Each sample groups how it was prepared, characterized, and measured. Click a card to expand.
Simulated twisted bilayer graphene moiré supercell under hydrostatic pressure in the flat-band regime, used to generate Hartree-Fock LDOS and FT-LDOS images for regression of the Hubbard U parameter.
Each sample groups how it was prepared, characterized, and measured. Click a card to expand.
Simulated twisted bilayer graphene moiré supercell under hydrostatic pressure in the flat-band regime, used to generate Hartree-Fock LDOS and FT-LDOS images for regression of the Hubbard U parameter.
Each sample groups how it was prepared, characterized, and measured. Click a card to expand.
Simulated twisted bilayer graphene moiré supercell under hydrostatic pressure in the flat-band regime, used to generate Hartree-Fock LDOS and FT-LDOS images for regression of the Hubbard U parameter.
Each sample groups how it was prepared, characterized, and measured. Click a card to expand.
Simulated twisted bilayer graphene moiré supercell under hydrostatic pressure in the flat-band regime, used to generate Hartree-Fock LDOS and FT-LDOS images for regression of the Hubbard U parameter.
Tight-binding model for twisted bilayer graphene with long-range Coulomb interaction and on-site Hubbard term; self-consistent restricted Hartree-Fock solution; Fourier transform of LDOS to generate FT-LDOS images.FORGE package
Convolutional-neural-network regression on FT-LDOS images using a custom CNN and a modified ResNet-18 pretrained on ImageNet; trained with mean squared error loss and Adam optimizer; evaluated with MAE, RMSE, and R2.PyTorch
Tight-binding model for twisted bilayer graphene with long-range Coulomb interaction and on-site Hubbard term; self-consistent restricted Hartree-Fock solution; Fourier transform of LDOS to generate FT-LDOS images.FORGE package
Convolutional-neural-network regression on FT-LDOS images using a custom CNN and a modified ResNet-18 pretrained on ImageNet; trained with mean squared error loss and Adam optimizer; evaluated with MAE, RMSE, and R2.PyTorch
Tight-binding model for twisted bilayer graphene with long-range Coulomb interaction and on-site Hubbard term; self-consistent restricted Hartree-Fock solution; Fourier transform of LDOS to generate FT-LDOS images.FORGE package
Convolutional-neural-network regression on FT-LDOS images using a custom CNN and a modified ResNet-18 pretrained on ImageNet; trained with mean squared error loss and Adam optimizer; evaluated with MAE, RMSE, and R2.PyTorch
Tight-binding model for twisted bilayer graphene with long-range Coulomb interaction and on-site Hubbard term; self-consistent restricted Hartree-Fock solution; Fourier transform of LDOS to generate FT-LDOS images.FORGE package
Convolutional-neural-network regression on FT-LDOS images using a custom CNN and a modified ResNet-18 pretrained on ImageNet; trained with mean squared error loss and Adam optimizer; evaluated with MAE, RMSE, and R2.PyTorch
