-- https://towardsdatascience.com/mnist-handwritten-digits-classification-from-scratch-using-python-numpy-b08e401c4dab
-- def softmax(x):
--      exp_element=np.exp(x-x.max())
--      return exp_element/np.sum(exp_element,axis=0)
λxs.
  { m ⟜ (⋉)/* _1 xs; a ⟜ [e:(x-m)]`{0} xs
  ; sum ← [(+)/x]
  ; n ⟜ sum`{1∘[1]} (a::M float)
  ; ⍉([(%x)'y]`{0,1∘[1]} n a)
  }