A Perspective on Training Machine Learning Force Fields for Solid-State Electrolyte Materials

TL;DR

This paper discusses how to effectively train machine learning force fields for solid-state electrolytes, emphasizing data quality, model architecture, and benchmarking to improve battery materials.

Contribution

It provides practical guidelines for training ML force fields for SSEs, highlighting the importance of data quality and model architecture over dataset size.

Findings

Rigid SSE frameworks enable more efficient learning

Force RMSE does not correlate well with transport performance

Benchmarking frameworks help accelerate battery material development

Abstract

Machine learning force fields enable high-accuracy modeling of solid-state electrolytes (SSEs). This perspective evaluates dataset size, reference quality, and model architectures. We show that rigid SSE frameworks favor efficient learning, prioritizing data quality over quantity. Crucially, force RMSE does not reliably predict transport performance. By analyzing locality and benchmarking frameworks, we provide practical guidelines to accelerate the development of next-generation solid-state batteries.