Machine-learning interatomic potentials (CMD) and density-functional ab initio molecular dynamics (AIMD) are used to investigate the mechanical properties and phase stability of novel high-entropy ceramics and superlattice structures up to temperatures (>1000 K) or relevance for practical uses. The project is a theoretical-experimental collaboration between LiU, TUWien (Austria), Comenius University (Slovakia) and University of California San Diego and includes a large part of activities of FunMat II consortium. The theoretical investigations are based on methods developed by the PI. Financial support: VR Etablering Grant Nº VR-2021-04426, VINN Excellence Center Functional Nanoscale Materials FunMat-2 prolonged to 2027 (Grant 2022-03071), Olle Engkvist Foundation, Austrian Academy of Sciences, ‎ÖAW, via the DOC fellowship, KUWI grant from TU Wien, Hertha Firnberg Programme. During 2021-present, we published 19 papers +3 in review in high-quality journals including Science Advances, Nature Communications, Physical Review Materials, Acta Materialia, and npj Computational Materials. SNIC and NAISS resources are acknowledged in all publications (see activity report).