ED-DiT: Physics-Guided Diffusion Pretraining for Transferable Molecular Representations from Electron Density

• URL: http://arxiv.org/abs/2608.03260v1
• Date: Tue, 04 Aug 2026 07:33:35 GMT
• Title: ED-DiT: Physics-Guided Diffusion Pretraining for Transferable Molecular Representations from Electron Density
• Abstract summary: We propose ED-DiT, a physics-guided Diffusion Transformer for self-supervised pretraining on electron-density point clouds.<n> ED-DiT learns reusable representations by reconstructing corrupted and partially masked log-density fields across diffusion noise levels.<n>Experiments on six EDBench tasks show that ED-DiT consistently outperforms the same architecture trained from scratch, especially under limited supervision.
• Score: 14.29876606291099
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Pretraining has shown strong potential for learning transferable representations, yet it remains underexplored for electron-density-based molecular learning. Electron density provides a continuous three-dimensional description of molecular electronic structure, capturing both local spatial patterns and global physical quantities. This raises a key question: can electron-density fields be used for self-supervised pretraining to learn a shared representation that transfers across diverse electronic-structure-related tasks? We propose ED-DiT, a physics-guided Diffusion Transformer for self-supervised pretraining on electron-density point clouds. ED-DiT learns reusable representations by reconstructing corrupted and partially masked log-density fields across diffusion noise levels. An electron-number consistency constraint is further introduced to preserve the total electronic mass. The pretrained encoder can be adapted to property prediction, open-/closed-shell classification, molecule-electron-density retrieval, and molecule-conditioned electron-density prediction. Experiments on six EDBench tasks show that ED-DiT consistently outperforms the same architecture trained from scratch, especially under limited supervision. For molecule-conditioned electron-density prediction, it reduces RMSE from 2.2474 to 1.3753 and surpasses the available baseline. With only 10% labels, it improves orbital energy prediction RMSE from 0.0293 to 0.0138. These results demonstrate the effectiveness of physics-guided electron-density pretraining for learning transferable molecular representations.

Related papers

• Learning the Kohn-Sham map with neural operators for quasi-linear scaling density functional theory [57.90994363143731]
  Kohn--Sham density functional theory (DFT) underpins electronic-structure simulations.<n>Repeated orbital diagonalizations lead to cubic scaling, restricting quantum calculations to modest scales only.<n>We develop a learning target for orbital-free DFT that maps a Kohn--Sham potential directly to the corresponding density and noninteracting kinetic energy.
  arXiv  Detail & Related papers  (2026-08-24T23:05:42Z)
• Enhancing molecular dynamics with equivariant machine-learned densities [10.247839738079698]
  We introduce DenSNet, a density-first approach to machine-learned electronic structure.<n>We validate DenSNet on ethanol, ethanethiol, and resorcinol, where infrared spectra from machine-learned trajectories show excellent agreement with experimental gas-phase measurements.<n>We show that reinstating the electron density as the central learned quantity opens a practical route to transferable prediction of spectroscopic and electronic observables in large-scale molecular simulations.
  arXiv  Detail & Related papers  (2026-04-27T14:49:53Z)
• V2Rho-FNO: Fourier Neural Operator for Electronic Density Prediction [10.028886116912622]
  We present a universal framework based on Fourier Neural Operators (FNOs) that directly learns the mapping from external potentials to electron density distributions.<n>Our work establishes neural operator learning as a promising route for fast, accurate, and transferable electronic structure prediction.
  arXiv  Detail & Related papers  (2026-03-13T06:53:13Z)
• Acceleration of Atomistic NEGF: Algorithms, Parallelization, and Machine Learning [61.12861060232382]
  The Non-equilibrium Green's function (NEGF) formalism is a powerful method to simulate the quantum transport properties of nanoscale devices.<n>This paper summarizes key (algorithmic) achievements that have allowed us to bring DFT+NEGF simulations closer to the dimensions and functionality of realistic systems.
  arXiv  Detail & Related papers  (2026-02-03T12:01:39Z)
• OrbitAll: A Unified Quantum Mechanical Representation Deep Learning Framework for All Molecular Systems [64.69217059173184]
  OrbitAll is a geometry- and physics-informed deep learning framework.<n>It can represent and process any molecular system with arbitrary charges, spins, and environmental effects.<n>It achieves chemical accuracy using 10 times fewer training data than competing AI models, with a speedup of approximately $103$ - $104$ compared to density functional theory.
  arXiv  Detail & Related papers  (2025-07-05T01:21:56Z)
• Tokenizing Electron Cloud in Protein-Ligand Interaction Learning [51.74909649330779]
  ECBind is a method for tokenizing electron cloud signals into quantized embeddings.<n>It helps uncover binding modes that cannot be fully represented by atom-level models.<n>To extend its applicability to a wider range of scenarios, we utilize knowledge distillation to develop an electron-cloud-agnostic prediction model.
  arXiv  Detail & Related papers  (2025-05-25T07:36:50Z)
• EDBench: Large-Scale Electron Density Data for Molecular Modeling [24.86737047895369]
  electron density (ED) $rho(r)$ determines all ground state properties of interactive multi-particle systems.<n>EDBench is a large-scale, high-quality dataset of ED designed to advance learning-based research at the electronic scale.
  arXiv  Detail & Related papers  (2025-05-14T10:23:22Z)
• E3STO: Orbital Inspired SE(3)-Equivariant Molecular Representation for Electron Density Prediction [0.0]
  We introduce a novel SE(3)-equivariant architecture, drawing inspiration from Slater-Type Orbitals (STO) Our approach offers an alternative functional form for learned orbital-like molecular representation. We showcase the effectiveness of our method by achieving SOTA prediction accuracy of molecular electron density with 30-70% improvement over other work on Molecular Dynamics data.
  arXiv  Detail & Related papers  (2024-10-08T15:20:33Z)
• Higher-Order Equivariant Neural Networks for Charge Density Prediction in Materials [3.7655047338409893]
  ChargE3Net is an E(3)-equivariant graph neural network for predicting electron density in atomic systems. We show that ChargE3Net exceeds the performance of prior work on diverse sets of molecules and materials.
  arXiv  Detail & Related papers  (2023-12-08T21:56:19Z)
• Computing molecular excited states on a D-Wave quantum annealer [52.5289706853773]
  We demonstrate the use of a D-Wave quantum annealer for the calculation of excited electronic states of molecular systems. These simulations play an important role in a number of areas, such as photovoltaics, semiconductor technology and nanoscience.
  arXiv  Detail & Related papers  (2021-07-01T01:02:17Z)
• BIGDML: Towards Exact Machine Learning Force Fields for Materials [55.944221055171276]
  Machine-learning force fields (MLFF) should be accurate, computationally and data efficient, and applicable to molecules, materials, and interfaces thereof. Here, we introduce the Bravais-Inspired Gradient-Domain Machine Learning approach and demonstrate its ability to construct reliable force fields using a training set with just 10-200 atoms.
  arXiv  Detail & Related papers  (2021-06-08T10:14:57Z)
• Quantum deep field: data-driven wave function, electron density generation, and atomization energy prediction and extrapolation with machine learning [7.106986689736826]
  Deep neural networks (DNNs) have been used to successfully predict molecular properties calculated based on the Kohn--Sham density functional theory (KS-DFT) This letter presents the quantum deep field (QDF), which provides the electron density with an unsupervised but end-to-end physics-informed modeling by learning the atomization energy on a large-scale dataset.
  arXiv  Detail & Related papers  (2020-11-16T13:15:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.