Risk-Controlled DFT Selection for Machine-Learned Interatomic Potentials
Abstract
Automated materials discovery requires evaluating large pools of candidate configurations, with density functional theory (DFT) as a major throughput bottleneck. Pretrained machine-learned interatomic potentials (MLIPs) approximate DFT cheaply, but their errors vary by orders of magnitude across configurations, making it difficult to decide when to trust the MLIP predictions and when to fall back to DFT. We frame this decision as a risk-controlled selection problem and solve it within a conformal risk control (CRC) framework. Each configuration is mapped from atomic numbers and Cartesian coordinates to a fixed geometric descriptor, which a lightweight predictor uses to estimate an MLIP-DFT disagreement score in energy, average force, and maximum force. CRC then calibrates this score into a selection threshold for DFT labeling such that the MLIP error on the unselected configurations, which keep their MLIP labels, stays below a user-chosen fraction α of the expected error. Under an identical guarantee across eight dataset-MLIP pairs, our predictor routes 7% and 14% fewer configurations to DFT at α=0.3 than committee baselines. Because committee disagreement overlooks errors the MLIPs share, our gains are largest when the DFT labels and MLIP training data differ in DFT functional and basis set, and narrow when the two agree. Without model access, the structure-only predictor also ranks errors as well as the target MLIP's latent representation and outperforms widely used descriptors such as Smooth Overlap of Atomic Positions (SOAP) and Atom-Centered Symmetry Functions (ACSF).
Related
- Why Multi-Layer Message Passing Works: Completeness Theory for Graph Neural Network Interatomic Potentials
- Learned smoothing in operator-adapted multigrid: controlled accuracy gains and computational limits
- Quantum-Assisted Optimization Guided by Machine-Learned Risk Maps for Aerial Surveillance Routing
- Model selection in linear mixed-effects analysis of craving outcomes in a randomized controlled trial of medications for opioid use disorders
- FDR-Controlled Variable Selection for Generalized Linear Models and Cox Regression with Virtual Dummies
BibTeX
@misc{ryu2026riskcontrolled,
title = {{Risk-Controlled DFT Selection for Machine-Learned Interatomic Potentials}},
author = {Kun Ryu and Yeo Jin Jung and Shyue Ping Ong and Claire Donnat and Ying Shirley Meng},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.19630},
eprint = {2610.19630},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.19630},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Risk-Controlled DFT Selection for Machine-Learned Interatomic Potentials
AU - Kun Ryu
AU - Yeo Jin Jung
AU - Shyue Ping Ong
AU - Claire Donnat
AU - Ying Shirley Meng
PY - 2026
DA - 2026/10/14
PB - arXiv
DO - 10.48550/arxiv.2610.19630
UR - https://arxiv.org/abs/2610.19630
AB - Automated materials discovery requires evaluating large pools of candidate configurations, with density functional theory (DFT) as a major throughput bottleneck. Pretrained machine-learned interatomic potentials (MLIPs) approximate DFT cheaply, but their errors vary by orders of magnitude across configurations, making it difficult to decide when to trust the MLIP predictions and when to fall back to DFT. We frame this decision as a risk-controlled selection problem and solve it within a conformal risk control (CRC) framework. Each configuration is mapped from atomic numbers and Cartesian coordinates to a fixed geometric descriptor, which a lightweight predictor uses to estimate an MLIP-DFT disagreement score in energy, average force, and maximum force. CRC then calibrates this score into a selection threshold for DFT labeling such that the MLIP error on the unselected configurations, which keep their MLIP labels, stays below a user-chosen fraction α of the expected error. Under an identical guarantee across eight dataset-MLIP pairs, our predictor routes 7% and 14% fewer configurations to DFT at α=0.3 than committee baselines. Because committee disagreement overlooks errors the MLIPs share, our gains are largest when the DFT labels and MLIP training data differ in DFT functional and basis set, and narrow when the two agree. Without model access, the structure-only predictor also ranks errors as well as the target MLIP's latent representation and outperforms widely used descriptors such as Smooth Overlap of Atomic Positions (SOAP) and Atom-Centered Symmetry Functions (ACSF).
ER -