Attahced are the training dataset and potential files for the publication "SiC-TGAP: A machine learning interatomic potential for radiation damage simulations in 3C-SiC". The description of the uploaded files is as follows: dataset-SiC-TGAP.xyz: The reference training dataset, in which the atomic structures are stored in the "extxyz" format. Each structure is tagged with the total energy, atomic force components, and virial stresses, calculated with the VASP code using the PBE functional. Each structure is labeled with the "config_type" keyword, which defines the category to which the structure belongs. An additional "sub_config" keyword provides more information about the structure. The dataset contains 3000 atomic structures and 258543 atomic environments. SiC-TGAP_LAMMPS: The potential files for LAMMPS. SiC-TGAP_TurboGAP: The potential files and additional required files to use the potential in the TurboGAP code. To use the potential with TurboGAP, the associated files should be stored in (and called from) the "gap_files" directory. The unique identifier of the potential is: "GAP_2025_5_5_180_5_9_6_648".

2025