Why does matrix multiplication in NumPy not call the __add__ method?

Question

How does the method __mul__ in the class matrix in NumPy work? I want to implement a binary matrix multiplication, and I have a class Binary :

class Binary(int):
    def __init__(self, val):
        if val != 0:
            self.val = 1
        else:
            self.val = 0

    def __add__(self, other):
        print('add')
        return self.val ^ other

    def __radd__(self, other):
        print('radd')
        return self.val ^ other

My test:

from Binary import Binary
from numpy import matrix

i = Binary(1)
o = Binary(0)
a = matrix([i, i, o, i, i, o, o], dtype=Binary)
b = matrix([[o, o, i],
           [o, i, o],
           [o, i, i],
           [i, o, o],
           [i, o, i],
           [i, i, o],
           [i, i, i]], dtype=Binary)
print(a * b)

Result:

./test.py
[[2 1 2]]

The method __add__ is not used. Whereas there is a summation in matrix multiplication.