Search papers, labs, and topics across Lattice.
The authors introduce mlip v2, an open-source library designed to enhance the efficiency and scalability of molecular simulations using machine learning interatomic potentials (MLIPs). Key improvements include a redesigned API for modularity, a high-performance backend (e3j) for faster inference, and the introduction of the eSEN architecture with Mixture-of-Experts for scalable training. The library also features improved electrostatics handling and advanced simulation methods like NPT ensembles.
Open-source mlip v2 slashes the computational cost of atomistic simulations while maintaining near ab initio accuracy, thanks to a high-performance backend and scalable Mixture-of-Experts architecture.
Machine learning interatomic potentials (MLIPs) enable atomistic simulations with near ab initio accuracy at significantly reduced computational cost, but their broader adoption is often limited by fragmented tooling, limited scalability, and inflexible software design. We present mlip v2, a new generation of the mlip library that advances efficient and scalable molecular simulation through a unified and extensible framework. The new release features a targeted API redesign with improved modularity and control, enabling flexible customization of training, data processing, and simulation workflows. It further integrates a new high-performance backend for equivariant operations, e3j, significantly accelerating model inference and simulations. In addition, the framework introduces a range of entirely new capabilities, including the eSEN architecture with a Mixture-of-Experts formulation for scalable training on large and diverse datasets, improved handling of electrostatics through more physically grounded charge modeling and long-range interaction treatment, and advanced simulation features such as NPT ensembles and nudged elastic band methods. Together, these extensions significantly broaden the scope of MLIP applications, enabling efficient modeling of complex, reactive, and out-of-equilibrium systems, and bridging the gap between ML research and practical molecular simulation applications. The library is available on GitHub and on PyPI under the Apache license 2.0.
