Lobe Wants to Preview a Video in a Brain

Field notes

Lobe describes a three-part pipeline: footage is processed by a model trained on fMRI recordings, the resulting brain-response signal is visualized over the video, and a planned second model is meant to connect that signal to retention metrics. The page currently demonstrates a 45-second attention trace and a rotating fsaverage cortical surface with Destrieux parcellation. It also says the retention-prediction layer is still being trained against creator analytics and that the present brain model covers only the cortex's outer layer.