The Neural Encoding Dataset


The Neural Encoding Dataset (NED) consists of trained encoding models of fMRI and EEG responses to images of multiple subjects. The accompanying Python toolbox contains simple functions allowing you to use these models to synthesize fMRI and EEG responses to any image of your choice. NED also comes with pre-generated synthetic fMRI and EEG responses for ~150,000 naturalistic images, which you can use for research purposes. If you wish to contribute to improving and expanding NED (e.g., with more accurate encoding models, or new neural datasets), then please get in touch with Ale ( all feedback and help is strongly appreciated!