mlsanityData to 7 Oct 2026

Diffusion probabilistic models enhance variational autoencoder for crystal structure generative modeling

arXiv 2308.02165 ↗First posted 4 Aug 2023cs.LG cond-mat.mtrl-sci cond-mat.stat-mechcited by 17

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

Abstract (arXiv)

The crystal diffusion variational autoencoder (CDVAE) is a machine learning model that leverages score matching to generate realistic crystal structures that preserve crystal symmetry. In this study, we leverage novel diffusion probabilistic (DP) models to denoise atomic coordinates rather than adopting the standard score matching approach in CDVAE. Our proposed DP-CDVAE model can reconstruct and generate crystal structures whose qualities are statistically comparable to those of the original CDVAE. Furthermore, notably, when comparing the carbon structures generated by the DP-CDVAE model with relaxed structures obtained from density functional theory calculations, we find that the DP-CDVAE generated structures are remarkably closer to their respective ground states. The energy differences between these structures and the true ground states are, on average, 68.1 meV/atom lower than those generated by the original CDVAE. This significant improvement in the energy accuracy highlights the effectiveness of the DP-CDVAE model in generating crystal structures that better represent their ground-state configurations.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Placed in Flow matching, time series and science, one of 14 directions in Diffusion models. Builds on 24 of its 48 references that are in the map. First posted on arXiv 4 Aug 2023.

Builds on

24 of 48 references.

Closest ancestors · 1

Same direction, newest first

  1. 2022GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation6 Mar

Other references · 23

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  3. 2018How Powerful are Graph Neural Networks?cited by 4,355
  4. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,067
  5. 2021Improved Denoising Diffusion Probabilistic Modelscited by 3,736
  6. 2021Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gaugescited by 1,131
  7. 2019Strategies for Pre-training Graph Neural Networkscited by 964
  8. 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 735
  9. 2017SchNet: A continuous-filter convolutional neural network for modeling quantum interactionscited by 569
  10. 2020Directional Message Passing for Molecular Graphscited by 556
  11. 2021Solving Inverse Problems in Medical Imaging with Score-Based Generative Modelscited by 514
  12. 2018Spherical CNNscited by 452
11 more
  1. 2021GemNet: Universal Directional Graph Neural Networks for Moleculescited by 342
  2. 2021Crystal Diffusion Variational Autoencoder for Periodic Material Generationcited by 268
  3. 2020Fast and Uncertainty-Aware Directional Message Passing for Non-Equilibrium Moleculescited by 254
  4. 20233D Equivariant Diffusion for Target-Aware Molecule Generation and Affinity Predictioncited by 176
  5. 2021Learning Gradient Fields for Molecular Conformation Generationcited by 162
  6. 2023On the Expressive Power of Geometric Graph Neural Networkscited by 89
  7. 2018Molecular generative model based on conditional variational autoencoder for de novo molecular designcited by 73
  8. 2020An invertible crystallographic representation for general inverse design of inorganic crystals with targeted propertiescited by 53
  9. 2018Conditional molecular design with deep generative modelscited by 44
  10. 2023Star-Shaped Denoising Diffusion Probabilistic Modelscited by 13
  11. 2023Decoding Structure-Spectrum Relationships with Physically Organized Latent Spacescited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 3; 2024: 4; 2025: 7; 2026: 3.

17 papers.

Sort:

By area:

  1. 2023A Hitchhiker's Guide to Geometric GNNs for 3D Atomic Systemscited by 65
  2. 2023Scalable Diffusion for Materials Generationcited by 52
  3. 2024FlowLLM: Flow Matching for Material Generation with Large Language Models as Base Distributionscited by 52
  4. 2023Crystal-GFN: sampling crystals with desirable properties and constraintscited by 29
  5. 2024Ab Initio Structure Solutions from Nanocrystalline Powder Diffraction Datacited by 12
  6. 2025Materials Generation in the Era of Artificial Intelligence: A Comprehensive Surveycited by 9
  7. 2025CrystalGRW: Generative Modeling of Crystal Structures with Targeted Properties via Geodesic Random Walkscited by 5
  8. 2026DMFlow: Disordered Materials Generation by Flow Matchingcited by 5
  9. 2024Grand canonical generative diffusion model for crystalline phases and grain boundariescited by 1
  10. 2025Contrastive Learning of English Language and Crystal Graphs for Multimodal Representation of Materials Knowledgecited by 1
  11. 2025QE-Catalytic: A Graph-Language Multimodal Base Model for Relaxed-Energy Prediction in Catalytic Adsorptioncited by 1
  12. 2026Fourier Transformers for Latent Crystallographic Diffusion and Generative Modelingcited by 1
5 more
  1. 2024Unleashing the power of novel conditional generative approaches for new materials discoverycited by 0
  2. 2025On the Connection Between Diffusion Models and Molecular Dynamicscited by 0
  3. 2025AtomDiffuser: Time-Aware Degradation Modeling for Drift and Beam Damage in STEM Imagingcited by 0
  4. 2025Challenges in Non-Polymeric Crystal Structure Prediction: Why a Geometric, Permutation-Invariant Loss is Neededcited by 0
  5. 2026A generative machine learning model for designing metal hydrides applied to hydrogen storagecited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Crystal Structure Prediction by Joint Equivariant Diffusion
  2. 2025LLM Meets Diffusion: A Hybrid Framework for Crystal Material Generation
  3. 2025WyckoffDiff -- A Generative Diffusion Model for Crystal Symmetry
  4. 2024Generative Inverse Design of Crystal Structures via Diffusion Models with Transformers
  5. 2023Data-Driven Score-Based Models for Generating Stable Structures with Adaptive Crystal Cells
  6. 2025All-atom Diffusion Transformers: Unified generative modelling of molecules and materials
  7. 2025OXtal: An All-Atom Diffusion Model for Organic Crystal Structure Prediction
  8. 2023Investigating the Behavior of Diffusion Models for Accelerating Electronic Structure Calculations