
(1996
).51.
D. G.
Tempel
, T. J.
Martinez
, and N. T.
Maitra
, “Revisiting molecular dissociation in density functional theory: A simple model
,” J. Chem. Theory Comput.
5
, 770
–780
(2009
).52.
L. G.
Diniz
, N.
Kirnosov
, A.
Alijah
, J. R.
Mohallem
, and L.
Adamowicz
, “Accurate dipole moment curve and non-adiabatic effects on the high resolution spectroscopic properties of the LiH molecule
,” J. Mol. Spectrosc.
322
, 22
–28
(2016
).53.
M.
Motta
, D. M.
Ceperley
, G. K.-L.
Chan
, J. A.
Gomez
, E.
Gull
, S.
Guo
, C. A.
Jiménez-Hoyos
, T. N.
Lan
, J.
Li
, F.
Ma
et al, “Towards the solution of the many-electron problem in real materials: Equation of state of the hydrogen chain with state-of-the-art many-body methods
,” Phys. Rev. X
7
, 031059
(2017
).54.
N. C.
Rubin
, R.
Babbush
, and J.
McClean
, “Application of fermionic marginal constraints to hybrid quantum algorithms
,” New J. Phys.
20
, 053020
(2018
).55.
Y.
Qian
, W.
Fu
, W.
Ren
, and J.
Chen
, “Interatomic force from neural network based variational quantum Monte Carlo
,” J. Chem. Phys.
157
, 164104
(2022
).56.
R. J.
Le Roy
, N. S.
Dattani
, J. A.
Coxon
, A. J.
Ross
, P.
Crozet
, and C.
Linton
, “Accurate analytic potentials for Li2 () and Li2 () from 2 to 90 Å, and the radiative lifetime of Li (2p)
,” J. Chem. Phys.
131
, 204309
(2009
).57.
R. J.
Le Roy
, Y.
Huang
, and C.
Jary
, “An accurate analytic potential function for ground-state N2 from a direct-potential-fit analysis of spectroscopic data
,” J. Chem. Phys.
125
, 164310
(2006
).58.
R. P.
Feynman
, “Forces in molecules
,” Phys. Rev.
56
, 340
(1939
).59.
P.
Politzer
and J. S.
Murray
, “The Hellmann–Feynman theorem: A perspective
,” J. Mol. Model.
24
, 266
(2018
).60.
A. B.
Tsybakov
, Introduction to Nonparametric Estimation
(Springer
, New York
, 2009
).© 2025 Author(s). Published under an exclusive license by AIP Publishing.
2025
Author(s)
You do not currently have access to this content.
