Fixing 3 demos (#952)

`dev.state` does not work with the new device:
`variational_classifier`
`tutorial_noisy_circuits`

Typo:
`circuit_cutting`

demonstrations/tutorial_noisy_circuits.py

@@ -75,12 +75,18 @@ def circuit():
 print(f"QNode output = {circuit():.4f}")
 
 ######################################################################
-# The device stores the output state as a density matrix. In this case, the density matrix is
+# With a small modification of the circuit we can also ask for the density matrix. In this case, the density matrix is
 # equal to :math:`|\psi\rangle\langle\psi|`,
 # where :math:`|\psi\rangle=\frac{1}{\sqrt{2}}(|00\rangle + |11\rangle)`.
 
+@qml.qnode(dev)
+def density_matrix_circuit():
+    qml.Hadamard(wires=0)
+    qml.CNOT(wires=[0, 1])
+    return qml.state()
 
-print(f"Output state is = \n{np.real(dev.state)}")
+matrix = density_matrix_circuit()
+print(f"Output density matrix is = \n{np.real(matrix)}")
 
 ######################################################################
 # Incoherent noise is modelled by

demonstrations/tutorial_quantum_circuit_cutting.py

@@ -193,7 +193,7 @@ def circuit(x):
 
 
 ######################################################################
-# The double vertical lines between the two ``CZ`` gates on qubit 1 in the
+# The dashed line and the scissors between the two ``CZ`` gates on qubit 1 in the
 # above figure show where we have chosen to cut. This is where the ``WireCut``
 # operation is inserted. ``WireCut`` is used to manually mark locations for
 # wire cuts.

demonstrations/tutorial_variational_classifier.py

@@ -323,14 +323,14 @@ def test(angles):
 
     statepreparation(angles)
 
-    return qml.expval(qml.PauliZ(0))
+    return qml.state()
 
 
-test(ang)
+state = test(ang)
 
 print("x               : ", x)
 print("angles          : ", ang)
-print("amplitude vector: ", np.real(dev.state))
+print("amplitude vector: ", np.real(state))
 
 
 ##############################################################################
@@ -376,7 +376,7 @@ def cost(weights, bias, features, labels):
 
 ##############################################################################
 # Data
-# ~~~~
+# ~~~~ 
 #
 # We then load the Iris data set. There is a bit of preprocessing to do in
 # order to encode the inputs into the amplitudes of a quantum state. In