Preprint
Aug 2026
ED-DiT: Physics-Guided Diffusion Pretraining for Transferable Molecular Representations from Electron Density
ED-DiT, a physics-guided Diffusion Transformer for self-supervised pretraining on electron-density point clouds, demonstrates the effectiveness of physics-guided electron-density pretraining for learning transferable molecular representations.
Shuang Liang, Haocheng Wang, Jiayi Song et al.
· 0 citations