DINHR trained model checkpoints for CryoBench datasets and experimental cryo-EM datasets

This record provides trained DINHR (Denoising Implicit Neural High-order Representation) model checkpoints and auxiliary files to reproduce the results reported in our manuscript. The release includes checkpoints for CryoBench simulated datasets (IgG-1D at multiple SNRs, IgG-RL, TomoTwin-100, Ribosembly) and experimental datasets (EMPIAR-10073, EMPIAR-10076, INO80–Hexasome).Each subfolder contains the released .pt checkpoints (DINHR and INHR baselines where applicable) and dataset-specific auxiliary files (e.g., list_refs.txt, *.npy) used by our evaluation scripts.Code and usage instructions are available in the DINHR GitHub repository.