Would be nice if we could see the same kind of chain of response from other models.

I’d love to see what other implicit biases other groups have built in to their models.