Generative feedback explains distinct brain activity codes for seen and mental images

Notebooks and code

gabor_fwrf_model_training

A notebook to train the model parameters with k-fold validation for vision and imagery components of the experiment.

Input:

• Stimuli.h5py
• Stimuli_metadata.pkl
• Voxels.h5py
• Voxels_metadata.pkl

Produces a model parameter file:

• fwrf_{subject}_{timestamp}_data.pkl

linear_brain_model

Create a linear brain model from a set of structural prescriptions and perform vision and imagery-hypothetized inference from different area and explore its effects.

Produces:

• Figures