mlsanityData to 7 Oct 2026

SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects

arXiv 2105.00304 ↗First posted 1 May 2021physics.chem-ph cs.LGcited by 103

Not on the map. Why?

Abstract (arXiv)

Machine-learned force fields (ML-FFs) combine the accuracy of ab initio methods with the efficiency of conventional force fields. However, current ML-FFs typically ignore electronic degrees of freedom, such as the total charge or spin state, and assume chemical locality, which is problematic when molecules have inconsistent electronic states, or when nonlocal effects play a significant role. This work introduces SpookyNet, a deep neural network for constructing ML-FFs with explicit treatment of electronic degrees of freedom and quantum nonlocality. Chemically meaningful inductive biases and analytical corrections built into the network architecture allow it to properly model physical limits. SpookyNet improves upon the current state-of-the-art (or achieves similar performance) on popular quantum chemistry data sets. Notably, it is able to generalize across chemical and conformational space and can leverage the learned chemical insights, e.g. by predicting unknown spin states, thus helping to close a further important remaining gap for today's machine learning models in quantum chemistry.

Key facts. Cited by 103 papers in the map (data through 7 Oct 2026). Builds on 29 of its 132 references that are in the map. First posted on arXiv 1 May 2021.

Builds on

29 of 132 references.

References · 29

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  4. 2017Neural Message Passing for Quantum Chemistrycited by 3,460
  5. 2016Gaussian Error Linear Units (GELUs)cited by 3,431
  6. 2016Identity Mappings in Deep Residual Networkscited by 1,960
  7. 2020Rethinking Attention with Performerscited by 1,584
  8. 2017Sigmoid-Weighted Linear Units for Neural Network Function Approximation in Reinforcement Learningcited by 935
  9. 2019On the Convergence of Adam and Beyondcited by 834
  10. 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 735
  11. 2020Directional Message Passing for Molecular Graphscited by 556
  12. 2021Equivariant message passing for the prediction of tensorial properties and molecular spectracited by 438
17 more
  1. 2019Unmasking Clever Hans Predictors and Assessing What Machines Really Learncited by 359
  2. 2017The Expressive Power of Neural Networks: A View from the Widthcited by 318
  3. 2019Cormorant: Covariant Molecular Neural Networkscited by 262
  4. 2015Highway Networkscited by 257
  5. 2016Benefits of depth in neural networkscited by 248
  6. 2017Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanicscited by 213
  7. 2020Machine Learning Force Fieldscited by 204
  8. 2019Ab-Initio Solution of the Many-Electron Schr"odinger Equation with Deep Neural Networkscited by 195
  9. 2015The Power of Depth for Feedforward Neural Networkscited by 181
  10. 2019Deep neural network solution of the electronic Schr"odinger equationcited by 124
  11. 2019Machine learning for molecular simulationcited by 109
  12. 2015On the Expressive Power of Deep Learning: A Tensor Analysiscited by 97
  13. 2019Unifying machine learning and quantum chemistry -- a deep neural network for molecular wavefunctionscited by 73
  14. 2020OrbNet: Deep Learning for Quantum Chemistry Using Symmetry-Adapted Atomic-Orbital Featurescited by 63
  15. 2017Hierarchical modeling of molecular energies using a deep neural networkcited by 37
  16. 2017Machine Learning Molecular Dynamics for the Simulation of Infrared Spectracited by 30
  17. 2017WACSF - Weighted Atom-Centered Symmetry Functions as Descriptors in Machine Learning Potentialscited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 9; 2022: 19; 2023: 18; 2024: 18; 2025: 23; 2026: 16.

103 papers.

Sort:

By area:

  1. 2022Graph neural networks for materials science and chemistrycited by 181
  2. 2021Geometric Deep Learning on Molecular Representationscited by 113
  3. 2022Forces are not Enough: Benchmark and Critical Evaluation for Machine Learning Force Fields with Molecular Simulationscited by 92
  4. 2022SPICE, A Dataset of Drug-like Molecules and Peptides for Training Machine Learning Potentialscited by 80
  5. 2025Learning Smooth and Expressive Interatomic Potentials for Physical Property Predictioncited by 79
  6. 2023FAENet: Frame Averaging Equivariant GNN for Materials Modelingcited by 72
  7. 2021SE(3)-equivariant prediction of molecular wavefunctions and electronic densitiescited by 67
  8. 2022GemNet-OC: Developing Graph Neural Networks for Large and Diverse Molecular Simulation Datasetscited by 61
  9. 2021Software for Dataset-wide XAI: From Local Explanations to Global Insights with Zennit, CoRelAy, and ViRelAycited by 53
  10. 2023Ewald-based Long-Range Message Passing for Molecular Graphscited by 46
  11. 2023A new perspective on building efficient and expressive 3D equivariant graph neural networkscited by 46
  12. 2023OpenMM 8: Molecular Dynamics Simulation with Machine Learning Potentialscited by 44
38 more
  1. 2022Structure-based drug design with geometric deep learningcited by 40
  2. 2022So3krates: Equivariant attention for interactions on arbitrary length-scales in molecular systemscited by 39
  3. 2022Machine Learning Coarse-Grained Potentials of Protein Thermodynamicscited by 37
  4. 2021Graph Neural Networks Accelerated Molecular Dynamicscited by 33
  5. 2023Scaling the leading accuracy of deep equivariant models to biomolecular simulations of realistic sizecited by 33
  6. 2021Toward Explainable AI for Regression Modelscited by 30
  7. 2022Diffeomorphic Counterfactuals with Generative Modelscited by 29
  8. 2022SchNetPack 2.0: A neural network toolbox for atomistic machine learningcited by 29
  9. 2024Latent Ewald summation for machine learning of long-range interactionscited by 26
  10. 2022Simulate Time-integrated Coarse-grained Molecular Dynamics with Multi-Scale Graph Networkscited by 23
  11. 2023Navigating protein landscapes with a machine-learned transferable coarse-grained modelcited by 22
  12. 2022Equivariant Graph Attention Networks for Molecular Property Predictioncited by 19
  13. 2024TorchMD-Net 2.0: Fast Neural Network Potentials for Molecular Simulationscited by 15
  14. 2023CHGNet: Pretrained universal neural network potential for charge-informed atomistic modelingcited by 13
  15. 2024A predictive machine learning force field framework for liquid electrolyte developmentcited by 13
  16. 2025Force-Free Molecular Dynamics Through Autoregressive Equivariant Networkscited by 13
  17. 2024Euclidean Fast Attention -- Machine Learning Global Atomic Representations at Linear Costcited by 12
  18. 2023Uncertainty Estimation for Molecules: Desiderata and Methodscited by 11
  19. 2024Predicting solvation free energies with an implicit solvent machine learning potentialcited by 10
  20. 2025How simple can you go? An off-the-shelf transformer approach to molecular dynamicscited by 10
  21. 2022Accurate Machine Learned Quantum-Mechanical Force Fields for Biomolecular Simulationscited by 9
  22. 2024Molecular relaxation by reverse diffusion with time step predictioncited by 9
  23. 2024Neural P3M: A Long-Range Interaction Modeling Enhancer for Geometric GNNscited by 9
  24. 2025ELECTRA: A Cartesian Network for 3D Charge Density Prediction with Floating Orbitalscited by 9
  25. 2022NNP/MM: Accelerating molecular dynamics simulations with machine learning potentials and molecular mechaniccited by 8
  26. 2021Excited state, non-adiabatic dynamics of large photoswitchable molecules using a chemically transferable machine learning potentialcited by 7
  27. 2022Algorithmic Differentiation for Automated Modeling of Machine Learned Force Fieldscited by 7
  28. 2022Transferable E(3) equivariant parameterization for Hamiltonian of molecules and solidscited by 7
  29. 20233DReact: Geometric deep learning for chemical reactionscited by 7
  30. 2024Nutmeg and SPICE: Models and Data for Biomolecular Machine Learningcited by 7
  31. 2024Towards Symbolic XAI -- Explanation Through Human Understandable Logical Relationships Between Featurescited by 7
  32. 2021Surrogate- and invariance-boosted contrastive learning for data-scarce applications in sciencecited by 6
  33. 2024FeNNol: an Efficient and Flexible Library for Building Force-field-enhanced Neural Network Potentialscited by 6
  34. 2024Equivariant Graph Network Approximations of High-Degree Polynomials for Force Field Predictioncited by 6
  35. 2021BIGDML: Towards Exact Machine Learning Force Fields for Materialscited by 5
  36. 2023Neural network with optimal neuron activation functions based on additive Gaussian process regressioncited by 5
  37. 2023Learning Interatomic Potentials at Multiple Scalescited by 5
  38. 2025Learning Non-Local Molecular Interactions via Equivariant Local Representations and Charge Equilibrationcited by 5

50 most cited of 103 All 103 (JSON)