tfp.bijectors.Sigmoid | TensorFlow Probability
www.tensorflow.org › python › tfpIf the low and high parameters are not passed, the transformation is equivalent to Y = g (X) = 1 / (1 + exp (-X)), the same as tf.sigmoid. If the low and high parameters are passed, the transformation is equivalent to low + (high - low) * g (X) (with g (X) as defined above), a sigmoid that is shifted and scaled along the output axis.