tf.math.pow

TensorFlow 1 version View source on GitHub

Computes the power of one value to another.

Aliases: tf.compat.v1.math.pow, tf.compat.v1.pow, tf.compat.v2.math.pow, tf.compat.v2.pow, tf.pow

tf.math.pow(
    x,
    y,
    name=None
)

Given a tensor x and a tensor y, this operation computes \(x^y\) for corresponding elements in x and y. For example:

x = tf.constant([[2, 2], [3, 3]])
y = tf.constant([[8, 16], [2, 3]])
tf.pow(x, y)  # [[256, 65536], [9, 27]]

Args:

  • x: A Tensor of type float16, float32, float64, int32, int64, complex64, or complex128.
  • y: A Tensor of type float16, float32, float64, int32, int64, complex64, or complex128.
  • name: A name for the operation (optional).

Returns:

A Tensor.