Smooth Dynamic Cutoffs for Machine Learning Interatomic Potentials

2026-01-29

AI Summary

This work addresses the high computational cost and memory consumption of machine learning interatomic potentials (MLIPs) in molecular dynamics simulations. The authors propose a dynamic cutoff radius method that adaptively assigns a neighbor count to each atom, enabling efficient graph sparsification while preserving long-term simulation stability. By introducing, for the first time, a variable and smoothly varying cutoff mechanism, this approach overcomes the limitations of conventional fixed-radius cutoffs, achieving significant gains in efficiency without sacrificing accuracy. Evaluated across four state-of-the-art MLIP models—MACE, NequIP, OrbV3, and TensorNet—the method yields an average 2.26× reduction in memory usage and a 2.04× speedup in inference, while maintaining near-lossless prediction accuracy on both materials and molecular datasets.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Humans and AI: Human-in-the-loop Machine Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Machine learning interatomic potentials (MLIPs) have proven to be wildly useful for molecular dynamics simulations, powering countless drug and materials discovery applications. However, MLIPs face two primary bottlenecks preventing them from reaching realistic simulation scales: inference time and memory consumption. In this work, we address both issues by challenging the long-held belief that the cutoff radius for the MLIP must be held to a fixed, constant value. For the first time, we introduce a dynamic cutoff formulation that still leads to stable, long timescale molecular dynamics simulation. In introducing the dynamic cutoff, we are able to induce sparsity onto the underlying atom graph by targeting a specific number of neighbors per atom, significantly reducing both memory consumption and inference time. We show the effectiveness of a dynamic cutoff by implementing it onto 4 state of the art MLIPs: MACE, Nequip, Orbv3, and TensorNet, leading to 2.26x less memory consumption and 2.04x faster inference time, depending on the model and atomic system. We also perform an extensive error analysis and find that the dynamic cutoff models exhibit minimal accuracy dropoff compared to their fixed cutoff counterparts on both materials and molecular datasets. All model implementations and training code will be fully open sourced.

Problem

• machine learning interatomic potentials
• inference time
• memory consumption
• molecular dynamics simulations

Innovation

• dynamic cutoff
• machine learning interatomic potentials
• sparsity
• molecular dynamics
• inference acceleration