What Does the "Greater Than" Operator ">" Mean for Pytorch Tensors?

Using the > operator is the same as using the torch.gt() function.

In other words,

it > 0

is the same as

torch.gt(it, 0)

and it returns a ByteTensor (a Boolean tensor) of the same shape as it where out[i] is True if it[i] > 0 and False otherwise.

As seen from the question it is a 1-d tensor consisting of 10 elements. When we write it > 0 each element of the tensor it is compared with 0 and the output is set as True or False depending upon whether the number is greater than 0 or not. The result is also a 1-d boolean tensor consisting of True and False values. In your case, you will get a 1-d tensor like this: [False, False, False, False, False, False, False, False, False, False,] as none of the elements in it is equal to 0.

In more simpler terms,

If result is a variable (actually a 1-d tenosr) to store the output having the same shape as it then the equation result= it > 0 can be written as:

if it[i]>0:
    result[i]= True
else:
    result[i]= False 

But when it's performed as result= it > 0, the execution is much faster than writing our own for/while loop.

Hope this helps you.