AI Paper Atlas

Research map / Graph learning / GNN architectures

Geometric Graph Neural Diffusion for Stable Molecular Dynamics

Haokai Hong, Wanyu Lin, Zhang Chusong, KC Tan

ICLR 2026 · Graph learning › GNN architectures · approach: Equivariant networks for molecules

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 changedWhereMeasuredEffect
add GGND moduleVisNet · 3BPA 600Kstability (ps): 25.358 -> 100Improved
add GGND moduleVisNet · 3BPA 1200Kstability (ps): 0.004 -> 11.209Improved
add GGND moduleMACE · 3BPA 1200Kstability (ps): 1.965 -> 29.218Improved
add GGND moduleSEGNO · 3BPA 300Kenergy MAE (eV): 0.593 -> 0.293Improved
add GGND moduleSEGNO · 3BPA 600Kenergy MAE (eV): 0.908 -> 0.295Improved
add GGND moduleVisNet · 3BPA 300Kenergy MAE (eV): 0.002 -> 0.002No clear effect
use fully-connected diffusion instead of local diffusion (ablation)GGND · 3BPA 600Kstability (ps): 39.075 -> 100Improved
use diffusion instead of plain fully-connected message passing (ablation)GGND · 3BPA 600Kstability (ps): 69.292 -> 100Improved
add local diffusion only to baseline (ablation)VisNet · 3BPA 300Kenergy MAE (eV): 0.002 -> 0.013Hurt
add GGND module to VisNetVisNet · 3BPA training costtraining time (h): 6.822 -> 8.633Hurt
add GGND module to VisNetVisNet · 3BPA GPU memorytraining memory (GiB): 20.457 -> 23.582Hurt
use GGND versus best baseline on stabilityGGND · SAMD23 SiN OODstability (ps): 86.512 -> 99.892Improved

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.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy