If both a and b are 2-D arrays, it is matrix multiplication, but using matmul() or a @ b is preferred. If either a or b is 0-D (scalar), it is equivalent to multiply() ...
numpy-matmul-vs-dot
3925e8d270
numpy matmul, numpy matmul vs dot, numpy matmul vs @, numpy matmul function, numpy matmul broadcasting, numpy matmul along axis, numpy matmul source code, numpy matmul 3 matrices, numpy matmul operator, numpy matmul complexity, numpy matmul vs @ operator, numpy matmul order, numpy matmul shortcut