diff --git a/tensorflow/models/image/cifar10/cifar10.py b/tensorflow/models/image/cifar10/cifar10.py
index fb3a42cbb13..7df2149d40d 100644
--- a/tensorflow/models/image/cifar10/cifar10.py
+++ b/tensorflow/models/image/cifar10/cifar10.py
@@ -256,7 +256,10 @@ def inference(images):
     local4 = tf.nn.relu(tf.matmul(local3, weights) + biases, name=scope.name)
     _activation_summary(local4)
 
-  # softmax, i.e. softmax(WX + b)
+  # linear layer(WX + b),
+  # We don't apply softmax here because 
+  # tf.nn.sparse_softmax_cross_entropy_with_logits accepts the unscaled logits 
+  # and performs the softmax internally for efficiency.
   with tf.variable_scope('softmax_linear') as scope:
     weights = _variable_with_weight_decay('weights', [192, NUM_CLASSES],
                                           stddev=1/192.0, wd=0.0)