Data for: "Data efficiency and extrapolation trends in neural network interatomic potentials"

Schwalbe Koda, Daniel; Vita, Joshua

doi:10.11579/2335917

Title: Data for: "Data efficiency and extrapolation trends in neural network interatomic potentials"

Dataset
Other Related Research

Abstract

Data to reproduce the paper "Data efficiency and extrapolation trends in neural network interatomic potentials" (LLNL-JRNL-845001). Contains: plots, results, and visualization scripts. Preprint available at https://doi.org/10.48550/arXiv.2302.05823.

Authors:

Schwalbe Koda, Daniel; Vita, Joshua

Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States); PLS/MSD
Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)

Publication Date:: Sat Dec 31 23:00:00 EST 2022

Other Number(s):: LLNL-MI-852928

DOE Contract Number:: AC52-07NA27344

Research Org.:: LLNL-IPO (LLNL Innovation and Partnerships Office (IPO))

Sponsoring Org.:: USDOE National Nuclear Security Administration (NNSA)

OSTI Identifier:: 2335917

DOI:: https://doi.org/10.11579/2335917

Citation Formats


                    Schwalbe Koda, Daniel, and Vita, Joshua. Data for: "Data efficiency and extrapolation trends in neural network interatomic potentials".  United States: N. p., 2022. 
        Web.  doi:10.11579/2335917.

Copy to clipboard


                    Schwalbe Koda, Daniel, & Vita, Joshua. Data for: "Data efficiency and extrapolation trends in neural network interatomic potentials".  United States.  doi:https://doi.org/10.11579/2335917

Copy to clipboard


                    Schwalbe Koda, Daniel, and Vita, Joshua. 2022.  
"Data for: "Data efficiency and extrapolation trends in neural network interatomic potentials"".  United States.  doi:https://doi.org/10.11579/2335917.  https://www.osti.gov/servlets/purl/2335917. Pub date:Sat Dec 31 23:00:00 EST 2022

Copy to clipboard


                    
@article{osti_2335917,

  title        = {Data for: "Data efficiency and extrapolation trends in neural network interatomic potentials"},

  author       = {Schwalbe Koda, Daniel and Vita, Joshua},

  abstractNote = {Data to reproduce the paper "Data efficiency and extrapolation trends in neural network interatomic potentials" (LLNL-JRNL-845001). Contains: plots, results, and visualization scripts. Preprint available at https://doi.org/10.48550/arXiv.2302.05823.},

  doi          = {10.11579/2335917},

  journal      = {},

  number       = ,

  volume       = ,

  place        = {United States},

  year         = {Sat Dec 31 23:00:00 EST 2022},

  month        = {Sat Dec 31 23:00:00 EST 2022}

}

Copy to clipboard

Dataset:

View Dataset

DOI: https://doi.org/10.11579/2335917

Save / Share:

Export Metadata

Save to My Library

Similar records in DOE Data Explorer and OSTI.GOV collections:

Data efficiency and extrapolation trends in neural network interatomic potentials

Journal Article Vita, Joshua A. ; Schwalbe-Koda, Daniel - Machine Learning: Science and Technology

Abstract Recently, key architectural advances have been proposed for neural network interatomic potentials (NNIPs), such as incorporating message-passing networks, equivariance, or many-body expansion terms. Although modern NNIP models exhibit small differences in test accuracy, this metric is still considered the main target when developing new NNIP architectures. In this work, we show how architectural and optimization choices influence the generalization of NNIPs, revealing trends in molecular dynamics (MD) stability, data efficiency, and loss landscapes. Using the 3BPA dataset, we uncover trends in NNIP errors and robustness to noise, showing these metrics are insufficient to predict MD stability in the high-accuracymore »« less
Data efficiency and extrapolation trends in neural network interatomic potentials

Journal Article Vita, Joshua A. ; Schwalbe-Koda, Daniel - Machine Learning: Science and Technology

Recently, key architectural advances have been proposed for neural network interatomic potentials (NNIPs), such as incorporating message-passing networks, equivariance, or many-body expansion terms. Although modern NNIP models exhibit small differences in test accuracy, this metric is still considered the main target when developing new NNIP architectures. In this work, we show how architectural and optimization choices influence the generalization of NNIPs, revealing trends in molecular dynamics (MD) stability, data efficiency, and loss landscapes. Using the 3BPA dataset, we uncover trends in NNIP errors and robustness to noise, showing these metrics are insufficient to predict MD stability in the high-accuracy regime.more »« less
Solar PV Water Setbacks: Ordinances (2022) and Extrapolated Trends

Dataset Geospatial Data Science, NREL

This dataset represents solar energy setback requirements from bodies of water based on county ordinances as of April 2022. A setback requirement is a minimum distance from water that an energy project may be developed, and these varied widely across the counties in which they existed. Two versions are provided: one reflecting only the county ordinances and another incorporating extrapolated trends. In the extrapolated version, a default setback of 30 meters was applied in counties without specific water setback regulations. A TIF data file and a PNG map of the data are provided for both versions, showing areas where solarmore »« less
Solar PV Transmission Setbacks: Ordinances (2022) and Extrapolated Trends

Dataset Geospatial Data Science, NREL

This dataset represents solar energy setback requirements from transmission. A setback requirement is a minimum distance from transmission infrastructure that an energy project may be developed. As of April 2022, no ordinances were discovered for any counties. Such ordinances are likely to arise as regulations continue to expand. Therefore, this dataset applies a 30-meter setback, sourced from trends in other infrastructure. A TIF data file and a PNG map of the data are provided, showing areas where solar energy is prohibited or permitted across the contiguous United States. For further details and citation, please refer to the publication linked below:more »« less
Solar PV Structure Setbacks: Ordinances (2022) and Extrapolated Trends

Dataset Geospatial Data Science, NREL

This dataset represents solar energy setback requirements from structures based on county ordinances as of April 2022. A setback requirement is a minimum distance from a structure that an energy project may be developed, and these varied widely across the counties in which they existed. Two versions are provided: one reflecting only the county ordinances and another incorporating extrapolated trends. In the extrapolated version, a default setback of 61 meters was applied in counties without specific structure setback regulations. A TIF data file and a PNG map of the data are provided for both versions, showing areas where solar energymore »« less

Similar Records