Pith. sign in

REVIEW 3 major objections 8 minor 1 cited by

MOLPIPx: an end-to-end differentiable package for permutationally invariant polynomials in Python and Rust

T0 review · 3 major / 8 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read MOLPIPx makes permutationally invariant polynomial potential energy surfaces end-to-end differentiable, with automatic forces for linear, neural network, and Gaussian process models.

desk verdict A genuinely useful JAX/Rust PIP package whose load-bearing MSA-to-code translation lacks a direct permutation-invariance test; send to review with that fixed. read the letter →

arxiv 2411.17011 v3 pith:IR44OEI4 submitted 2024-11-26 physics.chem-ph physics.comp-ph

Manuel S. Drehwald , Asma Jamali , Rodrigo A. Vargas-Hernández This is my paper · ORCID
classification physics.chem-phphysics.comp-ph
keywords permutationallyinvariantpolynomialspotentialenergysurfacesautomaticdifferentiationJAXEnzymeADGaussianprocessregressionneuralnetworkpotentialsMorsevariables
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Potential energy surfaces describe how a molecule's energy changes with geometry, and permutationally invariant polynomials (PIPs) are a widely used way to build them because they respect the fact that swapping identical atoms leaves the energy unchanged. MOLPIPx is a software package that makes PIP-based models end-to-end differentiable: energies, forces, and higher derivatives are obtained automatically through automatic differentiation, removing the need for hand-coded gradients. It translates monomial and polynomial basis files, generated by the monomial symmetrization algorithm for 15 molecular symmetry types with polynomial degrees from 3 to 7, into Python (JAX) and Rust (EnzymeAD) code. The package also introduces anisotropic Morse variables with per-distance-type length scales optimized by gradient-based implicit differentiation. If this works as claimed, building and training accurate force fields for molecules becomes substantially more accessible.

What carries the argument

The central object is the PIP vector $\Phi_{\mathrm{PIP}}(x) = (f_{\mathrm{poly}} \circ f_{\mathrm{mono}} \circ \gamma \circ d)(x)$. It maps Cartesian coordinates $x$ to inter-nuclear distances $r$, then to Morse variables $\gamma = e^{-\lambda r}$, then to symmetrized monomials via $f_{\mathrm{mono}}$, and finally to symmetrized polynomials via $f_{\mathrm{poly}}$; the symmetrization enforces invariance under permutation of like atoms. Automatic differentiation engines (JAX for Python, EnzymeAD for Rust) differentiate through this composition so that energies, forces, and higher derivatives are obtained as gradients. The package also introduces anisotropic Morse variables $\gamma_{\mathrm{aniso}} = \exp(-\sum_{\ell} \lambda_\ell \omega_\ell \odot r)$, where $\omega_\ell$ are binary mask vectors selecting each type of atom-atom distance, with $\lambda_\ell$ optimized by implicit differentiation.

What would settle it

For each of the 15 symmetry types and each degree 3 to 7, evaluate the translated PIP vector on a random geometry and on the same geometry with identical-atom labels permuted; if any component of the vector or the resulting energy and force differs beyond numerical tolerance, the translation is not permutationally invariant and the central claim fails.

Watch

Extended reading notes

Core claim

The authors claim that a PIP-based potential energy model can be expressed as a smooth composition of functions, $\Phi_{\mathrm{PIP}}(x) = (f_{\mathrm{poly}} \circ f_{\mathrm{mono}} \circ \gamma \circ d)(x)$, where $x$ are Cartesian coordinates, $d$ gives inter-nuclear distances, $\gamma$ applies Morse variables $e^{-\lambda r}$, and $f_{\mathrm{mono}}$ and $f_{\mathrm{poly}}$ build symmetrized monomials and polynomials. Because every layer is differentiable, the same $\Phi_{\mathrm{PIP}}$ vector feeds linear regression (with least-squares fitting that can include force Jacobians), feed-forward neural networks, and Gaussian processes with invariant kernels. Forces are obtained as gradients of the model output, and anisotropic Morse variables introduce per-type length scales whose outer/inner optimization uses implicit differentiation. The library ships pre-generated functions for 15 molecular symmetries and a converter that turns MSA-generated files into JAX or Rust code, with the Rust/EnzymeAD backend currently supporting linear models.

Load-bearing premise

The load-bearing assumption is that the automated translation of MSA-generated monomial and polynomial files into JAX and Rust code produces correctly symmetrized PIP vectors for all 15 molecules and degrees 3 to 7, since no direct test of equal energies for permuted geometries is reported.

Editorial extensions

If this is right

  • A user can train linear PIP models, PIP neural networks, and PIP Gaussian processes from the same differentiable $\Phi_{\mathrm{PIP}}$ representation, with forces obtained as automatic gradients.
  • Force-augmented least squares for linear models becomes routine because the Jacobian $\nabla_x \Phi_{\mathrm{PIP}}$ is available automatically.
  • The anisotropic Morse variables allow per-distance-type length scales to be learned by gradient descent with implicit differentiation, replacing grid search.
  • Post-compilation joint energy-force evaluation is faster than $10^{-3}$ seconds for low-degree models, and the Rust/EnzymeAD backend compiles large systems (degree 8, five atoms) in 33.5 seconds where JAX jit takes hours.
  • The same machinery transfers to many-body expansions, since the supplied symmetry types can be used term by term.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the package is correct, PIP fitting can be embedded as a differentiable layer in larger end-to-end pipelines, for example training a force field jointly with downstream observables rather than fitting energies alone.
  • A natural extension is to treat the per-distance-type $\lambda_\ell$ as fully learned latent variables shared across a family of molecules, which the differentiable formulation makes straightforward.
  • The compile-time advantage of the Rust backend suggests ahead-of-time compilation could make very high-degree PIPs practical on larger molecules, a regime where JAX's jit cost is prohibitive.
  • A quick validation of permutational invariance on randomly permuted geometries for each of the 15 molecules would directly test whether the automated translation preserves the symmetry that the whole approach relies on.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 8 minor

Summary. The paper presents MOLPIPx, an open-source software package that constructs permutationally invariant polynomial (PIP) basis vectors from MSA-generated monomial and polynomial files and exposes them through JAX and Rust/EnzymeAD automatic differentiation. The library is designed to support linear PIP models with energy and force fitting, PIP-based neural networks, Gaussian processes, and anisotropic Morse variables whose length scales are optimized by implicit differentiation of an inner least-squares problem. Demonstrations are carried out for methane using the external CH4 dataset of Ref. 28, and the paper reports force/energy prediction times, RMSE values, and optimization trajectories for lambda and hyperparameters.

Significance. If the MSA-to-JAX/Rust translation is correct for all fifteen symmetry types and polynomial degrees listed, MOLPIPx fills a practical gap by removing the need to hand-code derivatives of PIP features and by making those features directly usable in standard ML frameworks such as Flax, GPJax, and Optax. The open-source release, the explicit code listings, and the use of external benchmark data are clear strengths; the benchmark data are independent of the package, so there is no evident circularity in the numerical results. The implicit-differentiation treatment of the Morse length scales is standard methodology rather than a new theoretical claim, but its integration with the PIP pipeline is useful. The main value of the paper is therefore software engineering: a flexible, differentiable PIP layer with automatic force computation.

major comments (3)
  1. [Section III, Fig. 2, Table I, Listings 2–3] The central load-bearing step is the automated translation of MSA .MONO/.POLY files into JAX and Rust functions, but the paper provides no direct verification of permutational invariance for the generated PIP vectors. Every downstream model in Eq. (1) inherits any indexing or code-generation error in fmono and fpoly; yet the only shown code is for A2B at degree 3, and the only tested molecule is CH4. I request a direct invariance test for all fifteen symmetry types in Table I and all degrees 3–7, for example by checking that Phi_PIP(x) is unchanged under permutations of like atoms, or by comparing the generated JAX/Rust output against the reference MSA output for a set of random geometries. In addition, Table I states that polynomial degrees range from 3 to 7, while Section IV.B reports timings for p = 8; the supported degree range should be stated consistently.
  2. [Sections II C and IV C, Eq. (11), Fig. 4] The anisotropic Morse-variable formulation is presented as a contribution, but the paper does not report whether optimizing lambda_AA and lambda_AB improves accuracy relative to an isotropic lambda. Figure 4 shows optimization trajectories for methane, but it does not give final lambda values, does not compare the outer validation loss against an isotropic or fixed-lambda baseline, and does not report test-set RMSE for the optimized model. Without such a comparison, the claim that the anisotropic approach is beneficial rather than merely differentiable is unsupported.
  3. [Abstract, Section IV.B, Section IV] The abstract states that MOLPIPx facilitates the computation of energy gradients and higher-order derivatives via JAX and EnzymeAD-Rust, but the only derivative computation demonstrated in the paper is first derivatives (forces). No Hessian or higher-order derivative example is provided for either backend, and Section IV explicitly states that the Rust version is currently limited to linear models. Please either demonstrate a higher-order derivative computation for at least one model (e.g., a Hessian calculation for a JAX linear or NN model) or qualify the abstract so that the validated scope is stated precisely.
minor comments (8)
  1. [Section II B, Eq. (7)] The block matrix in Eq. (7) is hard to read and appears to contain indexing inconsistencies, for example rows labeled with partial derivatives of φ_np(x1) with respect to x3na but columns listing φ_np(x2). Please rewrite it with clear block notation that defines the geometry index and the ordering of the force rows.
  2. [Section II C, Eq. (11)] The constraint lambda in R is misleading because positivity is enforced later via softplus; please write the constraint as lambda in R^d (with positivity handled separately) or as lambda in R_{>0} once the softplus parameterization is introduced.
  3. [Section IV.D, Fig. 5] The claim that the optimal value is lambda0 = 0.1 is based on a single grid-search run; please state the number of random seeds or repeated training runs used for each grid point, or otherwise explain why the comparison is robust to initialization noise.
  4. [Section IV.B, Fig. 3] The text says JAX jitting times for p = 8 are on the order of 10 hours, while the figure caption says the jitting time exceeded 5 hours for molecules with 5 atoms; please make the two statements consistent.
  5. [Section III, Listing 4] The decoration of a Flax Module class with @nn.jit may be confusing to readers unfamiliar with the exact Flax version used; a short note clarifying the intended Flax API version and the role of the decorator would help.
  6. [Section IV, Figs. 3 and 7] Timing and RMSE values are reported without uncertainties or replicate counts; either provide standard deviations over a small number of runs or explicitly state that the values are single-run measurements.
  7. [Table I, Section III] The entries of Table I are described as molecules, but most are abstract symmetry types such as A3 and ABCD; please clarify in the caption that these are generic atom-type labels rather than specific chemical species.
  8. [References] There are a few proofreading issues in the reference list, for example the text 'S. Manzhos, Sergei et al.' in the sentence preceding Ref. 91, and some entries mix author initials inconsistently; a careful proofread is recommended.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the PIP construction and benchmarks are grounded in external MSA code and external CH4 data; no prediction reduces to a fitted input.

full rationale

MOLPIPx's central derivation chain is self-contained. ΦPIP is defined by Eq. (1) as a composition of inter-nuclear distances, Morse variables, and symmetrized monomial/polynomial functions, where fmono and fpoly are produced by the external MSA algorithm of Xie and Bowman (Ref. 99) and translated into JAX/Rust by msa_file_generator. No equation in the paper defines a quantity in terms of the claim it is supposed to establish; the anisotropic Morse extension (Eqs. 8-10) constructs invariance by assigning one λ per distance type, which is a definitional construction rather than a circular validation. The reported CH4 errors come from fitting to, and testing against, data taken from Ref. 28, an external benchmark set; the RMSE numbers are not inputs to the model's construction. The only self-references (Refs. 101, 113 for implicit differentiation; several review citations) are used as methodological tools and not as the sole support of the paper's central claim; no uniqueness theorem or ansatz is imported from the authors' prior work to force the choice of representation. The absence of a direct permutational-invariance test for the translated MSA files is a validation gap and a correctness risk, but it is not a circularity: it concerns whether the external algorithm was transcribed faithfully, not whether the paper's inputs already contain its outputs.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical entities. The free parameters listed are model hyperparameters, not fundamental constants. The most consequential axiom is the assumed correctness of the MSA-generated polynomial basis; the paper does not independently verify it.

free parameters (3)
  • lambda (isotropic Morse length-scale) = 1.0 (default in Listing 4)
    Hyperparameter in the PIP layer that controls the exponential decay of Morse variables. In the linear and NN models it is either fixed or optimized, but its choice affects the representation.
  • lambda_AA, lambda_AB (anisotropic Morse length-scales) = Optimized values for CH4 (Fig. 4)
    The new anisotropic extension assigns per-distance-type length scales and optimizes them via bilevel optimization. The paper shows optimization trajectories but does not report final test accuracy gains over isotropic models.
  • lambda0 (energy-force loss weight) = 0.1 (grid search result, Sec IV.D)
    Relative weighting between energy and force terms in the PIP-NN loss function (Eq. 15). Found by grid search on a single molecule (CH4), so it is a fitted hyperparameter.
assumptions (4)
  • domain assumption The monomial symmetrization algorithm (MSA) produces a complete basis of permutationally invariant polynomials for the described molecular symmetries.
    The paper relies on MSA from Ref. 99 to generate f_mono and f_poly functions; correctness of this basis is standard in the PIP literature but is not proven in the paper.
  • domain assumption The inter-nuclear distance vector r = d(x) is a complete descriptor of the molecular geometry up to rotation and translation for the PES problems considered.
    Invoked in Sec. II A, Eq. 2; standard for gas-phase molecular PES fitting, where total energy is invariant under rigid motions.
  • ad hoc to paper The implicit differentiation of the inner least-squares solution w(lambda) with respect to lambda is valid, requiring the inner optimization to have a unique solution.
    Eq. 11 and surrounding text in Sec. II C assume the linear system (Eq. 7) has full rank so the solution is differentiable. This may fail for high-degree PIPs when the number of features exceeds the number of training points.
  • domain assumption The training data obtained from Ref. 28 accurately represents the PES of methane.
    All examples in Sec. IV use the methane data from Ref. 28; if this reference data contains errors, the reported RMSE values would be affected.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MOLPIPx: an end-to-end differentiable package for permutationally invariant polynomials in Python and Rust." pith.science (2026). https://pith.science/paper/IR44OEI4

@misc{pith2026241117011,
  author       = {Pith},
  title        = {Pith review of: MOLPIPx: an end-to-end differentiable package for permutationally invariant polynomials in Python and Rust},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IR44OEI4}},
  note         = {Machine review of arXiv:2411.17011}
}
read the original abstract

In this work, we present MOLPIPx, a versatile library designed to seamlessly integrate Permutationally Invariant Polynomials (PIPs) with modern machine learning frameworks, enabling the efficient development of linear models, neural networks, and Gaussian process models. These methodologies are widely employed for parameterizing potential energy surfaces across diverse molecular systems. MOLPIPx leverages two powerful automatic differentiation engines -JAX and EnzymeAD-Rust- to facilitate the efficient computation of energy gradients and higher-order derivatives, which are essential for tasks such as force field development and dynamic simulations. MOLPIPx is available at https://github.com/ChemAI-Lab/molpipx.

Figures

Figures reproduced from arXiv: 2411.17011 by the authors.

Figure 1
Figure 1. FIG. 1. Overview of the MOLPIPx package. For any model devel￾oped under MOLPIPx, the monomial and polynomial functions are constructed using the MSA algorithm and then translated into JAX or Rust, depending on the selected back-end. We provide some mono￾mial and polynomial functions, see Table I. The input for all models is the Cartesian coordinates of a molecule, which are used to con￾struct the PIP vector (Eq. 1).… view at source ↗
Figure 2
Figure 2. FIG. 2. Overview of the workflow to construct the fm [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Jitting times (a) and post-jitting run times (b) for the joint [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Optimization trajectories of ten random initial [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 6
Figure 6. Figure 6: FIG. 6. (a) Training loss and (b) validation loss for energy predictions [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Exact quantum dynamics of methanol: full-dimensional ab initio potential energy surface of spectroscopic quality and variational vibrational states

    physics.chem-ph 2025-05 conditional novelty 6.0 of 10

    A new ab initio potential energy surface for methanol, trained on 39,401 CCSD(T)-F12b/cc-pVTZ-F12 energies, yields variational vibrational band origins within about 5 cm-1 of experiment up to ~2500 cm-1.

Reference graph

Works this paper leans on

152 extracted references · 66 canonical work pages · cited by 1 Pith paper

  1. [1]

    Brown , author A

    author author A. Brown , author A. B. \ McCoy , author B. J. \ Braams , author Z. Jin , \ and\ author J. M. \ Bowman ,\ title title Quantum and classical studies of vibrational motion of CH5+ on a global potential energy surface obtained from a novel ab initio direct dynamics approach , \ @noop journal journal The Journal of chemical physics \ volume 121 ...

    2004

  2. [2]

    Czak\' o \ and\ author J

    author author G. Czak\' o \ and\ author J. M. \ Bowman ,\ title title Reaction dynamics of methane with F , O , Cl , and Br on ab initio potential energy surfaces , \ @noop journal journal The Journal of Physical Chemistry A \ volume 118 ,\ pages 2839--2864 ( year 2014 ) NoStop

    2014

  3. [3]

    author author K. A. \ Peterson , author S. Skokov , \ and\ author J. M. \ Bowman ,\ title title A theoretical study of the vibrational energy spectrum of the HOCl / HClO system on an accurate ab initio potential energy surface , \ @noop journal journal The Journal of chemical physics \ volume 111 ,\ pages 7446--7456 ( year 1999 ) NoStop

    1999

  4. [4]

    author author B. C. \ Shepler , author B. J. \ Braams , \ and\ author J. M. \ Bowman ,\ title title Quasiclassical trajectory calculations of acetaldehyde dissociation on a global potential energy surface indicate significant non-transition state dynamics , \ @noop journal journal The Journal of Physical Chemistry A \ volume 111 ,\ pages 8282--8285 ( year...

    2007

  5. [5]

    Czak \'o \ and\ author J

    author author G. Czak \'o \ and\ author J. M. \ Bowman ,\ title title Accurate ab initio potential energy surface, thermochemistry, and dynamics of the Cl(^2P, ^2P_ 3/2 ) + CH4 -> HCl + CH3 and H + CH3Cl reactions , \ @noop journal journal The Journal of Chemical Physics \ volume 136 ( year 2012 ) NoStop

    2012

  6. [6]

    Zou \ and\ author J

    author author S. Zou \ and\ author J. M. \ Bowman ,\ title title A new ab initio potential energy surface describing acetylene/vinylidene isomerization , \ @noop journal journal Chemical physics letters \ volume 368 ,\ pages 421--424 ( year 2003 ) NoStop

    2003

  7. [7]

    Huang , author S

    author author X. Huang , author S. Carter , \ and\ author J. Bowman ,\ title title Ab initio potential energy surface and rovibrational energies of H3O+ and its isotopomers , \ @noop journal journal The Journal of chemical physics \ volume 118 ,\ pages 5431--5441 ( year 2003 ) NoStop

    2003

  8. [8]

    Homayoon , author R

    author author Z. Homayoon , author R. Conte , author C. Qu , \ and\ author J. M. \ Bowman ,\ title title Full-dimensional, high-level ab initio potential energy surfaces for H2(H2O) and H2(H2O) with application to hydrogen clathrate hydrates , \ @noop journal journal The Journal of Chemical Physics \ volume 143 ,\ pages 084302 ( year 2015 ) NoStop

    2015

Show all 152 references
  1. [9]

    Wang , author P

    author author X. Wang , author P. L. \ Houston , \ and\ author J. M. \ Bowman ,\ title title A new (multi-reference configuration interaction) potential energy surface for H2CO and preliminary studies of roaming , \ @noop journal journal Philosophical Transactions of the Royal...

    2017

  2. [10]

    author author A. R. \ Sharma , author J. Wu , author B. J. \ Braams , author S. Carter , author R. Schneider , author B. Shepler , \ and\ author J. M. \ Bowman ,\ title title Potential energy surface and MULTIMODE vibrational analysis of C2H3+ , \ @noop journal journal The Jou...

    2006

  3. [11]

    Huang , author S

    author author X. Huang , author S. Carter , \ and\ author J. M. \ Bowman ,\ title title Ab initio potential energy surface and vibrational energies of H3O+ and its isotopomers , \ @noop journal journal The Journal of Physical Chemistry B \ volume 106 ,\ pages 8182--8188 ( year...

    2002

  4. [12]

    Xu , author H

    author author D. Xu , author H. Guo , author S. Zou , \ and\ author J. Bowman ,\ title title A scaled ab initio potential energy surface for acetylene and vinylidene , \ @noop journal journal Chemical physics letters \ volume 377 ,\ pages 582--588 ( year 2003 ) NoStop

    2003

  5. [13]

    Pandey , author C

    author author P. Pandey , author C. Qu , author A. Nandi , author Q. Yu , author P. L. \ Houston , author R. Conte , \ and\ author J. M. \ Bowman ,\ title title Ab initio potential energy surface for NaCl – H2 with correct long-range behavior , \ 10.1021/acs.jpca.3c07687 journ...

    doi

  6. [14]

    Conte , author P

    author author R. Conte , author P. L. \ Houston , \ and\ author J. M. \ Bowman ,\ title title Communication: A benchmark-quality, full-dimensional ab initio potential energy surface for Ar-HOCO , \ @noop journal journal The Journal of Chemical Physics \ volume 140 ,\ pages 151...

    2014

  7. [15]

    author author J. M. \ Bowman , author B. J. \ Braams , author S. Carter , author C. Chen , author G. Czak \'o , author B. Fu , author X. Huang , author E. Kamarchik , author A. R. \ Sharma , author B. C. \ Shepler , author Y. Wang , \ and\ author Z. Xie ,\ title title Ab-initi...

    2010

  8. [16]

    author author P. O. \ Dral ,\ title title Quantum chemistry in the age of machine learning , \ @noop journal journal The journal of physical chemistry letters \ volume 11 ,\ pages 2336--2347 ( year 2020 ) NoStop

    2020

  9. [17]

    Fr \"o hlking , author M

    author author T. Fr \"o hlking , author M. Bernetti , author N. Calonaci , \ and\ author G. Bussi ,\ title title Toward empirical force fields that match experimental observables , \ @noop journal journal The Journal of chemical physics \ volume 152 ,\ pages 230902 ( year 2020...

    2020

  10. [18]

    Jiang , author J

    author author B. Jiang , author J. Li , \ and\ author H. Guo ,\ title title High-fidelity potential energy surfaces for gas-phase and gas--surface scattering processes from machine learning , \ @noop journal journal The Journal of Physical Chemistry Letters \ volume 11 ,\ page...

    2020

  11. [19]

    Jinnouchi , author K

    author author R. Jinnouchi , author K. Miwa , author F. Karsai , author G. Kresse , \ and\ author R. Asahi ,\ title title On-the-fly active learning of interatomic potentials for large-scale atomistic simulations , \ @noop journal journal The Journal of Physical Chemistry Lett...

    2020

  12. [20]

    Koner , author S

    author author D. Koner , author S. M. \ Salehi , author P. Mondal , \ and\ author M. Meuwly ,\ title title Non-conventional force fields for applications in spectroscopy and chemical reaction dynamics , \ @noop journal journal The Journal of chemical physics \ volume 153 ,\ pa...

    2020

  13. [21]

    Poltavsky \ and\ author A

    author author I. Poltavsky \ and\ author A. Tkatchenko ,\ title title Machine learning force fields: Recent advances and remaining challenges , \ @noop journal journal The journal of physical chemistry letters \ volume 12 ,\ pages 6551--6564 ( year 2021 ) NoStop

    2021

  14. [22]

    Tong , author P

    author author Q. Tong , author P. Gao , author H. Liu , author Y. Xie , author J. Lv , author Y. Wang , \ and\ author J. Zhao ,\ title title Combining machine learning potential and structure prediction for accelerated materials design and discovery , \ @noop journal journal T...

    2020

  15. [23]

    Westermayr , author M

    author author J. Westermayr , author M. Gastegger , author K. T. \ Sch \"u tt , \ and\ author R. J. \ Maurer ,\ title title Perspective on integrating machine learning into computational chemistry and materials science , \ @noop journal journal The Journal of Chemical Physics ...

    2021

  16. [24]

    author author B. J. \ Braams \ and\ author J. M. \ Bowman ,\ title title Permutationally invariant potential energy surfaces in high dimensionality , \ @noop journal journal International Reviews in Physical Chemistry \ volume 28 ,\ pages 577--606 ( year 2009 ) NoStop

    2009

  17. [25]

    Qu , author Q

    author author C. Qu , author Q. Yu , \ and\ author J. M. \ Bowman ,\ title title Permutationally invariant potential energy surfaces , \ @noop journal journal Annual review of physical chemistry \ volume 69 ,\ pages 151--175 ( year 2018 ) NoStop

    2018

  18. [26]

    Pandey , author M

    author author P. Pandey , author M. Arandhara , author P. L. \ Houston , author C. Qu , author R. Conte , author J. M. \ Bowman , \ and\ author S. G. \ Ramesh ,\ title title Assessing permutationally invariant polynomial and symmetric gradient domain machine learning potential...

    doi

  19. [27]

    Xie \ and\ author J

    author author T. Xie \ and\ author J. M. \ Bowman ,\ title title On using potential, gradient, and hessian data in least squares fits of potentials: Application and tests for H2O , \ @noop journal journal The Journal of chemical physics \ volume 117 ,\ pages 10487--10492 ( yea...

    2002

  20. [28]

    Nandi , author C

    author author A. Nandi , author C. Qu , \ and\ author J. M. \ Bowman ,\ title title Using gradients in permutationally invariant polynomial potential fitting: A demonstration for CH4 using as few as 100 configurations , \ 10.1021/acs.jctc.9b00043 journal journal Journal of Che...

    doi

  21. [29]

    Conte , author C

    author author R. Conte , author C. Qu , author P. L. \ Houston , \ and\ author J. M. \ Bowman ,\ title title Efficient generation of permutationally invariant potential energy surfaces for large molecules , \ @noop journal journal Journal of Chemical Theory and Computation \ v...

    2020

  22. [30]

    author author P. L. \ Houston , author C. Qu , author A. Nandi , author R. Conte , author Q. Yu , \ and\ author J. M. \ Bowman ,\ title title Permutationally invariant polynomial regression for energies and gradients, using reverse differentiation, achieves orders of magnitude...

    doi

  23. [31]

    author author O. T. \ Unke , author S. Chmiela , author H. E. \ Sauceda , author M. Gastegger , author I. Poltavsky , author K. T. \ Schütt , author A. Tkatchenko , \ and\ author K.-R. \ Müller ,\ title title Machine learning force fields , \ 10.1021/acs.chemrev.0c01111 journa...

    doi

  24. [32]

    Murakami , author S

    author author T. Murakami , author S. Takahashi , author Y. Kikuma , \ and\ author T. Takayanagi ,\ title title Theoretical study of the thermal rate coefficients of the H3+ + C2H4 reaction: Dynamics study on a full-dimensional potential energy surface , \ @noop journal journa...

    2024

  25. [33]

    Liu , author W

    author author X. Liu , author W. Wang , \ and\ author J. P \'e rez-R \' os ,\ title title Molecular dynamics-driven global potential energy surfaces: Application to the alf dimer , \ @noop journal journal The Journal of Chemical Physics \ volume 159 ,\ pages 144103 ( year 2023...

    2023

  26. [34]

    author author O. T. \ Unke , author D. Koner , author S. Patra , author S. K \"a ser , \ and\ author M. Meuwly ,\ title title High-dimensional potential energy surfaces for molecular simulations: from empiricism to machine learning , \ @noop journal journal Machine Learning: S...

    2020

  27. [35]

    Zhang \ and\ author B

    author author Y. Zhang \ and\ author B. Jiang ,\ title title Universal machine learning for the response of atomistic systems to external fields , \ @noop journal journal Nature Communications \ volume 14 ,\ pages 6424 ( year 2023 ) NoStop

    2023

  28. [36]

    Behler ,\ title title Neural network potential-energy surfaces in chemistry: a tool for large-scale simulations , \ 10.1039/C1CP21668F journal journal Phys

    author author J. Behler ,\ title title Neural network potential-energy surfaces in chemistry: a tool for large-scale simulations , \ 10.1039/C1CP21668F journal journal Phys. Chem. Chem. Phys. \ volume 13 ,\ pages 17930--17955 ( year 2011 ) NoStop

    doi

  29. [37]

    author author J. M. \ Bowman , author C. Qu , author R. Conte , author A. Nandi , author P. L. \ Houston , \ and\ author Q. Yu ,\ title title -machine learned potential energy surfaces and force fields , \ @noop journal journal Journal of Chemical Theory and Computation \ volu...

    2022

  30. [38]

    Brown , author B

    author author A. Brown , author B. J. \ Braams , author K. Christoffel , author Z. Jin , \ and\ author J. M. \ Bowman ,\ title title Classical and quasiclassical spectral analysis of CH5+ using an ab initio potential energy surface , \ @noop journal journal The Journal of chem...

    2003

  31. [39]

    Xie , author B

    author author Z. Xie , author B. J. \ Braams , \ and\ author J. M. \ Bowman ,\ title title Ab initio global potential-energy surface for H5+ -> H3+ + H2 , \ @noop journal journal The Journal of chemical physics \ volume 122 ,\ pages 224307 ( year 2005 ) NoStop

    2005

  32. [40]

    Zhang , author B

    author author X. Zhang , author B. J. \ Braams , \ and\ author J. M. \ Bowman ,\ title title An ab initio potential surface describing abstraction and exchange for H + CH4 , \ @noop journal journal The Journal of chemical physics \ volume 124 ,\ pages 021104 ( year 2006 ) NoStop

    2006

  33. [41]

    Nandi , author C

    author author A. Nandi , author C. Qu , author P. L. \ Houston , author R. Conte , \ and\ author J. M. \ Bowman ,\ title title -machine learning for potential energy surfaces: A PIP approach to bring a DFT -based PES to CCSD(T) level of theory , \ @noop journal journal The Jou...

    2021

  34. [42]

    Wang , author S

    author author X. Wang , author S. Carter , \ and\ author J. M. \ Bowman ,\ title title Pruning the hamiltonian matrix in MULTIMODE : Test for C2H4 and application to CH3NO2 using a new ab initio potential energy surface , \ @noop journal journal The Journal of Physical Chemist...

    2015

  35. [43]

    Qu \ and\ author J

    author author C. Qu \ and\ author J. M. \ Bowman ,\ title title An ab initio potential energy surface for the formic acid dimer: zero-point energy, selected anharmonic fundamental energies, and ground-state tunneling splitting calculated in relaxed 1--4-mode subspaces , \ @noo...

    2016

  36. [44]

    Conte , author P

    author author R. Conte , author P. L. \ Houston , author C. Qu , author J. Li , \ and\ author J. M. \ Bowman ,\ title title Full-dimensional, ab initio potential energy surface for glycine with characterization of stationary points and zero-point energy calculations by means o...

    2020

  37. [45]

    Nandi , author C

    author author A. Nandi , author C. Qu , \ and\ author J. M. \ Bowman ,\ title title Full and fragmented permutationally invariant polynomial potential energy surfaces for trans and cis n-methyl acetamide and isomerization saddle points , \ @noop journal journal The Journal of ...

    2019

  38. [46]

    Qu \ and\ author J

    author author C. Qu \ and\ author J. M. \ Bowman ,\ title title A fragmented, permutationally invariant polynomial approach for potential energy surfaces of large molecules: Application to n-methyl acetamide , \ @noop journal journal The Journal of Chemical Physics \ volume 15...

    2019

  39. [47]

    Qu , author R

    author author C. Qu , author R. Conte , author P. L. \ Houston , \ and\ author J. M. \ Bowman ,\ title title Full-dimensional potential energy surface for acetylacetone and tunneling splittings , \ @noop journal journal Physical Chemistry Chemical Physics \ volume 23 ,\ pages ...

    2021

  40. [48]

    Houston , author R

    author author P. Houston , author R. Conte , author C. Qu , \ and\ author J. M. \ Bowman ,\ title title Permutationally invariant polynomial potential energy surfaces for tropolone and H and D atom tunneling dynamics , \ @noop journal journal The Journal of Chemical Physics \ ...

    2020

  41. [49]

    author author P. L. \ Houston , author C. Qu , author Q. Yu , author R. Conte , author A. Nandi , author J. K. \ Li , \ and\ author J. M. \ Bowman ,\ title title PESPIP : Software to fit complex molecular and many-body potential energy surfaces with permutationally invariant p...

    2023

  42. [50]

    first principles

    author author V. Babin , author G. R. \ Medders , \ and\ author F. Paesani ,\ title title Development of a “first principles” water potential with flexible monomers. II : Trimer potential energy surface, third virial coefficient, and small clusters , \ @noop journal journal Jo...

    2014

  43. [51]

    author author S. K. \ Reddy , author S. C. \ Straight , author P. Bajaj , author C. Huy Pham , author M. Riera , author D. R. \ Moberg , author M. A. \ Morales , author C. Knight , author A. W. \ G \"o tz , \ and\ author F. Paesani ,\ title title On the accuracy of the MB -pol...

    2016

  44. [52]

    Palos , author S

    author author E. Palos , author S. Dasgupta , author E. Lambros , \ and\ author F. Paesani ,\ title title Data-driven many-body potentials from density functional theory for aqueous phase chemistry , \ @noop journal journal Chemical Physics Reviews \ volume 4 ,\ pages 011301 (...

    2023

  45. [53]

    Palos , author A

    author author E. Palos , author A. Caruso , \ and\ author F. Paesani ,\ title title Consistent density functional theory-based description of ion hydration through density-corrected many-body representations , \ @noop journal journal The Journal of Chemical Physics \ volume 15...

    2023

  46. [54]

    author author E. F. \ Bull-Vulpe , author M. Riera , author A. W. \ G \"o tz , \ and\ author F. Paesani ,\ title title MB-Fit : Software infrastructure for data-driven many-body potential energy functions , \ @noop journal journal The Journal of Chemical Physics \ volume 155 ,...

    2021

  47. [55]

    Riera , author C

    author author M. Riera , author C. Knight , author E. F. \ Bull-Vulpe , author X. Zhu , author H. Agnew , author D. G. \ Smith , author A. C. \ Simmonett , \ and\ author F. Paesani ,\ title title MBX : A many-body energy and force calculator for data-driven many-body simulatio...

    2023

  48. [56]

    Riera , author E

    author author M. Riera , author E. P. \ Yeh , \ and\ author F. Paesani ,\ title title Data-driven many-body models for molecular fluids: CO2 / H2O mixtures as a case study , \ @noop journal journal Journal of Chemical Theory and Computation \ volume 16 ,\ pages 2246--2257 ( ye...

    2020

  49. [57]

    Bowman , author C

    author author J. Bowman , author C. Qu , author P. Houston , author T. Allison , \ and\ author B. Schneider ,\ title title DFT -based permutationally invariant polynomial potentials capture the twists and turns of C14H30 , \ @noop journal journal Journal of Chemical Theory and...

    2024

  50. [58]

    Gy \"oo ri \ and\ author G

    author author T. Gy \"oo ri \ and\ author G. Czako \'o ,\ title title Automating the development of high-dimensional reactive potential energy surfaces with the robosurfer program system , \ @noop journal journal Journal of Chemical Theory and Computation \ volume 16 ,\ pages ...

    2019

  51. [59]

    author author D. R. \ Moberg \ and\ author A. W. \ Jasper ,\ title title Permutationally invariant polynomial expansions with unrestricted complexity , \ @noop journal journal Journal of Chemical Theory and Computation \ volume 17 ,\ pages 5440--5455 ( year 2021 ) NoStop

    2021

  52. [60]

    author author D. R. \ Moberg , author A. W. \ Jasper , \ and\ author M. J. \ Davis ,\ title title Parsimonious potential energy surface expansions using dictionary learning with multipass greedy selection , \ @noop journal journal The Journal of Physical Chemistry Letters \ vo...

    2021

  53. [61]

    Koner \ and\ author M

    author author D. Koner \ and\ author M. Meuwly ,\ title title Permutationally invariant, reproducing kernel-based potential energy surfaces for polyatomic molecules: From formaldehyde to acetone , \ @noop journal journal Journal of chemical theory and computation \ volume 16 ,...

    2020

  54. [62]

    Varga \ and\ author D

    author author Z. Varga \ and\ author D. G. \ Truhlar ,\ title title Potential energy surface for high-energy N + N2 collisions , \ @noop journal journal Physical Chemistry Chemical Physics \ volume 23 ,\ pages 26273--26284 ( year 2021 ) NoStop

    2021

  55. [63]

    Varga , author Y

    author author Z. Varga , author Y. Liu , author J. Li , author Y. Paukku , author H. Guo , \ and\ author D. G. \ Truhlar ,\ title title Potential energy surfaces for high-energy N + O2 collisions , \ @noop journal journal The Journal of Chemical Physics \ volume 154 ,\ pages 0...

    2021

  56. [64]

    Li , author Z

    author author J. Li , author Z. Varga , author D. G. \ Truhlar , \ and\ author H. Guo ,\ title title Many-body permutationally invariant polynomial neural network potential energy surface for N4 , \ @noop journal journal Journal of Chemical Theory and Computation \ volume 16 ,...

    2020

  57. [65]

    Paukku , author K

    author author Y. Paukku , author K. R. \ Yang , author Z. Varga , \ and\ author D. G. \ Truhlar ,\ title title Global ab initio ground-state potential energy surface of N4 , \ @noop journal journal The Journal of chemical physics \ volume 139 ,\ pages 044309 ( year 2013 ) NoStop

    2013

  58. [66]

    Conte , author C

    author author R. Conte , author C. Qu , \ and\ author J. M. \ Bowman ,\ title title Permutationally invariant fitting of many-body, non-covalent interactions with application to three-body methane--water--water , \ @noop journal journal Journal of Chemical Theory and Computati...

    2015

  59. [67]

    van der Oord , author G

    author author C. van der Oord , author G. Dusson , author G. Csányi , \ and\ author C. Ortner ,\ title title Regularised atomic body-ordered permutation-invariant polynomials for the construction of interatomic potentials , \ 10.1088/2632-2153/ab527c journal journal Machine Le...

    doi

  60. [68]

    author author A. E. \ Allen , author G. Dusson , author C. Ortner , \ and\ author G. Cs \'a nyi ,\ title title Atomic permutationally invariant polynomials for fitting molecular force fields , \ @noop journal journal Machine Learning: Science and Technology \ volume 2 ,\ pages...

    2021

  61. [69]

    Uteva , author R

    author author E. Uteva , author R. S. \ Graham , author R. D. \ Wilkinson , \ and\ author R. J. \ Wheatley ,\ title title Interpolation of intermolecular potentials using gaussian processes , \ @noop journal journal The Journal of Chemical Physics \ volume 147 ,\ pages 161706 ...

    2017

  62. [70]

    Balan , author N

    author author R. Balan , author N. Haghani , \ and\ author M. Singh ,\ @noop title Permutation invariant representations with applications to graph deep learning , \ ( year 2022 ),\ http://arxiv.org/abs/2203.07546 arXiv:2203.07546 NoStop

    2022 arXiv

  63. [71]

    author author A. S. \ Abbott , author J. M. \ Turney , author B. Zhang , author D. G. A. \ Smith , author D. Altarawy , \ and\ author H. F. I. \ Schaefer ,\ title title PES-Learn : An open-source software package for the automated generation of machine learning models of molec...

    doi

  64. [72]

    author author J. P. \ Heindel , author Q. Yu , author J. M. \ Bowman , \ and\ author S. S. \ Xantheas ,\ title title Benchmark electronic structure calculations for H3O+(H2O) _n , n = 0–5 , clusters and tests of an existing 1,2,3-body potential energy surface with a new 4-body...

    doi

  65. [73]

    Nandi , author R

    author author A. Nandi , author R. Conte , author C. Qu , author P. L. \ Houston , author Q. Yu , \ and\ author J. M. \ Bowman ,\ title title Quantum calculations on a new CCSD(T) machine-learned potential energy surface reveal the leaky nature of gas-phase trans and gauche et...

    doi

  66. [74]

    author author J. L. \ Bin Jiang \ and\ author H. Guo ,\ title title Potential energy surfaces from high fidelity fitting of ab initio points: the permutation invariant polynomial - neural network approach , \ 10.1080/0144235X.2016.1200347 journal journal International Reviews ...

  67. [75]

    Jiang \ and\ author H

    author author B. Jiang \ and\ author H. Guo ,\ title title Permutation invariant polynomial neural network approach to fitting potential energy surfaces , \ @noop journal journal The Journal of Chemical Physics \ volume 139 ,\ pages 054112 ( year 2013 ) NoStop

    2013

  68. [76]

    Li , author J

    author author J. Li , author J. Chen , author D. H. \ Zhang , \ and\ author H. Guo ,\ title title Quantum and quasi-classical dynamics of the OH + CO -> H + CO2 reaction on a new permutationally invariant neural network potential energy surface , \ @noop journal journal The Jo...

    2014

  69. [77]

    Li , author B

    author author J. Li , author B. Jiang , \ and\ author H. Guo ,\ title title Permutation invariant polynomial neural network approach to fitting potential energy surfaces. II . four-atom systems , \ @noop journal journal The Journal of chemical physics \ volume 139 ,\ pages 204...

    2013

  70. [78]

    Li \ and\ author H

    author author J. Li \ and\ author H. Guo ,\ title title A new ab initio based global HOOH (1 ^3 A '' ) potential energy surface for the O(^3P) + H2O (X^ 1 A_ 1 ) <=> OH (X^ 2 ) + OH (X^ 2 ) reaction , \ @noop journal journal The Journal of Chemical Physics \ volume 138 ,\ page...

    2013

  71. [79]

    Li , author B

    author author J. Li , author B. Jiang , \ and\ author H. Guo ,\ title title Spin-orbit corrected full-dimensional potential energy surfaces for the two lowest-lying electronic states of FH2O and dynamics for the F + H2O -> HF + OH reaction , \ @noop journal journal The Journal...

    2013

  72. [80]

    Li \ and\ author H

    author author A. Li \ and\ author H. Guo ,\ title title A nine-dimensional ab initio global potential energy surface for the H2O+ + H2 -> H3O+ + H reaction , \ @noop journal journal The Journal of Chemical Physics \ volume 140 ,\ pages 224313 ( year 2014 a ) NoStop

    2014

  73. [81]

    Li \ and\ author H

    author author A. Li \ and\ author H. Guo ,\ title title A full-dimensional global potential energy surface of H3O+( a ^3A) for the OH+( X ^3 ^-) + H2( X ^1 _g^+) -> H(^2S) + H2O+( X ^2B1) reaction , \ @noop journal journal The Journal of Physical Chemistry A \ volume 118 ,\ pa...

    2014

  74. [82]

    Li \ and\ author H

    author author J. Li \ and\ author H. Guo ,\ title title A nine-dimensional global potential energy surface for NH4 (X^2 A_1) and kinetics studies on the H + NH3 <-> H2 + NH2 reaction , \ @noop journal journal Physical Chemistry Chemical Physics \ volume 16 ,\ pages 6753--6763 ...

    2014

  75. [83]

    Li \ and\ author H

    author author J. Li \ and\ author H. Guo ,\ title title Communication: An accurate full 15 dimensional permutationally invariant potential energy surface for the OH + CH4 -> H2O+ CH3 reaction , \ @noop journal journal The Journal of Chemical Physics \ volume 143 ,\ pages 22110...

    2015

  76. [84]

    Li , author H

    author author J. Li , author H. Song , \ and\ author H. Guo ,\ title title Insights into the bond-selective reaction of Cl + HOD (n_ OH ) -> HCl + OD , \ @noop journal journal Physical Chemistry Chemical Physics \ volume 17 ,\ pages 4259--4267 ( year 2015 ) NoStop

    2015

  77. [85]

    Li , author S

    author author J. Li , author S. Carter , author J. M. \ Bowman , author R. Dawes , author D. Xie , \ and\ author H. Guo ,\ title title High-level, first-principles, full-dimensional quantum calculation of the ro-vibrational spectrum of the simplest criegee intermediate (CH2OO)...

    2014

  78. [86]

    \ Yu , author S

    author author H.-G. \ Yu , author S. Ndengue , author J. Li , author R. Dawes , \ and\ author H. Guo ,\ title title Vibrational energy levels of the simplest Criegee intermediat (CH2OO) from full-dimensional Lanczos , MCTDH , and MULTIMODE calculations , \ @noop journal journa...

    2015

  79. [87]

    Li \ and\ author H

    author author J. Li \ and\ author H. Guo ,\ title title Full-dimensional potential energy surface and ro-vibrational levels of dioxirane , \ @noop journal journal The Journal of Physical Chemistry A \ volume 120 ,\ pages 2991--2998 ( year 2016 ) NoStop

    2016

  80. [88]

    Han , author A

    author author H. Han , author A. Li , \ and\ author H. Guo ,\ title title Toward spectroscopically accurate global ab initio potential energy surface for the acetylene-vinylidene isomerization , \ @noop journal journal The Journal of chemical physics \ volume 141 ,\ pages 2443...

    2014

  81. [89]

    Guo , author H

    author author L. Guo , author H. Han , author J. Ma , \ and\ author H. Guo ,\ title title Quantum dynamics of vinylidene photodetachment on an accurate global acetylene-vinylidene potential energy surface , \ @noop journal journal The Journal of Physical Chemistry A \ volume 1...

    2015

  82. [90]

    Jiang \ and\ author H

    author author B. Jiang \ and\ author H. Guo ,\ title title Permutation invariant polynomial neural network approach to fitting potential energy surfaces. III . molecule-surface interactions , \ @noop journal journal The Journal of chemical physics \ volume 141 ,\ pages 034109 ...

    2014

  83. [91]

    Manzhos \ and\ author T

    author author S. Manzhos \ and\ author T. J. \ Carrington ,\ title title Neural network potential energy surfaces for small molecules and reactions , \ 10.1021/acs.chemrev.0c00665 journal journal Chemical Reviews \ volume 121 ,\ pages 10187--10217 ( year 2021 ) NoStop

    doi

  84. [92]

    Guan , author H

    author author Y. Guan , author H. Guo , \ and\ author D. R. \ Yarkony ,\ title title Neural network based quasi-diabatic hamiltonians with symmetry adaptation and a correct description of conical intersections , \ @noop journal journal The Journal of Chemical Physics \ volume ...

    2019

  85. [93]

    Xie , author X

    author author C. Xie , author X. Zhu , author D. R. \ Yarkony , \ and\ author H. Guo ,\ title title Permutation invariant polynomial neural network approach to fitting potential energy surfaces. IV . coupled diabatic potential energy matrices , \ @noop journal journal The Jour...

    2018

  86. [94]

    Liu \ and\ author J

    author author Y. Liu \ and\ author J. Li ,\ title title Permutation-invariant-polynomial neural-network-based -machine learning approach: A case for the HO2 self-reaction and its dynamics study , \ @noop journal journal The Journal of Physical Chemistry Letters \ volume 13 ,\ ...

    2022

  87. [95]

    Qu , author Q

    author author C. Qu , author Q. Yu , author B. L. J. \ Van Hoozen , author J. M. \ Bowman , \ and\ author R. A. \ Vargas-Hern \'a ndez ,\ title title Assessing gaussian process regression and permutationally invariant polynomial approaches to represent high-dimensional potenti...

    doi

  88. [96]

    Moses \ and\ author V

    author author W. Moses \ and\ author V. Churavy ,\ title title Instead of rewriting foreign code for machine learning, automatically synthesize fast gradients , \ @noop journal journal Advances in neural information processing systems \ volume 33 ,\ pages 12472--12485 ( year 2...

    2020

  89. [97]

    author author W. S. \ Moses , author V. Churavy , author L. Paehler , author J. H \"u ckelheim , author S. H. K. \ Narayanan , author M. Schanen , \ and\ author J. Doerfert ,\ title title Reverse-mode automatic differentiation and optimization of gpu kernels via enzyme , \ in\...

    2021

  90. [98]

    author author W. S. \ Moses , author S. H. K. \ Narayanan , author L. Paehler , author V. Churavy , author M. Schanen , author J. H\" u ckelheim , author J. Doerfert , \ and\ author P. Hovland ,\ title title Scalable automatic differentiation of multiple parallel paradigms thr...

    2022

  91. [99]

    Xie \ and\ author J

    author author Z. Xie \ and\ author J. M. \ Bowman ,\ title title Permutationally invariant polynomial basis for molecular energy surface fitting via monomial symmetrization , \ @noop journal journal Journal of Chemical Theory and Computation \ volume 6 ,\ pages 26--34 ( year 2...

    2010

  92. [100]

    author author A. G. \ Baydin , author B. A. \ Pearlmutter , author A. A. \ Radul , \ and\ author J. M. \ Siskind ,\ title title Automatic differentiation in machine learning: a survey , \ @noop journal journal J. Mach. Learn. Res. \ volume 18 ,\ pages 5595–5637 ( year 2017 ) NoStop

    2017

  93. [101]

    Blondel , author Q

    author author M. Blondel , author Q. Berthet , author M. Cuturi , author R. Frostig , author S. Hoyer , author F. Llinares-L \'o pez , author F. Pedregosa , \ and\ author J.-P. \ Vert ,\ title title Efficient and modular implicit differentiation , \ @noop journal journal Advan...

    2022

  94. [102]

    author author P. O. \ Dral , author B. Hourahine , \ and\ author S. Grimme ,\ title title Modern semiempirical electronic structure methods , \ @noop journal journal The Journal of Chemical Physics \ volume 160 ,\ pages 040401 ( year 2024 ) NoStop

    2024

  95. [103]

    Tamayo-Mendoza , author C

    author author T. Tamayo-Mendoza , author C. Kreisbeck , author R. Lindh , \ and\ author A. Aspuru-Guzik ,\ title title Automatic differentiation in quantum chemistry with applications to fully variational Hartree--Fock , \ @noop journal journal ACS central science \ volume 4 ,...

    2018

  96. [104]

    author author C. W. \ Tan , author C. J. \ Pickard , \ and\ author W. C. \ Witt ,\ title title Automatic differentiation for orbital-free density functional theory , \ @noop journal journal The Journal of Chemical Physics \ volume 158 ,\ pages 124801 ( year 2023 ) NoStop

    2023

  97. [105]

    author author R. A. \ Vargas-Hern \'a ndez , author K. Jorner , author R. Pollice , \ and\ author A. Aspuru-Guzik ,\ title title Inverse molecular design and parameter optimization with H \"u ckel theory using automatic differentiation , \ @noop journal journal The Journal of ...

    2023

  98. [106]

    de Camargo , author R

    author author A. de Camargo , author R. T. Q. \ Chen , \ and\ author R. A. \ Vargas-Hern \'a ndez ,\ title title Leveraging normalizing flows for orbital-free density functional theory , \ @noop journal journal Machine Learning: Science and Technology \ volume 5 ,\ pages 03506...

    2024

  99. [107]

    Zhang , author C

    author author X. Zhang , author C. Li , author H.-Z. \ Ye , author T. C. \ Berkelbach , \ and\ author G. K. \ Chan ,\ title title Performant automatic differentiation of local coupled cluster theories: Response properties and ab initio molecular dynamics , \ @noop journal jour...

    2024

  100. [108]

    author author J. M. \ Arrazola , author S. Jahangiri , author A. Delgado , author J. Ceroni , author J. Izaac , author A. Sz \'a va , author U. Azad , author R. A. \ Lang , author Z. Niu , author O. Di Matteo , author R. Moyard , author J. Soni , author M. Schuld , author R. A...

    2021 arXiv

  101. [109]

    author author P. A. \ M Casares , author J. S. \ Baker , author M. Medvidovi \'c , author R. d. \ Reis , \ and\ author J. M. \ Arrazola ,\ title title GradDFT . a software library for machine learning enhanced density functional theory , \ @noop journal journal The Journal of ...

    2024

  102. [110]

    Dawid , author J

    author author A. Dawid , author J. Arnold , author B. Requena , author A. Gresch , author M. P odzie \'n , author K. Donatella , author K. A. \ Nicoli , author P. Stornati , author R. Koch , author M. B \"u ttner , author R. Oku a , author G. Mu \ n oz-Gil , author R. A. \ Var...

    2022

  103. [111]

    author author M. F. \ Kasim \ and\ author S. M. \ Vinko ,\ title title Learning the exchange-correlation functional from nature with fully differentiable density functional theory , \ @noop journal journal Physical Review Letters \ volume 127 ,\ pages 126403 ( year 2021 ) NoStop

    2021

  104. [112]

    Schmidt , author C

    author author J. Schmidt , author C. L. \ Benavides-Riveros , \ and\ author M. A. \ Marques ,\ title title Machine learning the physical nonlocal exchange--correlation functional of density-functional theory , \ @noop journal journal The journal of physical chemistry letters \...

    2019

  105. [113]

    author author R. A. \ Vargas-Hern \'a ndez , author R. T. \ Chen , author K. A. \ Jung , \ and\ author P. Brumer ,\ title title Fully differentiable optimization protocols for non-equilibrium steady states , \ @noop journal journal New Journal of Physics \ volume 23 ,\ pages 1...

    2021

  106. [114]

    Zhang \ and\ author G

    author author X. Zhang \ and\ author G. K. \ Chan ,\ title title Differentiable quantum chemistry with PySCF for molecules and materials at the mean-field level and beyond , \ @noop journal journal The Journal of Chemical Physics \ volume 157 ,\ pages 204801 ( year 2022 ) NoStop

    2022

  107. [115]

    author author C. E. \ Rasmussen \ and\ author C. K. I. \ Williams ,\ 10.7551/mitpress/3206.001.0001 title Gaussian Processes for Machine Learning \ ( publisher The MIT Press ,\ year 2005 ) NoStop

    2005 doi

  108. [116]

    note A mask vector is a binary vector used to selectively apply operations to specific elements of another vector or matrix. Stop

  109. [117]

    Garnett ,\ @noop title Bayesian optimization \ ( publisher Cambridge University Press ,\ year 2023 ) NoStop

    author author R. Garnett ,\ @noop title Bayesian optimization \ ( publisher Cambridge University Press ,\ year 2023 ) NoStop

    2023

  110. [118]

    Vargas-Hern \'a ndez , author Y

    author author R. Vargas-Hern \'a ndez , author Y. Guan , author D. Zhang , \ and\ author R. Krems ,\ title title Bayesian optimization for the inverse scattering problem in quantum reaction dynamics , \ @noop journal journal New Journal of Physics \ volume 21 ,\ pages 022001 (...

    2019

  111. [119]

    author author R. A. \ Vargas-Hern \'a ndez ,\ title title Bayesian optimization for calibrating and selecting hybrid-density functional models , \ @noop journal journal The Journal of Physical Chemistry A \ volume 124 ,\ pages 4053--4061 ( year 2020 ) NoStop

    2020

  112. [120]

    Kocer , author T

    author author E. Kocer , author T. W. \ Ko , \ and\ author J. Behler ,\ title title Neural network potentials: A concise overview of methods , \ @noop journal journal Annual review of physical chemistry \ volume 73 ,\ pages 163--186 ( year 2022 ) NoStop

    2022

  113. [121]

    Shao , author J

    author author K. Shao , author J. Chen , author Z. Zhao , \ and\ author D. H. \ Zhang ,\ title title Communication: Fitting potential energy surfaces with fundamental invariant neural network , \ @noop journal journal The Journal of Chemical Physics \ volume 145 ,\ pages 07110...

    2016

  114. [122]

    van der Wilk , author M

    author author M. van der Wilk , author M. Bauer , author S. John , \ and\ author J. Hensman ,\ title title Learning invariances using the marginal likelihood , \ in\ https://proceedings.neurips.cc/paper_files/paper/2018/file/d465f14a648b3d0a1faa6f447e526c60-Paper.pdf booktitle...

    2018

  115. [123]

    author author V. L. \ Deringer , author A. P. \ Bart \'o k , author N. Bernstein , author D. M. \ Wilkins , author M. Ceriotti , \ and\ author G. Cs \'a nyi ,\ title title Gaussian process regression for materials and molecules , \ @noop journal journal Chemical Reviews \ volu...

    2021

  116. [124]

    Cui , author Z

    author author J. Cui , author Z. Li , \ and\ author R. V. \ Krems ,\ title title Gaussian process model for extrapolation of scattering observables for complex molecules: From benzene to benzonitrile , \ @noop journal journal The Journal of chemical physics \ volume 143 ,\ pag...

    2015

  117. [125]

    author author R. A. \ Vargas-Hern \'a ndez \ and\ author R. V. \ Krems ,\ title title Physical extrapolation of quantum observables by generalization with gaussian processes , \ @noop journal journal Machine Learning Meets Quantum Physics \ ,\ pages 171--194 ( year 2020 ) NoStop

    2020

  118. [126]

    Sugisawa , author T

    author author H. Sugisawa , author T. Ida , \ and\ author R. V. \ Krems ,\ title title Gaussian process model of 51-dimensional potential energy surface for protonated imidazole dimer , \ @noop journal journal The Journal of Chemical Physics \ volume 153 ,\ pages 114101 ( year...

    2020

  119. [127]

    Kolb , author P

    author author B. Kolb , author P. Marshall , author B. Zhao , author B. Jiang , \ and\ author H. Guo ,\ title title Representing global reactive potential energy surfaces using gaussian processes , \ 10.1021/acs.jpca.7b01182 journal journal The Journal of Physical Chemistry A ...

    doi

  120. [128]

    Denzel \ and\ author J

    author author A. Denzel \ and\ author J. Kästner ,\ title title Gaussian process regression for geometry optimization , \ 10.1063/1.5017103 journal journal The Journal of Chemical Physics \ volume 148 ,\ pages 094114 ( year 2018 ) NoStop

    doi

  121. [129]

    Cui \ and\ author R

    author author J. Cui \ and\ author R. V. \ Krems ,\ title title Efficient non-parametric fitting of potential energy surfaces for polyatomic molecules with gaussian processes , \ @noop journal journal Journal of Physics B: Atomic, Molecular and Optical Physics \ volume 49 ,\ p...

    2016

  122. [130]

    Christianen , author T

    author author A. Christianen , author T. Karman , author R. A. \ Vargas-Hern \'a ndez , author G. C. \ Groenenboom , \ and\ author R. V. \ Krems ,\ title title Six-dimensional potential energy surface for NaK-NaK collisions: Gaussian process representation with correct asympto...

    2019

  123. [131]

    Dai \ and\ author R

    author author J. Dai \ and\ author R. V. \ Krems ,\ title title Interpolation and extrapolation of global potential energy surfaces for polyatomic systems by gaussian processes with composite kernels , \ @noop journal journal Journal of Chemical Theory and Computation \ volume...

    2020

  124. [132]

    Dai \ and\ author R

    author author J. Dai \ and\ author R. V. \ Krems ,\ title title Quantum gaussian process model of potential energy surface for a polyatomic molecule , \ @noop journal journal The Journal of Chemical Physics \ volume 156 ,\ pages 184802 ( year 2022 ) NoStop

    2022

  125. [133]

    Dai \ and\ author R

    author author J. Dai \ and\ author R. V. \ Krems ,\ title title Neural network gaussian processes as efficient models of potential energy surfaces for polyatomic molecules , \ @noop journal journal Machine Learning: Science and Technology \ volume 4 ,\ pages 045027 ( year 2023...

    2023

  126. [134]

    Laude , author D

    author author G. Laude , author D. Calderini , author D. P. \ Tew , \ and\ author J. O. \ Richardson ,\ title title Ab initio instanton rate theory made efficient using gaussian process regression , \ 10.1039/C8FD00085A journal journal Faraday Discuss. \ volume 212 ,\ pages 23...

    doi

  127. [135]

    Kamath , author R

    author author A. Kamath , author R. A. \ Vargas-Hern \'a ndez , author R. V. \ Krems , author T. Carrington , \ and\ author S. Manzhos ,\ title title Neural networks vs gaussian process regression for representing potential energy surfaces: A comparative study of fit quality a...

    2018

  128. [136]

    author author R. A. \ Vargas-Hern \'a ndez \ and\ author J. R. \ Gardner ,\ title title Gaussian processes with spectral delta kernel for higher accurate potential energy surfaces for large molecules , \ @noop journal journal arXiv:2109.14074 \ ( year 2021 ) NoStop

    2021 arXiv

  129. [137]

    author author S. Y. \ Yafu Guan \ and\ author D. H. \ Zhang ,\ title title Construction of reactive potential energy surfaces with gaussian process regression: active data selection , \ @noop journal journal Molecular Physics \ volume 116 ,\ pages 823--834 ( year 2018 ) NoStop

    2018

  130. [138]

    author author A. G. \ Wilson , author Z. Hu , author R. Salakhutdinov , \ and\ author E. P. \ Xing ,\ title title Deep kernel learning , \ in\ https://proceedings.mlr.press/v51/wilson16.html booktitle Proceedings of the 19th International Conference on Artificial Intelligence ...

    2016

  131. [139]

    Gardner , author G

    author author J. Gardner , author G. Pleiss , author K. Q. \ Weinberger , author D. Bindel , \ and\ author A. G. \ Wilson ,\ title title Gpytorch: Blackbox matrix-matrix gaussian process inference with gpu acceleration , \ @noop journal journal Advances in neural information p...

    2018

  132. [140]

    Pinder \ and\ author D

    author author T. Pinder \ and\ author D. Dodd ,\ title title GPJax : A gaussian process framework in jax , \ 10.21105/joss.04455 journal journal Journal of Open Source Software \ volume 7 ,\ pages 4455 ( year 2022 ) NoStop

    2022 doi

  133. [141]

    \ Griffiths , author L

    author author R.-R. \ Griffiths , author L. Klarner , author H. Moss , author A. Ravuri , author S. Truong , author Y. Du , author S. Stanton , author G. Tom , author B. Rankovic , author A. Jamasb , et al. ,\ title title GAUCHE : A library for Gaussian processes in chemistry ...

    2023

  134. [142]

    u tt , \ and\ author K.-R. \ M \

    author author S. Chmiela , author A. Tkatchenko , author H. E. \ Sauceda , author I. Poltavsky , author K. T. \ Sch \"u tt , \ and\ author K.-R. \ M \"u ller ,\ title title Machine learning of accurate energy-conserving molecular force fields , \ @noop journal journal Science ...

    2017

  135. [143]

    Asnaashari \ and\ author R

    author author K. Asnaashari \ and\ author R. V. \ Krems ,\ title title Gradient domain machine learning with composite kernels: improving the accuracy of pes and force fields for large molecules , \ @noop journal journal Machine Learning: Science and Technology \ volume 3 ,\ p...

    2021

  136. [144]

    Babuschkin , author K

    author author DeepMind , author I. Babuschkin , author K. Baumli , author A. Bell , author S. Bhupatiraju , author J. Bruce , author P. Buchlovsky , author D. Budden , author T. Cai , author A. Clark , author I. Danihelka , author A. Dedieu , author C. Fantacci , author J. God...

    2020

  137. [145]

    author author G. S. \ Jung , author J. Y. \ Choi , \ and\ author S. M. \ Lee ,\ title title Active learning of neural network potentials for rare events , \ @noop journal journal Digital Discovery \ volume 3 ,\ pages 514--527 ( year 2024 ) NoStop

    2024

  138. [146]

    author author T. D. \ Loeffler , author T. K. \ Patra , author H. Chan , author M. Cherukara , \ and\ author S. K. \ Sankaranarayanan ,\ title title Active learning the potential energy landscape for water clusters from sparse training data , \ @noop journal journal The Journa...

    2020

  139. [147]

    Qi , author T

    author author J. Qi , author T. W. \ Ko , author B. C. \ Wood , author T. A. \ Pham , \ and\ author S. P. \ Ong ,\ title title Robust training of machine learning interatomic potentials with dimensionality reduction and stratified sampling , \ @noop journal journal npj Computa...

    2024

  140. [148]

    Rader , author T

    author author J. Rader , author T. Lyons , \ and\ author P. Kidger ,\ title title Lineax: unified linear solves and linear least-squares in JAX and Equinox , \ @noop journal journal AI for science workshop at Neural Information Processing Systems 2023, arXiv:2311.17283 \ ( yea...

    2023 arXiv

  141. [149]

    Chmiela , author H

    author author S. Chmiela , author H. E. \ Sauceda , author K.-R. \ M \"u ller , \ and\ author A. Tkatchenko ,\ title title Towards exact molecular dynamics simulations with machine-learned force fields , \ 10.1038/s41467-018-06169-2 journal journal Nature Communications \ volu...

    doi

  142. [150]

    Chmiela , author H

    author author S. Chmiela , author H. E. \ Sauceda , author I. Poltavsky , author K.-R. \ M \"u ller , \ and\ author A. Tkatchenko ,\ title title sgdml: Constructing accurate and data efficient molecular force fields using machine learning , \ https://doi.org/10.1016/j.cpc.2019...

    doi

  143. [151]

    Paszke , author S

    author author A. Paszke , author S. Gross , author F. Massa , author A. Lerer , author J. Bradbury , author G. Chanan , author T. Killeen , author Z. Lin , author N. Gimelshein , author L. Antiga , author A. Desmaison , author A. K \"o pf , author E. Yang , author Z. DeVito , ...

    2019

  144. [152]

    El Kazdadi ,\ https://github.com/sarah-ek/faer-rs title faer-rs , \ howpublished https://github.com/sarah-ek/faer-rs NoStop

    author author S. El Kazdadi ,\ https://github.com/sarah-ek/faer-rs title faer-rs , \ howpublished https://github.com/sarah-ek/faer-rs NoStop

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.