Assessing zero-shot generalisation behaviour in graph-neural-network interatomic potentials

With the rapidly growing availability of machine-learned interatomic potential (MLIP) models for chemistry, much current research focuses on the development of generally applicable and "foundational" MLIPs. An important question in this context is whether, and how well, such models can transfer from one application domain to another. Here, we assess this transferability for an MLIP model at the interface of materials and molecular chemistry. Specifically, we study GO-MACE-23, a model designed for the extended covalent network of graphene oxide, and quantify its zero-shot performance for small, isolated molecules outside its direct scope, as well as for examples of chemical reactions. Our work provides quantitative insight into the generalisation ability of graph-based MLIP models and, by exploring their limits, can help to inform future developments.