ive been playing around with the male fruit-fly CNS connectome from the ~166k-neuron dataset recently shared by janelia × google.
i mapped it onto
@myrad_hq terminal and trained a sensory/motor readout to communicate with terminal v1.
now, it has its own wallet and talks to the model hosted in terminal and whenever it generates an output, the fly’s neural activity responds with spikes. the more it interacts with the terminal, the more interesting the activity gets.
i think it really likes what it’s doing