Research map / Graph learning / GNN architectures
Geometric Graph Neural Diffusion for Stable Molecular Dynamics
Haokai Hong, Wanyu Lin, Zhang Chusong, KC Tan
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 GGND module | VisNet · 3BPA 600K | stability (ps): 25.358 -> 100 | Improved |
| add GGND module | VisNet · 3BPA 1200K | stability (ps): 0.004 -> 11.209 | Improved |
| add GGND module | MACE · 3BPA 1200K | stability (ps): 1.965 -> 29.218 | Improved |
| add GGND module | SEGNO · 3BPA 300K | energy MAE (eV): 0.593 -> 0.293 | Improved |
| add GGND module | SEGNO · 3BPA 600K | energy MAE (eV): 0.908 -> 0.295 | Improved |
| add GGND module | VisNet · 3BPA 300K | energy MAE (eV): 0.002 -> 0.002 | No clear effect |
| use fully-connected diffusion instead of local diffusion (ablation) | GGND · 3BPA 600K | stability (ps): 39.075 -> 100 | Improved |
| use diffusion instead of plain fully-connected message passing (ablation) | GGND · 3BPA 600K | stability (ps): 69.292 -> 100 | Improved |
| add local diffusion only to baseline (ablation) | VisNet · 3BPA 300K | energy MAE (eV): 0.002 -> 0.013 | Hurt |
| add GGND module to VisNet | VisNet · 3BPA training cost | training time (h): 6.822 -> 8.633 | Hurt |
| add GGND module to VisNet | VisNet · 3BPA GPU memory | training memory (GiB): 20.457 -> 23.582 | Hurt |
| use GGND versus best baseline on stability | GGND · SAMD23 SiN OOD | stability (ps): 86.512 -> 99.892 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SEGNO — SEGNO: Generalizing Equivariant Graph Neural Networks with Physical Inductive Biases (ICLR 2024) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Pushing the Limits of All-Atom Geometric Graph Neural Networks: Pre-Training, Scaling, and Zero-Shot Transfer (ICLR 2025 · similarity 0.89)
- Neural P$^3$M: A Long-Range Interaction Modeling Enhancer for Geometric GNNs (NeurIPS 2024 · similarity 0.88)
- Geometric-Facilitated Denoising Diffusion Model for 3D Molecule Generation (AAAI 2024 · similarity 0.88)
- Empowering Active Learning for 3D Molecular Graphs with Geometric Graph Isomorphism (NeurIPS 2024 · similarity 0.87)
- Bridging Geometric States via Geometric Diffusion Bridge (NeurIPS 2024 · similarity 0.87)
- Enhancing Generalizability in Molecular Conformation Generation with METRIZATION-Informed Geometric Diffusion Pretraining (AAAI 2025 · similarity 0.86)
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.
