bitwise_xor_op

→TensorImplPtr

int bitwise_xor_op(const int & a, const int & b)

Element-wise bitwise XOR.

Computes out[i] = a[i] ^ b[i] for every element. Defined on integer and Bool dtypes; on Bool inputs this is equivalent to logical exclusive-or.

Non-differentiable: no autograd node is registered.

Math

\mathrm{out}[i] = a[i] \,\oplus_{\text{bit}}\, b[i].

References

NumPy numpy.bitwise_xor.

See Also

bitwise_and_op, bitwise_or_op

a, bTensorImplPtr

Operands of identical shape, integer-or-Bool dtype, and device.

TensorImplPtr

Tensor of the same shape and dtype as the inputs.

LucidError

If shapes/devices differ, or if the input dtype is not integer or Bool.