Home▸Materials Science▸Active Learning for Machine-Learned Interatomic Potentials

Active Learning for Machine-Learned Interatomic Potentials (2D)

A 2D toy configuration space (two structural parameters) where a real ensemble of kernel-regression models disagrees most on unexplored regions; active learning queries the highest-uncertainty point each round and measurably beats random sampling on a live learning curve.

Materials Science2DAdvanced60 FPS📱 Mobile-adapted⇄ 3D version

A 2D toy configuration space stands in for the atomic-environment space of a real machine-learned interatomic potential: two structural parameters (x, y) map to a synthetic but genuinely multi-featured "energy" surface. A real ensemble of Nadaraya–Watson kernel regressors, each with a different bandwidth, is trained on a growing labeled set and disagrees most where it has seen the least data. Every round, the engine reads that disagreement off the whole grid and queries ground truth at the single highest-uncertainty point — while an independent random-sampling baseline runs in parallel. The live learning curve plots prediction error against number of labels for both strategies, so the efficiency gain of active selection over random sampling is a measured result, not a claim.

⚙ Under the hood

Train a query-by-committee active-learning loop for a machine-learned interatomic potential: watch an ensemble of models disagree most on unfamiliar atomic environments, trigger simulated DFT labeling, and see the uncertainty collapse.

machine learningmaterials scienceDFTactive learningneural network potentialmolecular dynamics

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install