Δ-Learning of High-Fidelity Electronic Structure Using Graph Neural Networks with Modified Node-Level Features
Published in ACS Materials Letters, 2025
This paper presents a Δ-learning approach using Graph Neural Networks (GNNs) with modified node-level features to predict hybrid functional HSE06 eigenvalues from PBE inputs for metal and nitrogen doped graphene catalysts.
Recommended citation: Karimitari, N., Pakornchote, T., Alherz, A. W., Clary, J. M., Tezak, C., Dey, S., et al. (2025). "Δ-Learning of High-Fidelity Electronic Structure Using Graph Neural Networks with Modified Node-Level Features." ACS Materials Letters.
Download Paper
