Pith.

Chemical Physics · arXiv:2611.11019 v1 · submitted 7 Nov 2026 · 51 references

Toward universal excited-state machine-learned potentials: a local excitation Hamiltonian for size consistency and transferability

Pascal Friederich

The paper says

“Universal machine-learned interatomic potentials could make excited-state simulations across chemical space routine, but excitation energies, unlike ground-state energies, are not sums of atomic contributions.”

more of the abstract

The excited states of noninteracting molecules are the union of their states, which pooled multistate readouts reproduce at best for the lowest state. We introduce LEH-MACE, whose excitation energies are the eigenvalues of a local excitation Hamiltonian assembled from atom-centred blocks, so this union holds exactly. We generated TD-DFT energies and forces for about 300,000 equilibrium, thermal, and excited-state geometries of over 100,000 molecules. LEH-MACE's advantage over direct readouts grew with the distance from the training data and was largest above S1, with about half their S2 and S3 errors for long conjugated oligomers. For molecule pairs, only LEH-MACE kept its single-molecule accuracy, while a softmin readout remained superior for S1 dynamics of larger chromophores. Exact size consistency is thus one step toward such potentials.

from the abstract

Pith says

Machine review · deepseek-v4-flash · 10 Nov 2026 · what was read was not recorded · not checked by a person

It can be wrong. Challenge this review or reply to it.

Holds up. LEH-MACE delivers exact size consistency for excited-state MLIPs via a local Hamiltonian readout, with thorough evaluation and honest limitations; the main trade-off is that direct softmin readouts still beat it for S1 dynamics of individual large chromophores. This line is wrong

2 major objections · 4 minor

How often objections from reviews like this one are right: not yet measured

Are you an author? What you can do.

Not checked: figures, code, what the 51 references say.

The paper says, at Table 1 and Fig. 3

The review did not quote this passage.

A classifier read this sentence of the paper as acknowledging the concern (probability 0.88):

“The benchmark models contain three of the seven validation molecules (styrene, benzaldehyde, and cinnamaldehyde) in their training set, and the sampling models of rounds 3 and 4, fine-tuned from the dye-augmented LEH model, additionally azobenzene (Methods), so the criterion values of these models are slightly optimistic.”

Pith says: Objection 1 of 2 · major

The benchmark models (LEH, Direct-softmin, X-MACE) include three validation molecules (styrene, benzaldehyde, cinnamaldehyde) in their training set, as acknowledged in Supplementary Note 14. This is a mild contamination of the validation set, though the main test sets are unaffected. The authors should discuss the potential impact on the reported validation errors and the active-learning stopping criteria, which rely on these validation trajectories. The paper already notes this, but a brief quantitative statement of the effect would strengthen the report.

The paper says, at Methods: Weighted loss decision

The review did not quote this passage.

Pith says: Objection 2 of 2 · major

The choice of the weighted loss uses validation data (mean signed S1 error below 2 eV must decrease by at least 30%, validation S1–S3 error and trajectory error must not increase by more than 5%). The authors apply this rule uniformly across architectures.

more of objection 2

While the procedure is transparent, using the same validation data both to design the loss and to monitor progress is a mild degree of optimism. The paper acknowledges this, but a note about potential overfitting to the validation set would be appropriate. The results on the held-out test trajectories (not used for any decision) largely confirm the trends, so the concern is limited.

4 minor objections

The paper says, at Abstract

Pith says: minor objection 1 of 4

The phrase 'with about half their S2 and S3 errors for long conjugated oligomers' could be clarified: 'about half the errors of the direct readouts for S2 and S3' would avoid ambiguity.

The paper says, at Fig. 4a

Pith says: minor objection 2 of 4

The x-axis labels include 'Dyes', 'Ladders', 'Dimers'. Consider adding a note that the 'Dimers' error is for both different and identical molecules, since the text discusses both cases.

show 2 more

The paper says, at Methods: Local excitation Hamiltonian

Pith says: minor objection 3 of 4

Equation (2) shows Hij with a symmetrization. It may be worth noting explicitly that this ensures the matrix is symmetric, which is already stated but could be highlighted for clarity.

The paper says, at Supplementary Note 2

Pith says: minor objection 4 of 4

Table S2 is very comprehensive. Consider adding a column for the number of molecules in each test set to help readers gauge statistical significance.

What would settle it

Train LEH and a direct-softmin baseline on a dataset that includes a few donor–acceptor complexes where low-lying charge-transfer states exist within the cutoff. Test both models on a held-out set of such complexes. If LEH fails to reproduce the CT states (expected, given the model's scope), this confirms the acknowledged limitation; if it unexpectedly succeeds, the scope of size consistency might be broader than claimed.

from the skeptic's reading

How the review went

4 readings by deepseek-v4-flash · 10 Nov 2026

  1. Reader holds up confidence high
  2. Skeptic agrees 22 s · 40,955 tokens read, 2,181 written
  3. Editor wrote the verdict above 13 s · 41,676 tokens read, 1,291 written
  4. Referee accept: 2 major objections, 4 minor 19 s · 41,122 tokens read, 1,840 written

Not run on this paper: the simulated authors' defence.

Pith did not check this review's objections against the paper.

The review in full

The review in brief, machine-written

This paper is a genuine step forward. The core idea — assembling a local excitation Hamiltonian from atom-centred blocks whose eigenvalues are the excitation energies — is mathematically clean and directly solves the size-consistency problem that all multistate pooled readouts have. The proof is trivial once you see it, but nobody had built it into a working transferable MLIP before. The empirical validation is thorough: large dataset (300k geometries, 100k+ molecules), multiple baselines (X-MACE, OMNI-P2x, MS-ANI, and their own direct readouts on the same encoder), and systematic tests on random molecules, unseen scaffolds, larger molecules, conjugated dyes, ladder molecules, and dimers. The advantage of LEH grows exactly where a universal model needs it most — far from the training distribution. The authors also ran an active-learning loop to improve coverage of twisted geometries, and they release code, data, and trained models. The limitations are stated with unusual candour: no charge-transfer states, no double excitations, no long-range couplings, and the direct softmin readout is still more accurate for S1 dynamics of single large chromophores. That last point is the main practical trade-off: for photodynamics of individual molecules, LEH is not yet the best choice. But for spectra, for states above S1, and for multiple molecules, it is. The paper's one soft spot is that the model's systematic overestimation of S1 for large conjugated molecules is not fully explained — the authors say they don't know why the eigenvalue readout learns this lowering less well. That's an honest admission, not a flaw in the evaluation. The results are reproducible: the math is exact, the data are public, and the trained models are released. This paper deserves a serious referee. I would bring it to reading group and would cite it for the architecture and dataset.

The strongest objection

The central claim—that LEH achieves exact size consistency by construction—is mathematically proven (Eq. 4) and empirically validated with dimer spectra matching monomer spectra to 10⁻¹⁴ eV. The paper acknowledges the crucial scope condition: this holds only for local single excitations within the cutoff radius; charge-transfer states, double excitations, and long-range couplings are not represented. Extensive comparisons against multiple baselines on varied test sets systematically support that LEH's advantage grows with distance from the training distribution. No internal inconsistency or hidden assumption undermines the core argument. The study is thorough, the code and data are released, and limitations are honestly stated. I find no load-bearing concern beyond those already noted by the authors and the Reader.

The Reader's weakest assumption concerns the local representation's ability to capture the relevant excited-state physics. The paper acknowledges this limitation and provides evidence that the model works well for the tested classes of molecules (closed-shell organics with C, H, N, O, F). No additional critical flaw emerged. The central claim (exact size consistency) is mathematically clean and empirically supported. Therefore my assessment does not change the verdict.

The referee's report

The paper addresses a fundamental challenge in excited-state machine-learned potentials: size consistency. The LEH architecture is a principled solution that guarantees the union-of-spectra property exactly, which no previous readout achieves. The empirical results convincingly show that this property translates to improved transferability to larger molecules, conjugated systems, and dimers, especially for higher excited states. The dataset and open-source code are valuable resources. The work is a clear step toward universal excited-state MLIPs and will likely influence subsequent research.

Pith's reading of the paper (a machine's summary, not the authors')

The paper introduces LEH-MACE, a machine-learned potential for excited states whose excitation energies are the eigenvalues of a local excitation Hamiltonian assembled from atom-centered blocks. This construction guarantees exact size consistency: for noninteracting molecules the predicted spectrum is exactly the union of the molecules' spectra. The approach is tested on a new dataset of 300,000 TD-DFT geometries and shown to transfer better than standard multistate readouts to larger molecules, conjugated dyes, ladder molecules, and molecule pairs, especially for states above the first excited state. The authors argue that exact size consistency is a necessary step toward universal excited-state MLIPs.

What carries the argument. The local excitation Hamiltonian (LEH) is a symmetric matrix built from atom-centred blocks: a diagonal onsite block from each atom's MACE features and off-diagonal coupling blocks for atom pairs within a 5 Å cutoff. Its sorted eigenvalues are the singlet excitation energies. For two fragments beyond the cutoff the matrix is block diagonal, so its spectrum is the union of the fragments' spectra by construction. This gives exact size consistency for local single excitations.

Core claim. Only a local excitation Hamiltonian readout is exactly size consistent for excited-state MLIPs: the spectrum of noninteracting fragments is the multiset union of their spectra. This property, which readouts based on summing, averaging, or soft-min pooling do not share, improves transferability to larger molecules, conjugated systems, and dimers, with the largest advantage for higher-lying excited states. LEH-MACE achieves half the errors of the softmin readout for S2 and S3 of long conjugated oligomers and retains single-molecule accuracy for dimers, whereas all other readouts degrade by factors of three to four.

The premise it rests on. The representation is limited to local single excitations within a 5 Å cutoff; charge-transfer states, double excitations, and long-range excitonic couplings are not captured by the model.

If the paper is right (Pith's inferences, not the authors')

LEH can be used to predict absorption spectra and photophysical processes involving higher-lying excited states with transferable accuracy.

The readout is well suited for simulations of chromophores in explicit solvent, where large solvation shells contain many noninteracting molecules.

Active learning on excited-state dynamics selects geometries that standard thermal sampling misses; combining it with a size-consistent readout may extend the reach of excited-state MLIPs to nanosecond photodynamics.

The open dataset and code provide a benchmark for developing size-consistent excited-state models and for testing extensions to charge-transfer and long-range couplings.

LEH could be combined with ground-state foundation models (e.g., MACE-MP-0) to create a universal excited-state potential that inherits ground-state transferability.

The failure to learn S1 lowering for large chromophores suggests that multiple latent channels per atom or longer-range couplings are needed; these are testable extensions that the paper's framework supports.

The exact size consistency property could be exploited in fragment-based methods to predict spectra of molecular clusters from monomer models without retraining.

The paper's dataset, covering many geometries per molecule, enables studying whether excited-state MLIPs benefit more from more molecules or more geometries per molecule; the authors find they are equally valuable.

Check this review

Pith stored no record of this review's passes and took no fingerprint of it. Reviews made by code from 13 Nov 2026 on carry both.

How often reviews like this one are right

Reviews like this one: written by deepseek-v4-flash, where what was read was not recorded.

How often objections from reviews like this one are right: not yet measured

Every value above, in the review's public record (JSON)

Reviews by people

This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.

Request a human review

How Pith checks papers · The full audit of this paper

The paper

Abstract

Universal machine-learned interatomic potentials could make excited-state simulations across chemical space routine, but excitation energies, unlike ground-state energies, are not sums of atomic contributions. The excited states of noninteracting molecules are the union of their states, which pooled multistate readouts reproduce at best for the lowest state. We introduce LEH-MACE, whose excitation energies are the eigenvalues of a local excitation Hamiltonian assembled from atom-centred blocks, so this union holds exactly. We generated TD-DFT energies and forces for about 300,000 equilibrium, thermal, and excited-state geometries of over 100,000 molecules. LEH-MACE's advantage over direct readouts grew with the distance from the training data and was largest above S1, with about half their S2 and S3 errors for long conjugated oligomers. For molecule pairs, only LEH-MACE kept its single-molecule accuracy, while a softmin readout remained superior for S1 dynamics of larger chromophores. Exact size consistency is thus one step toward such potentials.

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.

abstract compiled · 2026-11-10

The authors' main claim, as the classifier picked it out 0.90

“Exact size consistency is thus one step toward such potentials.”

How the main claim is stated: asserted 0.78

What the abstract itself states, as probabilities: claims first 0.08 · states limitation 0.83 · human participants 0.02 · controlled comparison 0.85 · randomized assignment 0.01 · states quantitative result 0.96 · compares with prior results 0.96 · mentions code or data availability 0.04

The abstract, sentence by sentence

This table labels the abstract only (arXiv listing abstract v1). The number is the probability the classifier gave the label.

SentenceRoleThe authors' own workStates the main conclusionStrength
Universal machine-learned interatomic potentials could make excited-state simulations across chemical space routine, but excitation energies, unlike ground-state energies, are not sums of atomic contributions. background 0.88 0.06 0.05 hedged possibility 0.81
The excited states of noninteracting molecules are the union of their states, which pooled multistate readouts reproduce at best for the lowest state. background 0.81 0.15 0.04 asserted 0.91
We introduce LEH-MACE, whose excitation energies are the eigenvalues of a local excitation Hamiltonian assembled from atom-centred blocks, so this union holds exactly. contribution 0.53 0.93 0.47 asserted 1.00
We generated TD-DFT energies and forces for about 300,000 equilibrium, thermal, and excited-state geometries of over 100,000 molecules. method 0.96 0.94 0.03 asserted 1.00
LEH-MACE's advantage over direct readouts grew with the distance from the training data and was largest above $S_1$, with about half their $S_2$ and $S_3$ errors for long conjugated oligomers. result 1.00 0.93 0.14 asserted 1.00
For molecule pairs, only LEH-MACE kept its single-molecule accuracy, while a softmin readout remained superior for $S_1$ dynamics of larger chromophores. result 0.98 0.90 0.10 asserted 1.00
Exact size consistency is thus one step toward such potentials. interpretation 0.99 0.27 0.51 asserted 0.78

What kind of paper the classifier reports this is: method or tool 1.00 · evidence from simulation or computation 1.00

What was examined, and what was not

Sources read

  • arXiv listing abstract v1

Not examined

  • full text
  • figures and tables
  • references and citing sentences
  • relations to other papers

Levels run

  • P abstract

Not run

  • S sections
  • T body sentences
  • C citations
  • R cross-paper relations

How the labels were made

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-11-10: 1 request, 5,725 input tokens.

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON

Figures (6)
Figure 1 of the paper
Figure 1. Local excitation Hamiltonian and data. a, A MACE encoder computes atom features hi within the cutoff rc = 5 ˚A. Onsite blocks and pair blocks form a symmetric matrix H, whose sorted eigenvalues are the excitation energies S1, S2, . . . b, Two molecules A and B further apart than rc. H is block diagonal, and its spectrum is the union of the spectra of A and B. c, Generation of the training data. QCDGE-derived molecules at equilibrium and thermal geometries and geometries from model-driven S1 and S2 dynamics are labelled with TD-ωB97X￾D/6-31G* energies and forces (S0–S3, ten singlet…
Figure 2 of the paper
Figure 2. Dataset. a, Number of molecules versus the number of heavy atoms for the QCDGE￾derived training molecules, the larger test molecules, the dyes (training and test scaffolds, the training scaffolds are part of the production training set only), and the ladder molecules (test only). b, Distribution of the S1 excitation energy of the production training geometries by geometry type (number of geometries in brackets). c, Distribution of the S0 energy above the minimum of the same molecule. d, Mean absolute S1 error of the sampling models of active-learning rounds 2–4 on the geometries…
Figure 3 of the paper
Figure 3. Learning curves. Mean absolute error of the S1–S3 excitation energies (a, b) and of the S1 forces (c, d) versus the number of training geometries, for random test molecules with up to 10 heavy atoms at thermal geometries (a, c) and larger test molecules with 11–20 heavy atoms (b, d). The training sets contain 1k–76.6k molecules with about 2.2 geometries per molecule. Markers show seed means and bars seed standard deviations. LEH, L = 2: encoder with higher angular resolution, full training set only. OMNI-P2x arch.: the OMNI-P2x architecture trained from scratch on our data. MS-ANI…
Figure 4 of the paper
Figure 4. Transferability. a, Mean absolute error of the S1–S3 excitation energies of the models trained on the full training set, for test sets ordered by their distance from the train￾ing distribution. Shaded: test molecules similar to the training molecules. Dyes: all 3,875 sulfur-free dyes at their equilibrium geometries, none of which these models saw in training. Ladders: all sulfur-free ladder molecules at their equilibrium geometries. Dimers: two non￾bonded test molecules at minimum distances of 2.5–15 ˚A, compared with TD-ωB97X-D dimer spectra (
Figure 5 of the paper
Figure 5. Separability. a, Excitation energies (S1–S6) of two different molecules A and B at a minimum interatomic distance of 10 ˚A, beyond the cutoff rc = 5 ˚A. For TD-ωB97X-D and each model, the left column (A, B) shows the levels of A and B computed separately, coloured by molecule, and the right column (A+B) the levels of the pair computed together. Solid lines connect each level of the pair with the matching level of the separate molecules (TD-ωB97X-D and LEH, dimer levels coloured by the molecule they belong to). Dashed lines connect the n-th levels of the two columns (pooled readouts,…
Figure 6 of the paper
Figure 6. Excited-state dynamics. a, Mean absolute error of the S1 excitation energy and b, of the change of the S0 energy along the TD-ωB97X-D S1 trajectories of seven validation molecules, for the sampling models after each active-learning round (small markers: the two models, line: mean, round 1: benchmark LEH) and for the production models (small markers: four seeds, large markers: mean). The dashed line marks the 0.2 eV threshold of the stopping criterion. c, Mean absolute S1 error along the five test trajectories for the benchmark LEH before active learning and for the production LEH…

Figures are the authors', reproduced from the paper.

References (51)
  1. [1] Friederich, P.; H¨ ase, F.; Proppe, J.; Aspuru-Guzik, A. Machine-learned potentials for next-generation matter simulations.Nature Materials2021,20, 750–761, DOI: 10.1038/s41563-020-0777-6 on Pithdoi:10.1038/s41563-020-0777-6
  2. [2] Westermayr, J.; Marquetand, P. Machine Learning for Electronically Excited States of Molecules.Chemical Reviews2021,121, 9873–9926, DOI:10.1021/acs.chemrev.0c00749 on Pithdoi:10.1021/acs.chemrev.0c00749
  3. [3] Dral, P. O.; Barbatti, M. Molecular excited states through a machine learning lens.Nature Reviews Chemistry2021,5, 388–405, DOI:10.1038/s41570-021-00278-1 on Pithdoi:10.1038/s41570-021-00278-1
  4. [4] Li, J.; Lopez, S. A. A Look Inside the Black Box of Machine Learning Photo- dynamics Simulations.Accounts of Chemical Research2022,55, 1972–1984, DOI: 10.1021/acs.accounts.2c00288. 20 on Pithdoi:10.1021/acs.accounts.2c00288
  5. [5] Hu, D.; Xie, Y.; Li, X.; Li, L.; Lan, Z. Inclusion of Machine Learning Kernel Ridge Regression Potential Energy Surfaces in On-the-Fly Nonadiabatic Molecular Dynam- ics Simulation.The Journal of Physical Chemistry Letters2018,9, 2725–2732, DOI: 10.1021/acs.jpclett.8b00684 on Pithdoi:10.1021/acs.jpclett.8b00684
  6. [6] Dral, P. O.; Barbatti, M.; Thiel, W. Nonadiabatic Excited-State Dynamics with Ma- chine Learning.The Journal of Physical Chemistry Letters2018,9, 5660–5663, DOI: 10.1021/acs.jpclett.8b02469 on Pithdoi:10.1021/acs.jpclett.8b02469
  7. [7] Chen, W.-K.; Liu, X.-Y.; Fang, W.-H.; Dral, P. O.; Cui, G. Deep Learning for Nonadiabatic Excited-State Dynamics.The Journal of Physical Chemistry Letters2018,9, 6702–6708, DOI:10.1021/acs.jpclett.8b03026 on Pithdoi:10.1021/acs.jpclett.8b03026
  8. [8] Westermayr, J.; Gastegger, M.; Menger, M. F. S. J.; Mai, S.; Gonz´ alez, L.; Marquetand, P. Machine learning enables long time scale molecular photodynamics simulations.Chemical Science2019,10, 8100–8107, DOI:10.1039/c9sc01742a on Pithdoi:10.1039/c9sc01742a
  9. [9] Li, J.; Reiser, P.; Boswell, B. R.; Eberhard, A.; Burns, N. Z.; Friederich, P.; Lopez, S. A. Automatic discovery of photoisomerization mechanisms with nanosecond ma- chine learning photodynamics simulations.Chemical Science2021,12, 5302–5314, DOI: 10.1039/d0sc05610c on Pithdoi:10.1039/d0sc05610c
  10. [10] Li, J.; Stein, R.; Adrion, D. M.; Lopez, S. A. Machine-Learning Photodynamics Sim- ulations Uncover the Role of Substituent Effects on the Photochemical Formation of Cubanes.Journal of the American Chemical Society2021,143, 20166–20175, DOI: 10.1021/jacs.1c07725 on Pithdoi:10.1021/jacs.1c07725
  11. [11] Axelrod, S.; Shakhnovich, E.; G´ omez-Bombarelli, R. Excited state non-adiabatic dynam- ics of large photoswitchable molecules using a chemically transferable machine learning potential.Nature Communications2022,13, 3440, DOI:10.1038/s41467-022-30999-w on Pithdoi:10.1038/s41467-022-30999-w
  12. [12] Axelrod, S.; Shakhnovich, E.; G´ omez-Bombarelli, R. Thermal Half-Lives of Azobenzene Derivatives: Virtual Screening Based on Intersystem Crossing Using a Machine Learning Potential.ACS Central Science2023,9, 166–176, DOI:10.1021/acscentsci.2c00897 on Pithdoi:10.1021/acscentsci.2c00897
  13. [13] Kumar, P.; Zhou, C.; Friederich, P.; Escudero, D. In silico study of degradation in host molecular materials for OLEDs.Advanced Theory and Simulations2026,9, e70584, DOI: 10.1002/adts.70584 on Pithdoi:10.1002/adts.70584
  14. [14] Westermayr, J.; Gastegger, M.; Marquetand, P. Combining SchNet and SHARC: The SchNarc Machine Learning Approach for Excited-State Dynamics.The Journal of Physical Chemistry Letters2020,11, 3828–3834, DOI:10.1021/acs.jpclett.0c00527 on Pithdoi:10.1021/acs.jpclett.0c00527
  15. [15] Mausenberger, S.; M¨ uller, C.; Tkatchenko, A.; Marquetand, P.; Gonz´ alez, L.; Wester- mayr, J. SPaiNN: equivariant message passing for excited-state nonadiabatic molecular dynamics.Chemical Science2024,15, 15880–15890, DOI:10.1039/d4sc04164j on Pithdoi:10.1039/d4sc04164j
  16. [16] Westermayr, J.; Marquetand, P. Deep learning for UV absorption spectra with SchNarc: First steps toward transferability in chemical compound space.The Journal of Chemical Physics2020,153, 154112, DOI:10.1063/5.0021915 on Pithdoi:10.1063/5.0021915
  17. [17] Martyka, M.; Zhang, L.; Ge, F.; Hou, Y.-F.; Jankowska, J.; Barbatti, M.; Dral, P. O. Charting electronic-state manifolds across molecules with multi-state learning and gap- driven dynamics via efficient and robust active learning.npj Computational Materials2025, 11, 132, DOI:1... on Pithdoi:10.1038/s41524-025-01636-z
  18. [18] Barrett, R.; Ortner, C.; Westermayr, J. Transferable Machine Learning Potential X-MACE for Excited States using Integrated DeepSets. 2025; arXiv:2502.12870 on PithPith's reviewarXiv
  19. [19] Martyka, M.; Tong, X.-Y.; Jankowska, J.; Dral, P. O. OMNI-P2x universal neural net- work potential for excited-state simulations.Nature Communications2026,17, 4949, DOI: 10.1038/s41467-026-71380-5 on Pithdoi:10.1038/s41467-026-71380-5
  20. [20] Malosso, C.; How, W. B.; D ´ ıaz Mir´ on, G.; Hassanali, A.; Ceriotti, M. Transferable machine learning of excited-state dynamics with extremal pooling. 2026; arXiv:2606.16859 on PithPith's reviewarXiv
show 31 more
  1. [21] Westermayr, J.; Maurer, R. J. Physically inspired deep learning of molecular exci- tations and photoemission spectra.Chemical Science2021,12, 10755–10764, DOI: 10.1039/d1sc01542g on Pithdoi:10.1039/d1sc01542g
  2. [22] Cignoni, E.; Suman, D.; Nigam, J.; Cupellini, L.; Mennucci, B.; Ceriotti, M. Electronic Excited States from Physically Constrained Machine Learning.ACS Cent. Sci.2024,10, 637–648, DOI:10.1021/acscentsci.3c01480 on Pithdoi:10.1021/acscentsci.3c01480
  3. [23] Ren, F.; Chen, X.; Liu, F. Size-Transferable Prediction of Excited State Properties for Molecular Assemblies with a Machine Learning Exciton Model.The Journal of Physical Chemistry Letters2025,16, 2541–2552, DOI:10.1021/acs.jpclett.4c03548 on Pithdoi:10.1021/acs.jpclett.4c03548
  4. [24] Juergens, D.; St¨ ohr, M.; Hillers-Bendtsen, A. E.; Fajen, O. J.; Mart ´ ınez, T. J. Latent unified smooth Hamiltonians for excited state chemistry. 2026; arXiv:2609.01871 on PithPith's reviewarXiv
  5. [25] Ramakrishnan, R.; Hartmann, M.; Tapavicza, E.; von Lilienfeld, O. A. Electronic spectra from TDDFT and machine learning in chemical space.The Journal of Chemical Physics 2015,143, 084111, DOI:10.1063/1.4928757 on Pithdoi:10.1063/1.4928757
  6. [26] Zhu, Y.; Li, M.; Xu, C.; Lan, Z. Quantum chemistry dataset with ground- and excited-state properties of 450 kilo molecules.Scientific Data2024,11, 948, DOI: 10.1038/s41597-024-03788-x on Pithdoi:10.1038/s41597-024-03788-x
  7. [27] Curth, R.; R¨ ohrkasten, T. E.; M¨ uller, C.; Westermayr, J. Surface Hopping Nested In- stances Training Set for Excited-state Learning.Scientific Data2025,12, 1300, DOI: 10.1038/s41597-025-05443-5 on Pithdoi:10.1038/s41597-025-05443-5
  8. [28] Pinheiro Jr, M.; Zhang, S.; Dral, P. O.; Barbatti, M. WS22 database, Wigner sampling and geometry interpolation for configurationally diverse molecular datasets.Scientific Data 2023,10, 95, DOI:10.1038/s41597-023-01998-3 on Pithdoi:10.1038/s41597-023-01998-3
  9. [29] Batatia, I.; Kov´ acs, D. P.; Simm, G. N. C.; Ortner, C.; Cs´ anyi, G. MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields. Advances in Neural Information Processing Systems. 2022; pp 11423–11436, DOI: 10.48550/arXiv.2206.07697, ar... on PithPith's reviewarXivdoi:10.48550/arxiv.2206.07697
  10. [30] Zhou, C.; Neubert, M.; Koide, Y.; Zhang, Y.; Vuong, V.-Q.; Schl¨ oder, T.; Dehnen, S.; Friederich, P. PAL – parallel active learning for machine-learned potentials.Digital Discov- ery2025,4, 1901–1911, DOI:10.1039/d5dd00073d on Pithdoi:10.1039/d5dd00073d
  11. [31] Jacquemin, D.; Wathelet, V.; Perp` ete, E. A.; Adamo, C. Extensive TD-DFT benchmark: singlet-excited states of organic molecules.Journal of Chemical Theory and Computation 2009,5, 2420–2435, DOI:10.1021/ct900298e on Pithdoi:10.1021/ct900298e
  12. [32] Liang, J.; Feng, X.; Hait, D.; Head-Gordon, M. Revisiting the performance of time- dependent density functional theory for electronic excitations: assessment of 43 popular and recently developed functionals from rungs one to four.Journal of Chemical Theory and Computation2022,... on Pithdoi:10.1021/acs.jctc.2c00160
  13. [33] Loos, P.-F.; Scemama, A.; Jacquemin, D. The quest for highly accurate excitation energies: a computational perspective.The Journal of Physical Chemistry Letters2020,11, 2374– 2383, DOI:10.1021/acs.jpclett.0c00014 on Pithdoi:10.1021/acs.jpclett.0c00014
  14. [34] He, Y.; Escudero, D. Beyond the three-state picture: when higher-lying excited states become quantitatively indispensable.Chemical Science2026,17, 10967–10981, DOI: 10.1039/d6sc01726f on Pithdoi:10.1039/d6sc01726f
  15. [35] Levine, B. G.; Ko, C.; Quenneville, J.; Mart ´ ınez, T. J. Conical intersections and dou- ble excitations in time-dependent density functional theory.Molecular Physics2006,104, 1039–1051, DOI:10.1080/00268970500417762 on Pithdoi:10.1080/00268970500417762
  16. [36] Batatia, I. et al. A foundation model for atomistic materials chemistry.The Journal of Chemical Physics2025,163, 184110, DOI:10.1063/5.0297006 on Pithdoi:10.1063/5.0297006
  17. [37] Levine, D. S. et al. The Open Molecules 2025 (OMol25) Dataset, Evaluations, and Models. 2025; arXiv:2505.08762 on PitharXiv
  18. [38] Smith, J. S.; Nebgen, B. T.; Zubatyuk, R.; Lubbers, N.; Devereux, C.; Barros, K.; Tre- tiak, S.; Isayev, O.; Roitberg, A. E. Approaching coupled cluster accuracy with a general- purpose neural network potential through transfer learning.Nature Communications2019, 10, 2903, DOI... on Pithdoi:10.1038/s41467-019-10827-4
  19. [39] Balasubramani, S. G. et al. TURBOMOLE: Modular program suite forab initioquantum- chemical and condensed-matter simulations.The Journal of Chemical Physics2020,152, 184107, DOI:10.1063/5.0004635 on Pithdoi:10.1063/5.0004635
  20. [40] Chai, J.-D.; Head-Gordon, M. Long-range corrected hybrid density functionals with damped atom–atom dispersion corrections.Phys. Chem. Chem. Phys.2008,10, 6615–6620, DOI: 10.1039/B810189B on Pithdoi:10.1039/b810189b
  21. [41] Hariharan, P. C.; Pople, J. A. The influence of polarization functions on molecu- lar orbital hydrogenation energies.Theoretica Chimica Acta1973,28, 213–222, DOI: 10.1007/BF00533485 on Pithdoi:10.1007/bf00533485
  22. [42] Furche, F.; Ahlrichs, R. Adiabatic time-dependent density functional methods for ex- cited state properties.The Journal of Chemical Physics2002,117, 7433–7447, DOI: 10.1063/1.1508368 on Pithdoi:10.1063/1.1508368
  23. [43] Bannwarth, C.; Ehlert, S.; Grimme, S. GFN2-xTB—An Accurate and Broadly Parametrized Self-Consistent Tight-Binding Quantum Chemical Method with Multipole Electrostatics and Density-Dependent Dispersion Contributions.Journal of Chemical The- ory and Computation2019,15, 1652–1671... on Pithdoi:10.1021/acs.jctc.8b01176
  24. [45] G´ omez-Bombarelli, R.; Wei, J. N.; Duvenaud, D.; Hern´ andez-Lobato, J. M.; S´ anchez- Lengeling, B.; Sheberla, D.; Aguilera-Iparraguirre, J.; Hirzel, T. D.; Adams, R. P.; Aspuru- Guzik, A. Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules.A... on Pithdoi:10.1021/acscentsci.7b00572
  25. [46] Irwin, J. J.; Shoichet, B. K. ZINC – A Free Database of Commercially Available Compounds for Virtual Screening.J. Chem. Inf. Model.2005,45, 177–182, DOI:10.1021/ci049714+ on Pithdoi:10.1021/ci049714
  26. [47] RDKit contributors RDKit: Open-source cheminformatics.https://www.rdkit.org, 2026. 23 on Pith
  27. [50] Dral, P. O. et al. MLatom 3: A Platform for Machine Learning-Enhanced Computational Chemistry Simulations and Workflows.Journal of Chemical Theory and Computation2024, 20, 1193–1213, DOI:10.1021/acs.jctc.3c01203. 24 Supplementary Information Toward universal excited-state mach... on Pithdoi:10.1021/acs.jctc.3c01203
  28. [51] Bemis, G. W.; Murcko, M. A. The properties of known drugs. 1. Molecular frameworks. Journal of Medicinal Chemistry 1996, 39, 2887–2893, DOI: 10.1021/jm9602928 on Pithdoi:10.1021/jm9602928
  29. [52] Joung, J. F.; Han, M.; Jeong, M.; Park, S. Experimental database of optical properties of organic compounds. Scientific Data 2020, 7, 295, DOI: 10.1038/s41597-020-00634-8 on Pithdoi:10.1038/s41597-020-00634-8
  30. [53] Beard, E. J.; Sivaraman, G.; V´ azquez-Mayagoitia,´A.; Vishwanath, V.; Cole, J. M. Compar- ative dataset of experimental and computational attributes of UV/vis absorption spectra. Scientific Data 2019, 6, 307, DOI: 10.1038/s41597-019-0306-0 on Pithdoi:10.1038/s41597-019-0306-0
  31. [54] Plasser, F.; Lischka, H. Analysis of Excitonic and Charge Transfer Interactions from Quan- 13 Supplementary Table 11: Production models (weighted loss) on the validation split and all test sets and along the TD- ωB97X-D S1 trajectories (frames before TD-DFT breaks down). Seeds... on Pith

Challenges and discussion

Disagree with the review, or know something it missed? Leave a comment, or challenge this review formally: a challenge becomes a permanent entry in this paper's record, and its decision is published here.

No comments yet.

Sign in to comment