.png){kind=icon}
.png){kind=icon}
Return
How to Train a Shallow Ensemble
.png){kind=icon}
PRE
AI
DOI：10.1021/acs.jctc.6c00310.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Shallow ensembles provide a convenient strategy for uncertainty quantification in machine-learning interatomic potentials, which is computationally efficient because the different ensemble members share a large part of the model weights. In this work, we systematically investigate training strategies for shallow ensembles to balance calibration performance with computational cost. We first demonstrate that explicit optimization of a negative log-likelihood (NLL) loss improves calibration with respect to approaches based on ensembles of randomly initialized models or on a last-layer Laplace approximation. However, models trained solely on energy objectives yield miscalibrated force estimates. We show that explicitly modeling force uncertainties via an NLL objective is essential for reliable calibration, though it typically incurs a significant computational overhead. To address this, we validate an efficient protocol: full-model fine-tuning of a shallow ensemble originally trained with a probabilistic energy loss or one sampled from the Laplace posterior. This approach yields a calibration quality comparable to training from scratch. While the computational benefits scale with system and model size, it reduces training time by up to 96% for data sets containing large structures. We evaluate this protocol across a diverse range of materials, including amorphous carbon, ionic liquids (BMIM BF4), liquid water (H2O), barium titanate (BaTiO3), and a model tetrapeptide (Ac-Ala3-NHMe), establishing practical guidelines for reliable uncertainty quantification in atomistic machine learning.
Keywords:
Approximation
Calibration
Hessians
Machine learning
Neural networks
Journal
IF:
5.5
Papers:
1.1W
Citations:
5.4W