The dressed state Hamiltonian is given by
$$ \hat{H}{\lambda} = \begin{bmatrix} \lambda+ && i\hbar\frac{\dot{\theta}}{2}\\ -i\hbar\frac{\dot{\theta}}{2} && \lambda_-\end{bmatrix} $$
An off-diagonal term $\hbar\dot{\theta}/2 \hat{\sigma}_y$ drives the diabatic process, where
$$ \dot{\theta} = \frac{-\dot{\Delta}\Omega+\Delta\dot{\Omega}}{\Delta^2+\Omega^2} $$
The approach of the RAP is to make $\dot{\Delta}=0$ at every point where it makes $\dot{\theta}=0$. It requires a huge amount of time $T_{\rm adia}$, where it satisfy $\dot{\Delta}=(\Delta_f-\Delta_i)/T_{\rm adia}=0$.
The shortcut-to-adiabaticity (STA) method is developed to reduce the time budget while maintaining the adiabatic character. The STA aims to make $\dot{\theta}=0$, by properly tuning time-dependent variables $(\Omega,\Delta)$.
XXXX
A. Lukin et al., Quantum Quench Dynamics as a Shortcut to Adiabaticity, arXiv:2405.21019.