Search NASASearch

DOE OSTI · 2999427

When more data hurts: Optimizing data coverage while mitigating diversity-induced underfitting in an ultrafast machine-learned potential

Abstract

Machine-learned interatomic potentials (MLIPs) are becoming an essential tool in materials modeling. However, optimizing the generation of training data used to parametrize the MLIPs remains a significant challenge. This is because MLIPs can fail when encountering local environments too different from those present in the training data. The difficulty of determining a priori the environments that will be encountered during molecular dynamics simulation necessitates diverse, high-quality training data. Here, this study investigates how training data diversity affects the performance of MLIPs using the Ultra-Fast force field (UF 3 ) to model amorphous silicon nitride. We employ expert and autonomously generated data to create the training data and fit four force field variants to subsets of the data. Our findings reveal a critical balance in training data diversity: insufficient diversity hinders generalization, while excessive diversity can exceed the MLIP's learning capacity, reducing simulation accuracy. Specifically, we found that the UF 3 variant trained on a subset of the training data, in which nitrogen-rich structures were removed, offered vastly better prediction and simulation accuracy than any other variant. By comparing these UF 3 variants, we highlight the nuanced requirements for creating accurate MLIPs, emphasizing the importance of application-specific training data to achieve optimal performance in modeling complex material behaviors.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Gibson, Jason B. [Univ. of Florida, Gainesville, FL (United States)] (ORCID:0000000179745264), Janicki, Tesia Danielle [Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)] (ORCID:0000000160086414), Hire, Ajinkya C. [Univ. of Florida, Gainesville, FL (United States)] (ORCID:0000000331472521), Bishop, Chris [Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)], Lane, J. Matthew Doyle [Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)] (ORCID:0000000278789158), Hennig, Richard G. [Univ. of Florida, Gainesville, FL (United States)] (ORCID:0000000349337686). 2025-04-22. When more data hurts: Optimizing data coverage while mitigating diversity-induced underfitting in an ultrafast machine-learned potential. https://doi.org/10.1103/physrevmaterials.9.043802

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related reports

Cation Effects on CO 2 Delivery to Cu Electrode in Reactive Capture of CO 2

The direct electrochemical conversion of captured CO 2 , known as reactive capture of CO 2 (RCC), remains a formidable challenge in heterogeneous catalysis. Given that amines are one of the most widely used capture agents for CO 2 , it would be desirable to electrochemically reduce the resultant adducts, such as carbamate, directly in RCC. However, current understanding suggests that the primary species undergoing reduction in RCC with amines is the CO 2 dissociated from the sorbent. Herein, we employ ab initio molecular dynamics (AIMD) with DFT to analyze how the nature of alkali metal cations in the electrolyte affects carbamate at the Cu surface, thereby assessing the possibility of promoting RCC by cation effects. The simulations show that the carbamate’s orientation with respect to the electrode is governed by the optimal distance between the carbamate and the cation, specifically how this distance aligns with the cation’s hydration spheres. Moreover, the slow-growth AIMD results indicate that the CO 2 dissociation barrier correlates with the orientation of carbamate at the interface. When the carbamate resides beyond the cation’s first hydration sphere, it adopts a flat orientation with respect to the surface that promotes the release of CO 2 from the capture agent. In contrast, when the carbamate disrupts the first hydration sphere and exhibits a strong cation−π interaction, it adopts an upright orientation that is less conducive to CO 2 release. These findings reveal a nontrivial cation effect in RCC, suggesting that it should be possible to optimize RCC via the choice of the electrolyte.

ab initio molecular dynamics