Oh I see, perhaps I misread. In any case, from what I understand neuroscience needs better experimental methods, not better math, so still sort of an odd challenge—unless the point is have "brain-like" function without regard to how human brains actually work, which is something I could imagine DARPA being very interested in.