10.1101/2020.06.11.147272

Fourier-transform-based attribution priors improve the interpretability and stability of deep learning models for genomics