Matrix Multiplication for 1D

[sociopath00]

In the course, we have tensor = torch.tensor([1,2,3]) which has 1D only. In this case how matrix multiplication works. As per the rules it should have same inner dimensions. When we do torch.matmul(tensor, tensor) how the 2 rules of matrix multiplications are applied here?

[AlphaGeek]

I believe that behind the scenes torch just returns the dot product of the two vectors

[lishav123]

Hello, There!
At first, I also didn't get it, I thought maybe it's doing kind of vector operation, like .dot or somthing, because, that's honestly dosen't make any sense to me that how matmul is breaking the multiplying rule of matrices, that, why it executed without any error at first place?

I was unsure about it too, but it seems that it was a bug, which is now fixed, I am using Pytorch version 2.7.1+cu126 and I have tried this code in this version

import torch
tensor = torch.tensor([1, 2, 3])
print(torch.matmul(tensor, tensor))

Output was:

RuntimeError: mat1 and mat2 shapes cannot be multiplied (1x3 and 1x3)

So, as in course it's said that for matrix multiplication inner dimention must be same, so I tried this

import torch
tensor = torch.tensor([1, 2, 3])
print(torch.matmul(tensor, tensor.T)) # After transpose

Output was:

tensor([[14]])

Inshort, yep it dosen't make sense in video that matmul works there, but it's just probabily a bug(which is fixed now) or they might wanted to make it more robust!