This proposal concerns applications of machine-learned interatomic potentials (MLIPs) and continuation calculations that started in our previous medium project (2025/5-183). The systems in focus are water structures "on top of" ceria (Cerium oxide) surfaces. Ceria has intriguing physical and chemical properties and is an essential part of many high-tech devices. In many of these applications water is present, by design or as an unwanted ingredient. The objective of our planned simulations is to find the characteristics of the interactions and structures found at the interface between water and different ceria surfaces by means of MD simulations and, of course, adequate post-processing. One goal is to be able to perform droplet calculations, i.e. deposit water on the surfaces and examine the water spreading tendency on the surface, as nano-ceria has been found experimentally to be water-repellent, a property that is technically exploitable. Such simulations require very long and large-size (the droplets must not touch initially) simulations, which would be unfeasible using ab initio MD simulations. The calculations planned for this project are as follows. 1. Perform static and dynamic water/solid DFT calculations (using CP2K) to fill the "holes" in our existing training datasets (collected during previous proposal). We already have a reasonably extensive dataset to start with as we have performed many smaller-scale MD simulations in the past and collected trajectories. 2. Training step: The MLIP training will be performed on Dardel. This requires the n2p2 package (already installed on Dardel). Typical MLIP training at the start requires approximately 1000 CPU hours, increasing to 5000 CPU hours as the dataset size and the number of atoms in the system grow. Correspondingly, memory usage also increases, requiring the use of fat nodes. 3. MLIPs will be tested by running MD simulation using lammps. The primary test consists of checking for stability of the MD trajectories. Snap-hots from failing trajectories will be used to augment the training-set in an active learning process where we go back to step 1. Once stable trajectories are obtained we can compare the results to the DFT MD results and decide whether or not to add more data or move to point 4. 4. Perform water-droplet-on-surface production MD simulations with the new MLIP and monitor the interface structures and water spreading/confinement. 5. Perform static DFT calculations to examine some of the most valuable electronic properties such as charge distributions, which get lost in the MLIP approach. The extended project will result in: Performing 1–2 additional refinement steps (stages 1–3) by adding more previously unseen structures will allow us to generate high-quality datasets and develop universal MLIPs that can be reused in related studies of ceria/water interfaces. This includes developing and validating MLIPs for water–ceria interactions across multiple surface orientations. Finally, the trained and validated MLIPs will be used for large-scale simulations aimed at revealing the microscopic origins of ceria’s hydrophobic behavior (stages 4–5).