QM9VASP, QM9Psi4 and Benchmark Datasets

・Nanoparticle_CO.tar.gzBenchmark dataset and results・QM9_VASP_Psi4.tar.gzQM9VASP, QM9Psi4 including TEA codes.<br>・fig2_qm9_psi4_vasp_mf_training.tar.gzThis archive contains the training outputs, evaluation tables, and plotting scripts used to generate Figure 2. It summarizes QM9 PSI4/VASP multi-fidelity training results across different target-level data percentages and alignment settings.<br>・fig3_scaling_factor_determination.tar.gzThis archive contains the scripts, intermediate files, and summary plots used to generate Figure 3. It includes the determination and stability analysis of scaling factors for mapping several source datasets to the VASP target across different sample sizes.<br><br>```<br>@article{ramakrishnan2014quantum,<br>title={Quantum chemistry structures and properties of 134 kilo molecules},<br>author={Ramakrishnan, Raghunathan and Dral, Pavlo O and Rupp, Matthias and Von Lilienfeld, O Anatole},<br>journal={Scientific data},<br>volume={1},<br>number={1},<br>pages={1--7},<br>year={2014},<br>publisher={Nature Publishing Group}<br>}<br>```<br><br>```<br>@article{shiota2025lowering,<br>title={Lowering the exponential wall: accelerating high-entropy alloy catalysts screening using local surface energy descriptors from neural network potentials},<br>author={Shiota, Tomoya and Ishihara, Kenji and Mizukami, Wataru},<br>journal={Digital Discovery},<br>volume={4},<br>number={3},<br>pages={738--751},<br>year={2025},<br>publisher={Royal Society of Chemistry}<br>}<br>```<br><br>```@misc{shiota2024tamingmultidomainfidelitydata,<br>title={Taming Multi-Domain, -Fidelity Data: Towards Foundation Models for Atomistic Scale Simulations},<br>author={Tomoya Shiota and Kenji Ishihara and Tuan Minh Do and Toshio Mori and Wataru Mizukami},<br>year={2024},<br>eprint={2412.13088},<br>archivePrefix={arXiv},<br>primaryClass={physics.chem-ph},<br>url={https://arxiv.org/abs/2412.13088},<br>}<br>```