Some of them are often wildly overstated, others are kept hidden. In polite society you can only mention a difference between male and female cognition if it either (a) advantages females or (b) it is strictly neutral. See also: Larry Summers.
Also there's some evidence of differences in cognition between different races; you definitely can't mention these in polite society so I'm not going to. See also: The Bell Curve, and that chick at Harvard Law who got reprimanded by the Dean for something she said in a private email.
Of course all these actual cognitive differences don't really matter since intragroup variability vastly swamps intergroup variability. The only interesting and significant differences between male and female minds are not in the way they solve abstract problems but in ... pretty much all the other stuff (emotional makeup, desires, responses to events, sociability, et cetera).
