What's meaning of "n tensors" in TensorBoard graph?

I'm reading the TensorFlow tutorial code mnist_deep.py and save the graph.

The ouput of scope fc1 should have shape [-1, 1024]. But it's 2 tensors in the graph in TensorBoard.

  # Fully connected layer 1 -- after 2 round of downsampling, our 28x28 image
  # is down to 7x7x64 feature maps -- maps this to 1024 features.
  with tf.name_scope('fc1'):
    W_fc1 = weight_variable([7 * 7 * 64, 1024])
    b_fc1 = bias_variable([1024])

    h_pool2_flat = tf.reshape(h_pool2, [-1, 7*7*64])
    h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)

  # Dropout - controls the complexity of the model, prevents co-adaptation of
  # features.
  with tf.name_scope('dropout'):
    keep_prob = tf.placeholder(tf.float32)
    h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)

Solution

It should mean that the output tensor of Relu is used twice in the Droupout node. If you try expanding it you should see the input go into two different nodes.