AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

Learning from the Electronic Structure of Molecules across the Periodic Table

Manasa Kaniselvan, Benjamin Kurt Miller, Meng Gao, Juno Nam, Daniel S. Levine

ICLR 2026 · Probabilistic and scientific modeling › Molecules and proteins · approach: Machine-learned interatomic potentials · 5 citations (OpenAlex)

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 node-parity expansion in the Hamiltonian head (ablation)HELM · nabla2DFT 2k testenergy error from H (meV): 329,923.53 -> 1,530.83Improved
add l=0 component scaling on top of node-parity expansion (ablation)HELM · nabla2DFT 2k testenergy error from H (meV): 1,530.83 -> 706.39Improved
add l=0 scaling on top of no parity expansion (ablation)HELM · nabla2DFT 2k testHamiltonian error (1e-6 Eh): 324.35 -> 120.18Improved
Hamiltonian pretraining with frozen backbone before energy trainingHELM · nabla2DFT 2k testenergy MAE (meV): 791.02 -> 324.64Improved
Hamiltonian pretraining then finetuning on energiesHELM · nabla2DFT 2k testenergy MAE (meV): 791.02 -> 266.23Improved
Hamiltonian pretraining with frozen backboneHELM · OMol_all_5kenergy MAE (meV): 5976.40 -> 3255.69Improved
finetune the pretrained backbone instead of keeping it frozen (ablation)HELM · OMol_CSH_58k testsenergy MAE (meV): 3255.69 -> 3926.12Hurt
finetune the pretrained backbone instead of keeping it frozen (ablation)HELM · nabla2DFT 2k testenergy MAE (meV): 324.64 -> 266.23Improved
Hamiltonian pretraining on OMol_CSH_58k, in-distribution energy training splitHELM · OMol_CSH_58k trainenergy MAE (meV): 353.15 -> 361.98No clear effect
use HELM architecture for Hamiltonian prediction instead of previous modelsHELM · nabla2DFTH error: ~3-5x lower than the previous bestImproved

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