PyTorch .logical_xor()

The .logical_xor() function in PyTorch computes the element-wise logical XOR (Exclusive OR) of two input tensors. The resulting tensor contains Boolean (True or False) values.

According to the XOR rule, an element is True only if exactly one of the corresponding inputs is truthy. Inputs are interpreted as Boolean values where non-zero means True and zero means False.

Syntax

torch.logical_xor(input, other, out)

Parameters:

• input: A Boolean tensor (or a tensor that can be cast to Boolean).
• other: A Boolean tensor of the same shape or broadcastable to input.
• out (optional): A tensor for storing the result in-place.

Return value:

Returns a Boolean tensor where each element is True when exactly one of the corresponding elements in input and other is True.

Example 1

This example shows the logical XOR operation applied to two 2×2 tensors:

import torch

# Tensor A:
# [True, False]
# [True, True]
tensor_a = torch.tensor([[1, 0], [5, 10]])

# Tensor B:
# [True, True]
# [False, True]
tensor_b = torch.tensor([[2, 3], [0, 1]])

# Compute logical XOR: A XOR B
# -----------------------------
# [1 XOR 2] -> True XOR True -> False
# [0 XOR 3] -> False XOR True -> True
# [5 XOR 0] -> True XOR False -> True
# [10 XOR 1] -> True XOR True -> False
result_tensor = torch.logical_xor(tensor_a, tensor_b)

print("Tensor A:\n", tensor_a)
print("\nTensor B:\n", tensor_b)
print("\nA.logical_xor(B) Result (Boolean):\n", result_tensor)

Copy to clipboard
Copy to clipboard

The output of this code is:

Tensor A:
 tensor([[ 1,  0],
        [ 5, 10]])

Tensor B:
 tensor([[2, 3],
        [0, 1]])

A.logical_xor(B) Result (Boolean):
 tensor([[False,  True],
        [ True, False]])

Copy to clipboard
Copy to clipboard

Example 2

This example demonstrates XOR on two 1-D tensors so each result maps cleanly to a single pair of elements:

import torch

x = torch.tensor([1, 0, 10, 0])
y = torch.tensor([0, 5, 0, 0])

# Compute x XOR y
xor_result = torch.logical_xor(x, y)

print(xor_result)

Copy to clipboard
Copy to clipboard

The output of this code is:

tensor([ True,  True,  True, False])

Copy to clipboard
Copy to clipboard