AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

From Evaluation to Design: Using Potential Energy Surface Smoothness Metrics to Guide ML Interatomic Potential Architectures

Ryan Liu, Eric Qu, Tobias Kreiman, Samuel M Blau, Aditi S. Krishnapriyan

ICML 2026 · Probabilistic and scientific modeling › Molecules and proteins · approach: Machine-learned interatomic potentials

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 controllable Gaussian smearing width (scale 5) (ablation)MinDScAIP · BSCTFSD (full): 97.4 -> 83.1Improved
add temperature-controlled attention (tau 10) (ablation)MinDScAIP · BSCTFSD (full): 97.4 -> 75.5Improved
add weight decay (ablation)MinDScAIP · BSCTFSD (full): 97.4 -> 76.3Improved
combine smearing and temperature (ablation)MinDScAIP · BSCTFSD (full): 97.4 -> 43.2Improved
combine smearing and temperature (ablation)MinDScAIP · SPICE testEnergy MAE (meV/atom): 0.09 -> 0.12Hurt
replace standard kNN with Diff-kNN (ablation)MinDScAIP (gradient forces) · NVE simulationEnergy drift (meV/atom): 19.10 -> 0.678Improved
use gradient-based forces instead of direct force head (ablation)MinDScAIP · SPICE/BSCTEnergy drift / FSD: 2.6e5 -> 0.678; FSD 71.8 -> 97.4Mixed
scale model from small (3.8M) to large (60M) (ablation)MinDScAIP · SPICE/BSCTFSD (full): 80.2 -> 97.4Hurt
scale model from small (3.8M) to large (60M) (ablation)MinDScAIP · SPICE testForces MAE (meV/A): 6.63 -> 3.02Improved
strengthen smoothness designs from weak to strong (ablation)MinDScAIP-30M · Matbench Discoverykappa_SRME: 0.77 -> 0.49Improved
lower FSD via smoothness designs (Vanilla to Smear. & Temp.) (ablation)MinDScAIP · MD22 at 2000KMax kinetic temperature jump: 9734 -> 490Improved

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