Pith. sign in

REVIEW 14 cited by

TorchMD-NET: Equivariant Transformers for Neural Network based Molecular Potentials

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2202.02541 v2 pith:WYDHYZ35 submitted 2022-02-05 cs.LG cs.AIphysics.chem-ph

TorchMD-NET: Equivariant Transformers for Neural Network based Molecular Potentials

classification cs.LG cs.AIphysics.chem-ph
keywords accuracymolecularpotentialscomputationalconformationsefficiencyequivarianttorchmd-net
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

The prediction of quantum mechanical properties is historically plagued by a trade-off between accuracy and speed. Machine learning potentials have previously shown great success in this domain, reaching increasingly better accuracy while maintaining computational efficiency comparable with classical force fields. In this work we propose TorchMD-NET, a novel equivariant transformer (ET) architecture, outperforming state-of-the-art on MD17, ANI-1, and many QM9 targets in both accuracy and computational efficiency. Through an extensive attention weight analysis, we gain valuable insights into the black box predictor and show differences in the learned representation of conformers versus conformations sampled from molecular dynamics or normal modes. Furthermore, we highlight the importance of datasets including off-equilibrium conformations for the evaluation of molecular potentials.

discussion (0)

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

Forward citations

Cited by 14 Pith papers

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

  1. Quotient-Space Diffusion Models

    cs.LG 2026-04 unverdicted novelty 8.0

    Quotient-space diffusion models generate correct symmetric distributions by removing redundancy on the quotient space, simplifying learning and improving results on small molecules and proteins under SE(3) symmetry.

  2. Geometric Algebra Meets Cartesian Tensors: Higher-Order Equivariance for Interatomic Potentials

    physics.chem-ph 2026-06 unverdicted novelty 7.0

    CliffordSTF couples Clifford multivectors to rank-2 and rank-3 symmetric-traceless tensor tracks through bilinear cross-track contractions, lifting force cosine similarity from 0.055 to 0.551 on rMD17 while outperform...

  3. DPA4: Pushing the Accuracy-Cost Frontier of Interatomic Potentials with EMFA SO(2) Convolution

    physics.chem-ph 2026-06 unverdicted novelty 7.0

    DPA4 is a new SE(3)-equivariant interatomic potential with EMFA SO(2) convolution that sets new accuracy-cost records on Matbench Discovery and SPICE benchmarks using fewer parameters than prior models.

  4. h-MINT: Modeling Pocket-Ligand Binding with Hierarchical Molecular Interaction Network

    cs.LG 2026-04 unverdicted novelty 7.0

    h-MINT improves ligand-protein binding affinity prediction by 2-4% and virtual screening metrics by 1-3% via overlapping fragment tokenization and hierarchical modeling.

  5. Align Your Structures: Generating Trajectories with Structure Pretraining for Molecular Dynamics

    cs.LG 2026-04 conditional novelty 6.5

    Structure-pretrained diffusion plus an equivariant temporal interpolator generates chemically realistic MD trajectories on small molecules, tetrapeptides, and proteins by separating spatial and temporal learning.

  6. Transformer Atomic Cluster Expansion: TRACE

    cond-mat.mtrl-sci 2026-07 conditional novelty 6.0

    A no-message-passing, attention-based local potential reproduces a perovskite phase transition, liquid-water O–O structure, and an organic rearrangement barrier with a single architecture.

  7. Streamlining Analysis and Design of Two-Dimensional Electronic Spectroscopy using Machine Learning

    physics.chem-ph 2026-06 unverdicted novelty 6.0

    A Gaussian mixture model is used to learn spectral densities from 2DES experiments, enabling extraction of vibronic couplings, spectral extrapolation, and optimized experiment selection across simulated and experiment...

  8. Quotient-Space Diffusion Models

    cs.LG 2026-04 unverdicted novelty 6.0

    Quotient-space diffusion models handle symmetries by diffusing on the space of equivalent configurations under group actions like SE(3), reducing learning complexity and guaranteeing correct sampling for molecular generation.

  9. Energy-Guided Generative Modeling for Low-Energy Molecular Structure Discovery

    cs.LG 2025-12 unverdicted novelty 6.0

    EnFlow integrates flow-based conformer generation with energy landscape modeling to enable joint ensemble generation and ground-state identification using only 1-2 ODE steps.

  10. Bayesian E(3)-Equivariant Interatomic Potential with Iterative Restratification of Many-body Message Passing

    cs.LG 2025-10 unverdicted novelty 6.0

    Bayesian E(3)-equivariant MLPs with joint energy-force NLL loss achieve competitive accuracy while enabling uncertainty-guided active learning, OOD detection, and calibration.

  11. Equivariant Graph Neural Networks Improve Optical Spectra Prediction for Materials Screening

    physics.optics 2026-06 unverdicted novelty 5.0

    Equivariant GNNs outperform prior models on optical spectra and static permittivity prediction using RPA datasets for materials screening.

  12. A Systematic Survey and Benchmark of Deep Learning for Molecular Property Prediction in the Foundation Model Era

    cs.LG 2026-04 accept novelty 5.0

    A systematic survey and benchmark of four deep learning paradigms for molecular property prediction that organizes the field, critiques current data practices, and outlines three future directions.

  13. Bayesian Prior Construction for Uncertainty Quantification in First-Principles Statistical Mechanics

    cond-mat.stat-mech 2025-09 accept novelty 5.0

    Bayesian hyperparameter selection and ground-state-enforcing priors are compared for cluster expansions; standard posteriors rarely reproduce DFT ground states, and cone-restricted priors fix this.

  14. Integrating Mechanistic and Data-Driven Models for Neurological Disorders through Differentiable Programming

    cs.AI 2026-06 unverdicted novelty 3.0

    This perspective paper categorizes hybrid architectures for combining mechanistic and data-driven models using residual learning, Neural ODEs, and solver-in-the-loop to model neurological disorder progression.