Teaching Molecular Dynamics to a Non-Autoregressive Ionic Transport Predictor

2026-05-10

AI Summary

Existing approaches struggle to balance efficiency and accuracy in predicting ionic transport properties: molecular dynamics simulations are computationally expensive, while current learning models are either slow or inaccurate. This work proposes a novel non-autoregressive learning framework that, for the first time, incorporates atomic trajectories as an auxiliary modality during training, enabling the model to efficiently predict dynamic transport properties at inference time without requiring trajectory inputs. The method seamlessly accommodates both trajectory-inclusive and trajectory-free data and significantly outperforms existing non-autoregressive baselines on two benchmark datasets, achieving lower prediction errors and accelerating inference by over 200× compared to autoregressive models.

Abstract

Unlike most static material properties widely studied in the machine learning literature, ionic transport properties are inherently dynamic, making their fast and accurate prediction from static atomic structures challenging. The current standard approach, molecular dynamics (MD) simulations, suffers from prohibitively high computational cost. Recent autoregressive learning-based MD acceleration methods requiring sequential inference remain slow and prone to error accumulation; in contrast, existing non-autoregressive material property prediction models are less accurate because they fail to exploit dynamics. Moreover, existing methods typically benefit from datasets either with or without atomic trajectories, but not both. To overcome these limitations, we propose a non-autoregressive learning framework based on auxiliary modality learning, which treats atomic trajectories as an auxiliary modality during training but does not require them at inference. This enables the predictor to learn dynamics without sequential inference while benefiting from both types of datasets. As a result, our framework achieves over 200 times speedup compared to autoregressive models on the dataset with atomic trajectories while substantially reducing prediction error relative to non-autoregressive benchmarks across both types of datasets. Our code is available at https://github.com/jykim-git/MD.

Problem

• ionic transport
• molecular dynamics
• non-autoregressive prediction
• machine learning
• atomic trajectories

Innovation

• non-autoregressive learning
• ionic transport prediction
• auxiliary modality learning
• molecular dynamics acceleration
• dynamic property prediction