Research map / Probabilistic and scientific modeling / Molecules and proteins
Equivariant Atomic and Lattice Modeling Using Geometric Deep Learning for Crystal Structure Optimization
Ziduo Yang, Yi-Ming Zhao, Xian Wang, Wei Zhuo, Xiaoqing Liu, Lei Shen
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add lattice nodes with lattice-atom interaction (ablation) | E3Relax · MP | MAE coords: 0.075 -> 0.057 | Improved |
| add lattice nodes with lattice-atom interaction (ablation) | E3Relax · X-Mn-O | MAE coords: 0.158 -> 0.105 | Improved |
| add layer-wise supervision refinement (ablation) | E3Relax · MP | MAE vol: 8.465 -> 7.417 | Improved |
| add layer-wise supervision refinement (ablation) | E3Relax · X-Mn-O | MAE coords: 0.146 -> 0.105 | Improved |
| add self-connecting edges (ablation) | E3Relax · MP | MAE coords: 0.058 -> 0.057 | No clear effect |
| add self-connecting edges (ablation) | E3Relax · X-Mn-O | MAE coords: 0.144 -> 0.105 | Improved |
| E3Relax vs best baseline | E3Relax · C2DB | MAE vol: 56.909 vs 55.800 (GotenNet) | Hurt |
| E3Relax vs best baseline | E3Relax · MP | MAE coords: 0.066 -> 0.057 | Improved |
| E3Relax vs best baseline | E3Relax · Layered vdW | MAE vol: 1.006 -> 0.871 | Improved |
| single-step iteration-free model vs iterative potentials | E3Relax · MoS2 inference time | seconds: 3.72 -> 0.021 | Improved |
| vary layers, hidden size, layer weights (ablation) | E3Relax · X-Mn-O | MAE | No clear effect |
Closest papers by meaning
- Learning the Electronic Hamiltonian of Large Atomic Structures (ICML 2025 · similarity 0.92)
- ELoRA: Low-Rank Adaptation for Equivariant GNNs (ICML 2025 · similarity 0.88)
- Learning Equivariant Non-Local Electron Density Functionals (ICLR 2025 · similarity 0.88)
- Crystalformer: Infinitely Connected Attention for Periodic Structure Encoding (ICLR 2024 · similarity 0.88)
- Vector Field Oriented Diffusion Model for Crystal Material Generation (AAAI 2024 · similarity 0.87)
- Higher-Rank Irreducible Cartesian Tensors for Equivariant Message Passing (NeurIPS 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
