SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects

TL;DR

SpookyNet is a deep neural network that models force fields with explicit electronic degrees of freedom and nonlocal effects, improving accuracy and generalization in quantum chemistry applications.

Contribution

It introduces a novel architecture incorporating electronic degrees of freedom and quantum nonlocality, addressing limitations of previous ML-FFs.

Findings

Achieves state-of-the-art performance on quantum chemistry datasets.

Generalizes across chemical and conformational space.

Can predict unknown spin states using learned chemical insights.

Abstract

Machine-learned force fields (ML-FFs) combine the accuracy of ab initio methods with the efficiency of conventional force fields. However, current ML-FFs typically ignore electronic degrees of freedom, such as the total charge or spin state, and assume chemical locality, which is problematic when molecules have inconsistent electronic states, or when nonlocal effects play a significant role. This work introduces SpookyNet, a deep neural network for constructing ML-FFs with explicit treatment of electronic degrees of freedom and quantum nonlocality. Chemically meaningful inductive biases and analytical corrections built into the network architecture allow it to properly model physical limits. SpookyNet improves upon the current state-of-the-art (or achieves similar performance) on popular quantum chemistry data sets. Notably, it is able to generalize across chemical and conformational space and can leverage the learned chemical insights, e.g. by predicting unknown spin states, thus helping to close a further important remaining gap for today's machine learning models in quantum chemistry.