mlsanityData to 14 Oct 2026

Swallowing the Bitter Pill: Simplified Scalable Conformer Generation

arXiv 2311.17932 ↗First posted 27 Nov 2023physics.chem-ph cs.LGcited by 56

On the map: Diffusion models → Flow matching, time series and science

Abstract (arXiv)

We present a novel way to predict molecular conformers through a simple formulation that sidesteps many of the heuristics of prior works and achieves state of the art results by using the advantages of scale. By training a diffusion generative model directly on 3D atomic positions without making assumptions about the explicit structure of molecules (e.g. modeling torsional angles) we are able to radically simplify structure learning, and make it trivial to scale up the model sizes. This model, called Molecular Conformer Fields (MCF), works by parameterizing conformer structures as functions that map elements from a molecular graph directly to their 3D location in space. This formulation allows us to boil down the essence of structure prediction to learning a distribution over functions. Experimental results show that scaling up the model capacity leads to large gains in generalization performance without enforcing inductive biases like rotational equivariance. MCF represents an advance in extending diffusion models to handle complex scientific problems in a conceptually simple, scalable and effective manner.

Key facts. Cited by 56 papers in the map (data through 14 Oct 2026). Placed in Flow matching, time series and science, one of 14 directions in Diffusion models. Builds on 44 of its 60 references that are in the map. First posted on arXiv 27 Nov 2023.

Builds on

44 of 60 references.

Closest ancestors · 4

Same direction, newest first

  1. 2023Two for One: Diffusion Models and Force Fields for Coarse-Grained Molecular Dynamics1 Feb
  2. 2022Flow Matching for Generative Modeling6 Oct
  3. 2022Equivariant Diffusion for Molecule Generation in 3D31 Mar
  4. 2022GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation6 Mar

Other references · 40

  1. 2017Attention Is All You Needcited by 60,361
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  4. 2020Denoising Diffusion Implicit Modelscited by 9,781
  5. 2018How Powerful are Graph Neural Networks?cited by 4,393
  6. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,673
  7. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,689
  8. 2016HyperNetworkscited by 1,027
  9. 2020Learning to Simulate Complex Physics with Graph Networkscited by 880
  10. 2021E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentialscited by 708
  11. 2020Benchmarking Graph Neural Networkscited by 680
  12. 2021Perceiver IO: A General Architecture for Structured Inputs & Outputscited by 629
28 more
  1. 2022DiffDock: Diffusion Steps, Twists, and Turns for Molecular Dockingcited by 313
  2. 2022Torsional Diffusion for Molecular Conformer Generationcited by 263
  3. 2020Equivariant Flows: Exact Likelihood Generative Learning for Symmetric Densitiescited by 250
  4. 2020GEOM: Energy-annotated molecular conformations for property prediction and molecular generationcited by 243
  5. 2022Scalable Adaptive Computation for Iterative Generationcited by 195
  6. 2021Are Transformers More Robust Than CNNs?cited by 187
  7. 2022Sign and Basis Invariant Networks for Spectral Graph Representation Learningcited by 179
  8. 2020DiffusionNet: Discretization Agnostic Learning on Surfacescited by 164
  9. 2021Learning Gradient Fields for Molecular Conformation Generationcited by 163
  10. 2022Pre-training via Denoising for Molecular Property Predictioncited by 124
  11. 2023TRACT: Denoising Diffusion Models with Transitive Closure Time-Distillationcited by 116
  12. 2021GeoMol: Torsional Geometric Generation of Molecular 3D Conformer Ensemblescited by 107
  13. 2020Equivariant flow-based sampling for lattice gauge theorycited by 103
  14. 2022A Generalization of ViT/MLP-Mixer to Graphscited by 93
  15. 2021Generative Models as Distributions of Functionscited by 89
  16. 2021Learning Neural Generative Dynamics for Molecular Conformation Generationcited by 88
  17. 2019A Generative Model for Molecular Distance Geometrycited by 73
  18. 2021An End-to-End Framework for Molecular Conformation Generation via Bilevel Programmingcited by 59
  19. 2023Language models can generate molecules, materials, and protein binding sites directly in three dimensions as XYZ, CIF, and PDB filescited by 59
  20. 2022The Lie Derivative for Measuring Learned Equivariancecited by 54
  21. 2021Learning Signal-Agnostic Manifolds of Neural Fieldscited by 45
  22. 20233D molecule generation by denoising voxel gridscited by 40
  23. 2022Generalised Implicit Neural Representationscited by 31
  24. 2022Gauge-equivariant flow models for sampling in lattice field theories with pseudofermionscited by 31
  25. 2023Diffusion Probabilistic Fieldscited by 28
  26. 2023Normalizing flows for lattice gauge theory in arbitrary space-time dimensioncited by 23
  27. 2022Generalized Laplacian Positional Encoding for Graph Representation Learningcited by 21
  28. 2022Intrinsic Neural Fields: Learning Functions on Manifoldscited by 20

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 11; 2025: 29; 2026: 16.

56 papers.

Sort:

By area:

  1. 2024Let Your Graph Do the Talking: Encoding Structured Data for LLMscited by 118
  2. 2025All-atom Diffusion Transformers: Unified generative modelling of molecules and materialscited by 75
  3. 2024Orb: A Fast, Scalable Neural Network Potentialcited by 66
  4. 2024ET-Flow: Equivariant Flow-Matching for Molecular Conformer Generationcited by 43
  5. 2025Wyckoff Transformer: Generation of Symmetric Crystalscited by 27
  6. 2025NExT-Mol: 3D Diffusion Meets 1D Language Modeling for 3D Molecule Generationcited by 17
  7. 2025SimpleFold: Folding Proteins is Simpler than You Thinkcited by 16
  8. 2026Scaling Atomistic Protein Binder Design with Generative Pretraining and Test-Time Computecited by 15
  9. 2024Relaxed Equivariance via Multitask Learningcited by 13
  10. 2025Applications of Modular Co-Design for De Novo 3D Molecule Generationcited by 13
  11. 2025TABASCO: A Fast, Simplified Model for Molecular Generation with Improved Physical Qualitycited by 12
  12. 2025Flexible MOF Generation with Torsion-Aware Flow Matchingcited by 10
38 more
  1. 2024AdsorbDiff: Adsorbate Placement via Conditional Denoising Diffusioncited by 8
  2. 2025Feynman-Kac-Flow: Inference Steering of Conditional Flow Matching to an Energy-Tilted Posteriorcited by 7
  3. 2024Recurrent Aggregators in Neural Algorithmic Reasoningcited by 6
  4. 2025Sampling 3D Molecular Conformers with Diffusion Transformerscited by 6
  5. 2025Rao-Blackwell Gradient Estimators for Equivariant Denoising Diffusioncited by 5
  6. 2025Learning Inter-Atomic Potentials without Explicit Equivariancecited by 5
  7. 2024Equivariant Neural Tangent Kernelscited by 4
  8. 2025Efficient Molecular Conformer Generation with SO(3)-Averaged Flow Matching and Reflowcited by 4
  9. 2025Energy-Based Flow Matching for Generating 3D Molecular Structurecited by 4
  10. 2025Scaling Laws and Symmetry, Evidence from Neural Force Fieldscited by 4
  11. 2025Functional Adjoint Sampler: Scalable Sampling on Infinite Dimensional Spacescited by 4
  12. 2025Atomic Diffusion Models for Small Molecule Structure Elucidation from NMR Spectracited by 4
  13. 2024MolMix: A Simple Yet Effective Baseline for Multimodal Molecular Representation Learningcited by 3
  14. 2025PolyConf: Unlocking Polymer Conformation Generation through Hierarchical Generative Modelscited by 3
  15. 2025Quick ViTs: Speeding up Vision Transformers through Equivariancecited by 3
  16. 2025Scalable Non-Equivariant 3D Molecule Generation via Rotational Alignmentcited by 3
  17. 2025Unraveling the Potential of Diffusion Models in Small Molecule Generationcited by 3
  18. 2026Geometric Flow Matching for Molecular Conformation Generation via Manifold Decompositioncited by 3
  19. 2026Fast Organic Crystal Structure Prediction with Unit Cell Flow Matchingcited by 3
  20. 2024INRFlow: Flow Matching for INRs in Ambient Spacecited by 2
  21. 2026Generating Cyclic Conformers with Flow Matching in Cremer-Pople Coordinatescited by 2
  22. 2025Flow-Matching Based Refiner for Molecular Conformer Generationcited by 1
  23. 2025On Uncertainty Calibration for Equivariant Functionscited by 1
  24. 2025Training Dynamics of Learning 3D-Rotational Equivariancecited by 1
  25. 2026SimpleDesign: A Joint Model for Protein Sequence and Structure Codesigncited by 1
  26. 2024SymDiff: Equivariant Diffusion via Stochastic Symmetrisationcited by 0
  27. 2024MING: A Functional Approach to Learning Molecular Generative Modelscited by 0
  28. 2025polyGen: A Learning Framework for Atomic-level Polymer Structure Generationcited by 0
  29. 2025A standard transformer and attention with linear biases for molecular conformer generationcited by 0
  30. 2025A Conformation-Centric Generative Foundation Model for Linear Polymer Modeling and Designcited by 0
  31. 2025HyperDiffusionFields (HyDiF): Diffusion-Guided Hypernetworks for Learning Implicit Molecular Neural Fieldscited by 0
  32. 2025FunPhase: A Periodic Functional Autoencoder for Motion Generation via Phase Manifoldscited by 0
  33. 2025Energy-Guided Generative Modeling for Low-Energy Molecular Structure Discoverycited by 0
  34. 2026Breaking the Bottlenecks: Scalable Diffusion Models for 3D Molecular Generationcited by 0
  35. 2026Recurrent Equivariant Constraint Modulation: Learning Per-Layer Symmetry Relaxation from Datacited by 0
  36. 2026Quotient-Space Diffusion Modelscited by 0
  37. 2026SymDrift: One-Shot Generative Modeling under Symmetriescited by 0
  38. 2026When Does Equivariance Help? Canonical Alignment in Neural Fluid Surrogatescited by 0

50 most cited of 56 All 56 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024Physics-informed generative model for drug-like molecule conformers
  2. 2025Unlocking hidden biomolecular conformational landscapes in diffusion models at inference time
  3. 2024Equivariant Blurring Diffusion for Hierarchical Molecular Conformer Generation
  4. 2022DiffMD: A Geometric Diffusion Model for Molecular Dynamics Simulations
  5. 2025Chemistry-Enhanced Diffusion-Based Framework for Small-to-Large Molecular Conformation Generation
  6. 2025D3MES: Diffusion Transformer with multihead equivariant self-attention for 3D molecule generation
  7. 2023Modeling Molecular Structures with Intrinsic Diffusion Models
  8. 2024Diffusion Models in textitDe Novo Drug Design