DEQuify your force field: More efficient simulations using deep equilibrium models
This work addresses the need for faster and more accurate machine learning force fields for molecular dynamics simulations, offering incremental improvements by leveraging prior information about system continuity.
The paper tackled the problem of inefficient molecular dynamics simulations by exploiting the continuity of molecular states, recasting an equivariant base model as a deep equilibrium model to recycle features from previous time steps. This resulted in a 10%-20% improvement in accuracy and speed on MD17, MD22, and OC20 200k datasets, along with more memory-efficient training.
Machine learning force fields show great promise in enabling more accurate molecular dynamics simulations compared to manually derived ones. Much of the progress in recent years was driven by exploiting prior knowledge about physical systems, in particular symmetries under rotation, translation, and reflections. In this paper, we argue that there is another important piece of prior information that, thus fa,r hasn't been explored: Simulating a molecular system is necessarily continuous, and successive states are therefore extremely similar. Our contribution is to show that we can exploit this information by recasting a state-of-the-art equivariant base model as a deep equilibrium model. This allows us to recycle intermediate neural network features from previous time steps, enabling us to improve both accuracy and speed by $10\%-20\%$ on the MD17, MD22, and OC20 200k datasets, compared to the non-DEQ base model. The training is also much more memory efficient, allowing us to train more expressive models on larger systems.