pith. sign in

arxiv: 2605.24013 · v1 · pith:XQVS7KGSnew · submitted 2026-05-20 · ⚛️ physics.chem-ph

UniField: RBF-Guided Electron Density Fusion for Enhanced Molecular Representations

Wei Zhang , Kun Li , Jiameng Chen , Jiajun Yu , Yizhen Zheng , Duanhua Cao , Wenbin Hu This is my paper

Pith reviewed 2026-06-30 17:40 UTC · model grok-4.3

classification ⚛️ physics.chem-ph
keywords electron densitymolecular graphsmultimodal fusionSE(3)-equivariantquantum propertiesbenchmark datasetsRBF guidancecomputational chemistry
0
0 comments X

The pith

Fusing continuous electron density fields with discrete molecular graphs via RBF-guided equivariant fusion produces new state-of-the-art accuracy on quantum property benchmarks.

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

The paper claims that standard 3D molecular models limited to atomic skeletons miss the continuous electron density field that actually controls quantum behaviors such as delocalization and non-covalent interactions. To address this, the authors release the UniField-ED benchmark containing natively aligned graphs and high-fidelity ED point clouds from QM9-ED and QMugs-ED. They then present UniField, an SE(3)-equivariant multimodal network that uses radial basis function guidance to integrate the two modalities. This integration yields concrete gains, including 14.8 percent overall improvement over leading topology-only models on ED5-OE and 28.2 percent on frontier orbital tasks for drug-like molecules.

Core claim

UniField establishes that an SE(3)-equivariant multimodal architecture can intrinsically intertwine discrete topological graphs with continuous quantum electronic environments through RBF-guided fusion of ED point clouds, producing new state-of-the-art results across the ED5-OE, QM9-ED, and QMugs-ED benchmarks with measured gains of 14.8 percent over topology-only SOTA, 37.0 percent over pure-ED models, and 28.2 percent average precision on frontier orbitals.

What carries the argument

RBF-Guided Electron Density Fusion, the alignment and integration step that combines high-fidelity continuous ED point clouds with discrete graphs inside the SE(3)-equivariant network.

If this is right

  • Topology-only models encounter a performance ceiling on tasks involving long-range delocalization that multimodal ED fusion can exceed.
  • Predictions for complex drug-like molecules on QMugs-ED become more accurate, especially for frontier orbital properties.
  • A standardized, natively aligned benchmark now exists for testing future electron-density-enhanced molecular models.
  • Next-generation computational chemistry tools can incorporate continuous quantum fields without abandoning graph-based representations.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same fusion approach could be tested on experimental rather than computed density data to check whether gains persist outside simulated benchmarks.
  • Scaling the point-cloud representation to very large systems might reveal memory or alignment bottlenecks not visible in the current datasets.
  • Neighboring problems such as protein-ligand binding or materials property prediction could benefit from analogous continuous-field fusion if aligned data can be generated.

Load-bearing premise

The benchmark's ED point clouds accurately represent the continuous electron density field and remain properly aligned with the corresponding graphs so that fusion captures real physical effects rather than artifacts.

What would settle it

Training and evaluating a version of UniField on the same benchmarks but with the ED point clouds randomly permuted or replaced by noise, then observing that predictive performance drops to or below the level of a topology-only baseline, would falsify the value of the fusion.

Figures

Figures reproduced from arXiv: 2605.24013 by Duanhua Cao, Jiajun Yu, Jiameng Chen, Kun Li, Wei Zhang, Wenbin Hu, Yizhen Zheng.

Figure 1
Figure 1. Figure 1: The evolution of molecular representation paradigms and the introduction of UniField [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overview of UniField. It consists of an SE(3)-Equivariant GNN for the atomic skeleton [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
read the original abstract

Current 3D geometric molecular representations predominantly focus on discrete atomic skeletons, inherently overlooking the continuous electron density (ED) field that fundamentally governs microscopic quantum behaviors. Consequently, these purely topological models suffer from critical representational blind spots, particularly in capturing long-range electron delocalization and non-covalent interactions, imposing a severe theoretical ceiling on predicting complex quantum properties. To bridge this physical gap and standardize research in electron density-enhanced molecular learning, we first construct the large-scale UniField-ED Benchmark. Comprising the QM9-ED and QMugs-ED datasets, this benchmark provides natively aligned discrete graphs and high-fidelity ED point clouds. Building upon this data infrastructure, we introduce UniField, an SE(3)-equivariant multimodal architecture that intrinsically intertwines discrete topological graphs with continuous quantum electronic environments. Extensive empirical evaluations across all three benchmarks demonstrate that UniField establishes new state-of-the-art performance. Specifically, UniField achieves a 14.8% improvement in overall predictive performance against the leading topology-only SOTA on the ED5-OE benchmark, alongside a 37.0% performance gain over top pure-ED models. Furthermore, on the complex drug-like dataset QMugs-ED, it yields a striking 28.2% average precision improvement across frontier orbital properties. Alongside new SOTA results on QM9-ED, our method establishes a rigorous foundation for next-generation computational chemistry. Code and datasets are anonymously available at https://anonymous.4open.science/r/UniField-ED-5B1B.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript constructs the UniField-ED benchmark (QM9-ED and QMugs-ED) supplying natively aligned discrete molecular graphs and high-fidelity electron density point clouds. It proposes UniField, an SE(3)-equivariant multimodal model that fuses these graphs with continuous ED fields via RBF guidance, and reports new state-of-the-art results: 14.8% overall improvement versus the leading topology-only model on ED5-OE, 37% gain over top pure-ED models, and 28.2% average precision improvement on frontier orbital properties for the drug-like QMugs-ED set, plus new SOTA on QM9-ED.

Significance. If the ED point clouds prove to be high-fidelity samplings of the true continuous field and remain natively aligned with the graphs, the work could meaningfully raise the representational ceiling for properties governed by long-range delocalization and non-covalent interactions. The public release of code and datasets would further strengthen its utility for the field.

major comments (2)
  1. [Abstract] Abstract: the central SOTA claims (14.8% on ED5-OE versus topology SOTA; 28.2% on QMugs-ED frontier orbitals) rest on the premise that the supplied ED point clouds are both high-fidelity representations of the continuous quantum field and natively aligned to the discrete graphs. No quantitative validation (integrated density error versus grid DFT references, alignment error statistics, or sampling-density sensitivity) is supplied to rule out discretization artifacts or data leakage as alternative explanations for the reported gains.
  2. [Abstract] The manuscript provides no architecture diagram, loss formulation, or training protocol for the RBF-guided fusion mechanism, making it impossible to assess whether the multimodal gains are attributable to genuine ED incorporation or simply to increased model capacity.
minor comments (1)
  1. [Abstract] The title references RBF guidance, yet the abstract does not specify how radial basis functions are used to interpolate or condition the continuous ED field onto the discrete graph.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address the two major comments point by point below. Where the comments correctly identify gaps, we commit to revisions that will incorporate the requested elements.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central SOTA claims (14.8% on ED5-OE versus topology SOTA; 28.2% on QMugs-ED frontier orbitals) rest on the premise that the supplied ED point clouds are both high-fidelity representations of the continuous quantum field and natively aligned to the discrete graphs. No quantitative validation (integrated density error versus grid DFT references, alignment error statistics, or sampling-density sensitivity) is supplied to rule out discretization artifacts or data leakage as alternative explanations for the reported gains.

    Authors: We agree that quantitative validation of the ED point clouds is necessary to support the SOTA claims and exclude artifacts or leakage. In the revised manuscript we will add a dedicated subsection (Methods) reporting integrated density errors versus grid DFT references, alignment error statistics between graphs and point clouds, and sampling-density sensitivity ablations. These results will also appear in the supplement. revision: yes

  2. Referee: [Abstract] The manuscript provides no architecture diagram, loss formulation, or training protocol for the RBF-guided fusion mechanism, making it impossible to assess whether the multimodal gains are attributable to genuine ED incorporation or simply to increased model capacity.

    Authors: The full manuscript contains an architecture diagram (Figure 1), the RBF-guided fusion loss (Equation 4, Section 3.2), and the complete training protocol (Section 4.1). To ensure these elements are immediately visible and to directly address capacity concerns, we will insert a short overview paragraph describing the fusion mechanism into the abstract and add an explicit capacity-matched ablation discussion in the main text. revision: partial

Circularity Check

0 steps flagged

No circularity; empirical benchmark results are independent of any self-referential derivation

full rationale

The paper constructs new datasets (QM9-ED, QMugs-ED) and reports empirical performance gains of a multimodal model on those datasets. No equations, uniqueness theorems, or fitted parameters are presented that reduce a claimed prediction back to the input by construction. The central claims rest on benchmark numbers rather than any derivation chain, and the provided abstract and description contain no self-citation load-bearing steps or ansatz smuggling. This is the normal case of an empirical ML paper whose results are falsifiable against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Abstract provides no explicit free parameters or invented entities; relies on standard domain assumptions of SE(3) equivariance and the physical relevance of electron density fields.

axioms (2)
  • domain assumption SE(3)-equivariance is required for physically consistent molecular representations
    Invoked by the choice of architecture in the abstract.
  • domain assumption Electron density point clouds can be fused with discrete graphs to capture quantum behaviors missed by topology
    Central premise stated in the abstract.

pith-pipeline@v0.9.1-grok · 5821 in / 1251 out tokens · 44248 ms · 2026-06-30T17:40:53.156720+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

61 extracted references · 25 canonical work pages · 1 internal anchor

  1. [1]

    Electron density-enhanced molecular geometry learning

    Hongxin Xiang, Jun Xia, Xin Jin, Wenjie Du, Li Zeng, and Xiangxiang Zeng. Electron density-enhanced molecular geometry learning. InProceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, pages 7840–7848, 2025. doi: 10.24963/ijcai.2025/872

    work page doi:10.24963/ijcai.2025/872 2025

  2. [2]

    Msanchor: De novo molecular generation from mass spectrometry data with anchor-extended molecular scaffolds

    Xiaohan Qin, Chao Wang, Zhengyang Zhou, Linjiang Chen, Wenjie Du, and Yang Wang. Msanchor: De novo molecular generation from mass spectrometry data with anchor-extended molecular scaffolds. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 953–961, 2026

    2026

  3. [3]

    Collaborative expert llms guided multi-objective molecular optimization.arXiv preprint arXiv:2503.03503, 2025

    Jiajun Yu, Yizhen Zheng, Huan Yee Koh, Shirui Pan, Tianyue Wang, and Haishuai Wang. Collaborative expert llms guided multi-objective molecular optimization.arXiv preprint arXiv:2503.03503, 2025

    work page arXiv 2025

  4. [4]

    Large language models for drug discovery and development.Patterns, 6(10), 2025

    Yizhen Zheng, Huan Yee Koh, Jiaxin Ju, Madeleine Yang, Lauren T May, Geoffrey I Webb, Li Li, Shirui Pan, and George Church. Large language models for drug discovery and development.Patterns, 6(10), 2025

    2025

  5. [5]

    Graph-structured small molecule drug discovery through deep learning: Progress, challenges, and opportunities

    Kun Li, Yida Xiong, Hongzhi Zhang, Xiantao Cai, Jia Wu, Bo Du, and Wenbin Hu. Graph-structured small molecule drug discovery through deep learning: Progress, challenges, and opportunities. In2025 IEEE International Conference on Web Services (ICWS), pages 1033–1042, 2025. doi: 10.1109/ICWS67624. 2025.00135

    work page doi:10.1109/icws67624 2025

  6. [6]

    Can molecular evolution mechanism enhance molecular representation? InProceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 15108–15116, 2026

    Kun Li, Longtao Hu, Jiameng Chen, Hongzhi Zhang, Yida Xiong, Xiantao Cai, Wenbin Hu, and Jia Wu. Can molecular evolution mechanism enhance molecular representation? InProceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 15108–15116, 2026

    2026

  7. [7]

    Pcevo: Path-consistent molecular representation via virtual evolutionary

    Kun Li, Longtao Hu, Yida Xiong, Jiajun Yu, Hongzhi Zhang, Jiameng Chen, Xiantao Cai, Jia Wu, and Wenbin Hu. Pcevo: Path-consistent molecular representation via virtual evolutionary. InProceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-26. International Joint Conferences on Artificial Intelligence Organization...

    2026

  8. [8]

    A centrality-based graph learning framework

    Jiajun Yu, Zhihao Wu, Jielong Lu, Tianyue Wang, and Haishuai Wang. A centrality-based graph learning framework. InProceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, pages 3588–3596, 2025

    2025

  9. [9]

    Geometric deep learning on molecular representa- tions.Nature Machine Intelligence, 3(12):1023–1032, 2021

    Kenneth Atz, Francesca Grisoni, and Gisbert Schneider. Geometric deep learning on molecular representa- tions.Nature Machine Intelligence, 3(12):1023–1032, 2021. doi: 10.1038/s42256-021-00418-8

    work page doi:10.1038/s42256-021-00418-8 2021

  10. [10]

    Deep learning methods for molecular representation and property prediction.Drug Discovery Today, 27(12):103373, 2022

    Zhen Li, Mingjian Jiang, Shuang Wang, and Shugang Zhang. Deep learning methods for molecular representation and property prediction.Drug Discovery Today, 27(12):103373, 2022. doi: 10.1016/j.drudis. 2022.103373

    work page doi:10.1016/j.drudis 2022

  11. [11]

    Drugpilot: Llm-based parameterized reasoning agent for drug discovery.arXiv preprint arXiv:2505.13940, 2025

    Kun Li, Zhennan Wu, Shoupeng Wang, Jia Wu, Shirui Pan, and Wenbin Hu. Drugpilot: Llm-based parameterized reasoning agent for drug discovery.arXiv preprint arXiv:2505.13940, 2025

    work page arXiv 2025

  12. [12]

    Bsl: A unified and generalizable multitask learning platform for virtual drug discovery from design to synthesis.arXiv preprint arXiv:2508.01195, 2025

    Kun Li, Zhennan Wu, Yida Xiong, Hongzhi Zhang, Longtao Hu, Zhonglie Liu, Junqi Zeng, Wenjie Wu, Mukun Chen, Jiameng Chen, et al. Bsl: A unified and generalizable multitask learning platform for virtual drug discovery from design to synthesis.arXiv preprint arXiv:2508.01195, 2025

    work page arXiv 2025

  13. [13]

    Smiles, a chemical language and information system

    David Weininger. Smiles, a chemical language and information system. 1. introduction to methodology and encoding rules.Journal of Chemical Information and Computer Sciences, 28(1):31–36, 1988. doi: 10.1021/ci00057a005

    work page doi:10.1021/ci00057a005 1988

  14. [14]

    Heller, Alan McNaught, Igor Pletnev, Stephen Stein, and Dmitrii Tchekhovskoi

    Stephen R. Heller, Alan McNaught, Igor Pletnev, Stephen Stein, and Dmitrii Tchekhovskoi. Inchi, the iupac international chemical identifier.Journal of Cheminformatics, 7:23, 2015. doi: 10.1186/ s13321-015-0068-4

    2015

  15. [15]

    Self- referencing embedded strings (SELFIES): A 100% robust molecular string representation.Machine Learning: Science and Technology, 1(4):045024, 2020

    Mario Krenn, Florian Häse, AkshatKumar Nigam, Pascal Friederich, and Alán Aspuru-Guzik. Self- referencing embedded strings (SELFIES): A 100% robust molecular string representation.Machine Learning: Science and Technology, 1(4):045024, 2020. doi: 10.1088/2632-2153/aba947

    work page doi:10.1088/2632-2153/aba947 2020

  16. [16]

    Kernel readout for graph neural networks

    Jiajun Yu, Zhihao Wu, Jinyu Cai, Adele Lu Jia, and Jicong Fan. Kernel readout for graph neural networks. InIJCAI, pages 2505–2514, 2024

    2024

  17. [17]

    Gotennet: Rethinking efficient 3d equivariant graph neural networks

    Sarp Aykent and Tian Xia. Gotennet: Rethinking efficient 3d equivariant graph neural networks. InThe Thirteenth International Conference on Learning Representations, 2025. URL https://openreview. net/forum?id=5wxCQDtbMo. 10

    2025

  18. [18]

    Enhancing geometric representations for molecules with equivariant vector-scalar interactive message passing.Nature Communications, 15(1):313, 2024

    Yusong Wang, Tong Wang, Shaoning Li, Xinheng He, Mingyu Li, Zun Wang, Nanning Zheng, Bin Shao, and Tie-Yan Liu. Enhancing geometric representations for molecules with equivariant vector-scalar interactive message passing.Nature Communications, 15(1):313, 2024. doi: 10.1038/s41467-023-43720-2

    work page doi:10.1038/s41467-023-43720-2 2024

  19. [19]

    Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds

    Nathaniel Thomas, Tess Smidt, Steven Kearnes, Lusann Yang, Li Li, Kai Kohlhoff, and Patrick Riley. Tensor field networks: Rotation- and translation-equivariant neural networks for 3d point clouds.arXiv preprint arXiv:1802.08219, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  20. [20]

    Cormorant: Covariant molecular neural networks

    Brandon Anderson, Truong-Son Hy, and Risi Kondor. Cormorant: Covariant molecular neural networks. InAdvances in Neural Information Processing Systems, volume 32, 2019

    2019

  21. [21]

    E(n) equivariant graph neural networks

    Victor Garcia Satorras, Emiel Hoogeboom, and Max Welling. E(n) equivariant graph neural networks. In Proceedings of the 38th International Conference on Machine Learning, volume 139 ofProceedings of Machine Learning Research, pages 9323–9332, 2021

    2021

  22. [22]

    Wood, Abhishek Das, and Tess E

    Yi-Lun Liao, Brandon M. Wood, Abhishek Das, and Tess E. Smidt. Equiformerv2: Improved equivariant transformer for scaling to higher-degree representations. InThe Twelfth International Conference on Learning Representations, 2024. URLhttps://openreview.net/forum?id=mCOBKZmrzD

    2024

  23. [23]

    Inhomogeneous electron gas.Physical Review, 136(3B):B864–B871,

    Pierre Hohenberg and Walter Kohn. Inhomogeneous electron gas.Physical Review, 136(3B):B864–B871,

  24. [24]

    doi: 10.1103/PhysRev.136.B864

    work page doi:10.1103/physrev.136.b864

  25. [25]

    Kohn and L

    W. Kohn and L. J. Sham. Self-consistent equations including exchange and correlation effects.Physical Review, 140(4A):A1133–A1138, 1965. doi: 10.1103/PhysRev.140.A1133

    work page doi:10.1103/physrev.140.a1133 1965

  26. [26]

    Parr and Weitao Yang.Density-Functional Theory of Atoms and Molecules, volume 16

    Robert G. Parr and Weitao Yang.Density-Functional Theory of Atoms and Molecules, volume 16. Oxford University Press, New York, 1989

    1989

  27. [27]

    Richard F. W. Bader. A quantum theory of molecular structure and its applications.Chemical Reviews, 91 (5):893–928, 1991. doi: 10.1021/cr00005a013

    work page doi:10.1021/cr00005a013 1991

  28. [28]

    Johnson, Shahar Keinan, Paula Mori-Sánchez, Julia Contreras-García, Aron J

    Erin R. Johnson, Shahar Keinan, Paula Mori-Sánchez, Julia Contreras-García, Aron J. Cohen, and Weitao Yang. Revealing noncovalent interactions.Journal of the American Chemical Society, 132(18):6498–6506,

  29. [29]

    doi: 10.1021/ja100936w

    work page doi:10.1021/ja100936w

  30. [30]

    Manby, and Thomas F

    Zhuoran Qiao, Matthew Welborn, Animashree Anandkumar, Frederick R. Manby, and Thomas F. Miller III. Orbnet: Deep learning for quantum chemistry using symmetry-adapted atomic-orbital features.The Journal of Chemical Physics, 153(12):124111, 2020. doi: 10.1063/5.0021955

    work page doi:10.1063/5.0021955 2020

  31. [31]

    Deepdft: Neural message passing network for accurate charge density prediction.arXiv preprint arXiv:2011.03346, 2020

    Peter Bjørn Jørgensen and Arghya Bhowmik. Deepdft: Neural message passing network for accurate charge density prediction.arXiv preprint arXiv:2011.03346, 2020

    work page arXiv 2011

  32. [32]

    Equivariant graph neural networks for fast electron density estimation of molecules, liquids, and solids.npj Computational Materials, 8:183, 2022

    Peter Bjørn Jørgensen and Arghya Bhowmik. Equivariant graph neural networks for fast electron density estimation of molecules, liquids, and solids.npj Computational Materials, 8:183, 2022. doi: 10.1038/ s41524-022-00863-y

    2022

  33. [33]

    Machine learning accurate exchange and correlation function- als of the electronic density.Nature Communications, 11:3509, 2020

    Sebastian Dick and Marivi Fernandez-Serra. Machine learning accurate exchange and correlation function- als of the electronic density.Nature Communications, 11:3509, 2020. doi: 10.1038/s41467-020-17265-7

    work page doi:10.1038/s41467-020-17265-7 2020

  34. [34]

    Deep-learning density functional theory hamiltonian for efficient ab initio electronic-structure calculation

    He Li, Zun Wang, Nianlong Zou, Meng Ye, Runzhang Xu, Xiaoxun Gong, Wenhui Duan, and Yong Xu. Deep-learning density functional theory hamiltonian for efficient ab initio electronic-structure calculation. Nature Computational Science, 2(6):367–377, 2022. doi: 10.1038/s43588-022-00265-6

    work page doi:10.1038/s43588-022-00265-6 2022

  35. [35]

    Dral, Matthias Rupp, and O

    Raghunathan Ramakrishnan, Pavlo O. Dral, Matthias Rupp, and O. Anatole von Lilienfeld. Quantum chemistry structures and properties of 134 kilo molecules.Scientific Data, 1:140022, 2014. doi: 10.1038/ sdata.2014.22

    2014

  36. [36]

    Qmugs, quantum mechanical properties of drug-like molecules.Scientific Data, 9(1):273, 2022

    Clemens Isert, Kenneth Atz, José Jiménez-Luna, and Gisbert Schneider. Qmugs, quantum mechanical properties of drug-like molecules.Scientific Data, 9(1):273, 2022. doi: 10.1038/s41597-022-01390-7

    work page doi:10.1038/s41597-022-01390-7 2022

  37. [37]

    Edbench: Large-scale electron density data for molecular modeling

    Hongxin Xiang, Ke Li, Mingquan Liu, Zhixiang Cheng, Bin Yao, Wenjie Du, Jun Xia, Li Zeng, Xin Jin, and Xiangxiang Zeng. Edbench: Large-scale electron density data for molecular modeling. InAdvances in Neural Information Processing Systems, 2025. Datasets and Benchmarks Track

    2025

  38. [38]

    Contrastive learning-based drug screening model for glun1/glun3a inhibitors.Acta Pharmacologica Sinica, pages 1–13, 2025

    Kun Li, Yue Zeng, Yi-da Xiong, Hao-chen Wu, Sui Fang, Zhi-yan Qu, Yan Zhu, Bo Du, Zhao-bing Gao, and Wen-bin Hu. Contrastive learning-based drug screening model for glun1/glun3a inhibitors.Acta Pharmacologica Sinica, pages 1–13, 2025. 11

    2025

  39. [39]

    Cohen, Paula Mori-Sánchez, and Weitao Yang

    Aron J. Cohen, Paula Mori-Sánchez, and Weitao Yang. Challenges for density functional theory.Chemical Reviews, 112(1):289–320, 2012

    2012

  40. [40]

    Perspective on density functional theory.The Journal of Chemical Physics, 136(15):150901, 2012

    Kieron Burke. Perspective on density functional theory.The Journal of Chemical Physics, 136(15):150901, 2012

    2012

  41. [41]

    Tanja Van Mourik, Michael Bühl, and Marie-Pierre Gaigeot. Density functional theory across chemistry, physics and biology.Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 372(2011):20120488, 2014

    2011

  42. [42]

    Nobel lecture: Electronic structure of matter—wave functions and density functionals

    Walter Kohn. Nobel lecture: Electronic structure of matter—wave functions and density functionals. Reviews of Modern Physics, 71(5):1253–1266, 1999

    1999

  43. [43]

    Schoenholz, Patrick F

    Justin Gilmer, Samuel S. Schoenholz, Patrick F. Riley, Oriol Vinyals, and George E. Dahl. Neural message passing for quantum chemistry. InProceedings of the 34th International Conference on Machine Learning, volume 70 ofProceedings of Machine Learning Research, pages 1263–1272, 2017

    2017

  44. [44]

    Convolutional networks on graphs for learning molecular fingerprints

    David K Duvenaud, Dougal Maclaurin, Jorge Iparraguirre, Rafael Bombarell, Timothy Hirzel, Alan Aspuru-Guzik, and Ryan P Adams. Convolutional networks on graphs for learning molecular fingerprints. In C. Cortes, N. Lawrence, D. Lee, M. Sugiyama, and R. Garnett, editors,Advances in Neural Information Processing Systems, volume 28. Curran Associates, Inc., 2...

    2015

  45. [45]

    Schütt, Pieter-Jan Kindermans, Huziel E

    Kristof T. Schütt, Pieter-Jan Kindermans, Huziel E. Sauceda, Stefan Chmiela, Alexandre Tkatchenko, and Klaus-Robert Müller. Schnet: A continuous-filter convolutional neural network for modeling quantum interactions. InAdvances in Neural Information Processing Systems, volume 30, 2017

    2017

  46. [46]

    Unke and Markus Meuwly

    Oliver T. Unke and Markus Meuwly. Physnet: A neural network for predicting energies, forces, dipole moments, and partial charges.Journal of Chemical Theory and Computation, 15(6):3678–3693, 2019. doi: 10.1021/acs.jctc.9b00181

    work page doi:10.1021/acs.jctc.9b00181 2019

  47. [47]

    Margraf, and Stephan Günnemann

    Johannes Gasteiger, Shankari Giri, Johannes T. Margraf, and Stephan Günnemann. Fast and uncertainty- aware directional message passing for non-equilibrium molecules. InMachine Learning for Molecules Workshop, NeurIPS, 2020

    2020

  48. [48]

    Spherical message passing for 3d molecular graphs

    Yi Liu, Limei Wang, Meng Liu, Yuchao Lin, Xuan Zhang, Bora Oztekin, and Shuiwang Ji. Spherical message passing for 3d molecular graphs. InInternational Conference on Learning Representations (ICLR), 2022. URLhttps://openreview.net/forum?id=givsRXsOt9r

    2022

  49. [49]

    Mailoa, Mordechai Kornbluth, Nicola Molinari, Tess E

    Simon Batzner, Albert Musaelian, Lixin Sun, Mario Geiger, Jonathan P. Mailoa, Mordechai Kornbluth, Nicola Molinari, Tess E. Smidt, and Boris Kozinsky. E(3)-equivariant graph neural networks for data- efficient and accurate interatomic potentials.Nature Communications, 13(1):2453, 2022. doi: 10.1038/ s41467-022-29939-5

    2022

  50. [50]

    Schütt, Oliver T

    Kristof T. Schütt, Oliver T. Unke, and Michael Gastegger. Equivariant message passing for the prediction of tensorial properties and molecular spectra. InProceedings of the 38th International Conference on Machine Learning, volume 139 ofProceedings of Machine Learning Research, pages 9377–9388, 2021

    2021

  51. [51]

    Qi, Hao Su, Kaichun Mo, and Leonidas J

    Charles R. Qi, Hao Su, Kaichun Mo, and Leonidas J. Guibas. Pointnet: Deep learning on point sets for 3d classification and segmentation. InProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 77–85, 2017

    2017

  52. [52]

    Qi, Li Yi, Hao Su, and Leonidas J

    Charles R. Qi, Li Yi, Hao Su, and Leonidas J. Guibas. Pointnet++: Deep hierarchical feature learning on point sets in a metric space. InAdvances in Neural Information Processing Systems, volume 30, 2017

    2017

  53. [53]

    Sarma, Michael M

    Yue Wang, Yongbin Sun, Ziwei Liu, Sanjay E. Sarma, Michael M. Bronstein, and Justin M. Solomon. Dynamic graph cnn for learning on point clouds.ACM Transactions on Graphics, 38(5):146, 2019. doi: 10.1145/3326362

    work page doi:10.1145/3326362 2019

  54. [54]

    Hengshuang Zhao, Li Jiang, Jiaya Jia, Philip H. S. Torr, and Vladlen Koltun. Point transformer. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 16259–16268, 2021

    2021

  55. [55]

    Pointnext: Revisiting pointnet++ with improved training and scaling strategies

    Guocheng Qian, Yuchen Li, Houwen Peng, Jinjie Mai, Hasan Hammoud, Mohamed Elhoseiny, and Bernard Ghanem. Pointnext: Revisiting pointnet++ with improved training and scaling strategies. InAdvances in Neural Information Processing Systems (NeurIPS), 2022

    2022

  56. [56]

    Pointvector: A vector representation in point cloud analysis

    Xin Deng, WenYu Zhang, Qing Ding, and XinMing Zhang. Pointvector: A vector representation in point cloud analysis. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 9455–9465, 2023. 12

    2023

  57. [57]

    Ismail Fawaz, H., Forestier, G., Weber, J., Idoumghar, L., and Muller, P.-A

    Xiaoyang Wu, Li Jiang, Peng-Shuai Wang, Zhijian Liu, Xihui Liu, Yu Qiao, Wanli Ouyang, Tong He, and Hengshuang Zhao. Point transformer v3: Simpler, faster, stronger. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 4840–4851, 2024. doi: 10.1109/CVPR52733.2024.00463

    work page doi:10.1109/cvpr52733.2024.00463 2024

  58. [58]

    Psi4 1.4: Open-source software for high-throughput quantum chemistry.The Journal of Chemical Physics, 152(18), 2020

    Daniel GA Smith, Lori A Burns, Andrew C Simmonett, Robert M Parrish, Matthew C Schieber, Raimondas Galvelis, Peter Kraus, Holger Kruse, Roberto Di Remigio, Asem Alenaizan, et al. Psi4 1.4: Open-source software for high-throughput quantum chemistry.The Journal of Chemical Physics, 152(18), 2020

    2020

  59. [59]

    Gomez, Łukasz Kaiser, and Illia Polosukhin

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention Is All You Need. InAdvances in Neural Information Processing Systems, volume 30, pages 5998–6008, 2017

    2017

  60. [60]

    Comenet: Towards complete and efficient message passing for 3d molecular graphs

    Limei Wang, Yi Liu, Yuchao Lin, Haoran Liu, and Shuiwang Ji. Comenet: Towards complete and efficient message passing for 3d molecular graphs. InAdvances in Neural Information Processing Systems, volume 35, pages 650–664, 2022. URL https://proceedings.neurips.cc/paper_files/paper/ 2022/hash/0418973e545b932939302cb605d06f43-Abstract-Conference.html

    work page arXiv 2022

  61. [61]

    Equiformer: Equivariant graph attention transformer for 3d atomistic graphs

    Yi-Lun Liao and Tess Smidt. Equiformer: Equivariant graph attention transformer for 3d atomistic graphs. InInternational Conference on Learning Representations, 2023. URL https://openreview.net/ forum?id=KwmPfARgOTD. A Background and Theoretical Foundations A.1 Electron Density as a Molecular Quantum Field A.1.1 Definition of Electron Density Electron den...

    work page arXiv 2023