Understanding simplified Recurrent Neural network

How is Matrix multiplication working in this case (in reference to equation of a on left)?
Waa is 100x100 and a is 100x1 so, this would output a 100x1 Matrix
Now, Wax is 100x10,000 an x is 10,000x1 so, this would output a 10,000x1 Matrix
so, how can two matrix of different shapes be added ?
Maybe I am going entirely in wrong direction, please excuse for any stupid question.

Please check the dimensions again: if you multiply 100 x 10,000 dot 10,000 x 1, the result is 100 x 1, right? So the shapes should be compatible.

1 Like

You maybe familiar with this already, in matrix multiplication the edge dimensions are the dimensions of resultant matrix. Like you have : 100 * 10000 and 10000 * 1
resultant will have dimensions 100*1

2 Likes

thank you both!