TY - CONF A1 - Grandel, Jonas T1 - Fine-tuning universal interatomic potentials for phonon properties without catastrophic forgetting N2 - Accurate phonon predictions are essential for evaluating material stability and thermal behavior. Traditional DFT-based methods are computationally intensive, driving interest in faster, machine-learning-based alternatives. In this work, we fine-tune the machine learning interatomic potential MACE-MP-0b3 (https://arxiv.org/abs/2401.00096) to improve the prediction of harmonic phonons and thermal properties. A key challenge is the need for highly accurate force calculations. While fine-tuning can enhance phonon accuracy, it may reduce generalizability to other properties. Despite this, our study shows that fine-tuned models can accurately predict phonon properties (including renormalization of imaginary phonons) at the same time as volume-dependent properties like the bulk modulus. We fine-tune the base model using small sets of rattled structures and introduce a novel fine-tuning method, benchmarking it against existing techniques. These results suggest that refined MACE-based universal potentials are a promising path for efficient and accurate phonon modeling. All implementations are included in the equitrain training code (https://github.com/BAMeScience/equitrain). T2 - Psi-k Konferenz CY - Lausanne, France DA - 25.08.2025 KW - Machine Learning KW - MACE KW - Phonon KW - Thermal properites PY - 2025 AN - OPUS4-64010 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - GEN A1 - Grandel, Jonas A1 - George, Janine T1 - Mace-mp-03b Phonon Benchmark N2 - This repository contains phonon calculations and evaluations for the MACE MP-0b3 model. First release of the scripts used for the phonon benchmark in the paper benchmarking the MACE-MP-03b model. See https://arxiv.org/abs/2401.00096 for a previous version of the paper. Full Changelog: https://github.com/JaGeo/mace-mp-03b-phonon-benchmark/commits/v0.0.1 KW - Machine Learned Interatomic Potentials KW - Phonons KW - Thermal Conductivity KW - Materials Searches KW - Foundation Model PY - 2025 DO - https://doi.org/10.5281/zenodo.15462975 PB - Zenodo CY - Geneva AN - OPUS4-63174 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Benner, Philipp A1 - Grandel, Jonas T1 - A shortcut towards phonon predictions N2 - Phonon calculations with ab-initio methods are computationally expensive. The use of universal machine learning models reduces the cost, but raises concerns about prediction quality. Fine-tuning with only a few structures, improves predictions of phonons, thermal properties and especially diffusive thermal conductivity, while reducing computational cost by a factor of 10 in average compared to DFT methods. T2 - DPG Frühjahrstagung CY - Regensburg, Germany DA - 16.03.2025 KW - Machine Learning KW - MACE KW - Phonons PY - 2025 AN - OPUS4-62770 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - GEN A1 - Grandel, Jonas T1 - Harmonic phonon dataset N2 - Phonon Dataset calculated with DFT PBE and the foundational MACE-MP-0b3 model. The data are presented and discussed in the following publication: https://doi.org/10.1063/5.0297006 KW - Phonon KW - MACE KW - DFT PY - 2025 UR - https://github.com/ACEsuit/mace-foundations PB - GitHub CY - San Francisco, CA, USA AN - OPUS4-64747 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - JOUR A1 - Batatia, Ilyes A1 - Benner, Philipp A1 - Chiang, Yuan A1 - Elena, Alin M. A1 - Kovács, Dávid P. A1 - Riebesell, Janosh A1 - Advincula, Xavier R. A1 - Asta, Mark A1 - Avaylon, Matthew A1 - Baldwin, William J. A1 - Berger, Fabian A1 - Bernstein, Noam A1 - Bhowmik, Arghya A1 - Bigi, Filippo A1 - Blau, Samuel M. A1 - Cărare, Vlad A1 - Ceriotti, Michele A1 - Chong, Sanggyu A1 - Darby, James P. A1 - De, Sandip A1 - Della Pia, Flaviano A1 - Deringer, Volker L. A1 - Elijošius, Rokas A1 - El-Machachi, Zakariya A1 - Fako, Edvin A1 - Falcioni, Fabio A1 - Ferrari, Andrea C. A1 - Gardner, John L. A. A1 - Gawkowski, Mikołaj J. A1 - Genreith-Schriever, Annalena A1 - George, Janine A1 - Goodall, Rhys E. A. A1 - Grandel, Jonas A1 - Grey, Clare P. A1 - Grigorev, Petr A1 - Han, Shuang A1 - Handley, Will A1 - Heenen, Hendrik H. A1 - Hermansson, Kersti A1 - Ho, Cheuk Hin A1 - Hofmann, Stephan A1 - Holm, Christian A1 - Jaafar, Jad A1 - Jakob, Konstantin S. A1 - Jung, Hyunwook A1 - Kapil, Venkat A1 - Kaplan, Aaron D. A1 - Karimitari, Nima A1 - Naik, Aakash A. A1 - Csányi, Gábor T1 - A foundation model for atomistic materials chemistry N2 - Atomistic simulations of matter, especially those that leverage first-principles (ab initio) electronic structure theory, provide a microscopic view of the world, underpinning much of our understanding of chemistry and materials science. Over the last decade or so, machine-learned force fields have transformed atomistic modeling by enabling simulations of ab initio quality over unprecedented time and length scales. However, early machine-learning (ML) force fields have largely been limited by (i) the substantial computational and human effort required to develop and validate potentials for each particular system of interest and (ii) a general lack of transferability from one chemical system to the next. Here, we show that it is possible to create a general-purpose atomistic ML model, trained on a public dataset of moderate size, that is capable of running stable molecular dynamics for a wide range of molecules and materials. We demonstrate the power of the MACE-MP-0 model—and its qualitative and at times quantitative accuracy—on a diverse set of problems in the physical sciences, including properties of solids, liquids, gases, chemical reactions, interfaces, and even the dynamics of a small protein. The model can be applied out of the box as a starting or “foundation” model for any atomistic system of interest and, when desired, can be fine-tuned on just a handful of application-specific data points to reach ab initio accuracy. Establishing that a stable force-field model can cover almost all materials changes atomistic modeling in a fundamental way: experienced users obtain reliable results much faster, and beginners face a lower barrier to entry. Foundation models thus represent a step toward democratizing the revolution in atomic-scale modeling that has been brought about by ML force fields. KW - Materials Design KW - Thermal Conducitivity KW - Nanoparticles KW - Batteries PY - 2025 UR - https://nbn-resolving.org/urn:nbn:de:kobv:b43-647829 DO - https://doi.org/10.1063/5.0297006 SN - 0021-9606 VL - 163 IS - 18 SP - 1 EP - 89 PB - AIP Publishing AN - OPUS4-64782 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER -