fix typo in docs

Author: ytdHuang

docs/src/users_guide/time_evolution/mesolve.md

@@ -38,8 +38,8 @@ tlist = LinRange(0.0, 10.0, 20)
 sol = mesolve(H, state0, tlist, e_ops = [sigmaz()])
 ```
 
-!!! note "Type of initial state"
-    The initial state `state0` here can be given as a state vector ``|\psi(0)\rangle`` (in the type of [`Ket`](@ref)) or a density matrix ``\hat{\rho}(0)`` (in the type of [`Operator`](@ref)). If it is given as a [`Ket`](@ref), it will be transformed to density matrix ``\hat{\rho}(0) = |\psi(0)\rangle\langle\psi(0)|`` internally in [`mesolve`](@ref).
+!!! note "Use sesolve for improved efficiency"
+    Here, if the Hamiltonian `H` is given as an [`Operator`](@ref), and the initial state `state0` is given as a state vector ``|\psi(0)\rangle`` (in the type of [`Ket`](@ref)), it will automatically call [`sesolve`](@ref) for improved efficiency.
 
 The function returns [`TimeEvolutionSol`](@ref), as described in the previous section [Time Evolution Solutions](@ref doc-TE:Time-Evolution-Solutions). The stored `states` will always be in the type of [`Operator`](@ref) (density matrix).