We present label gradient alignment, a novel algorithm for semi-supervised
learning which imputes labels for the unlabeled data and trains on the imputed
labels. We define a semantically meaningful distance metric on the input space
by mapping a point (x, y) to the gradient of the mode