for-loop in function does not append arrays while same loop outside function does append arrays

I am new to python. I am trying to understand the behavior of a function.

I have two arrays:

1. a_indices, a 5x2 array for which each row contains row and column indices of specific cells of a matrix (not relevant for this question)
2. a_template a 9x2 array containing values that should be subtracted from each row of 1) in order to get the indices of the neighboring cells

To subtract a_template from each row of a_indices I cooked up the following code:

import numpy as np

# define arrays
a_indices = np.array([[0, 1],[1, 2],[2, 2],[0, 2],[1, 2]])
a_template = np.array([[-1,-1],[-1,0],[-1,1],[ 0,-1],[0,  0],[ 0,  1],[ 1, -1],[ 1,  0],[ 1,  1]])

# add a_template to each row of a_indices
a1 = []
for x in a_indices:
    a1.append(np.add(x,a_template))

print(a1)

Starting with an empty list it appends arrays with added values for each row of a_indices. This gives a list of 5 numpy arrays of 2x9, containing the values I need. Good!

However I will produce a lot of these a_indices arrays and they will likely have different numbers of rows. Therefore I wanted to wrap the above for loop into a function that should return my desired objects. So I tried this:

def find_neighbors(idc):
    b = []
    for x in idc:
        b.append(np.add(x,a_template))
        return b

a2 = find_neighbors(a_indices)
print(a2)

The function just contains the for loop from the above example. I expected that this function returns the same object as a1. However it returns a list of only one array (the same as the first array from a1).

My question is why does my function not return a list for each row of a_indices and of course how can I fix this?