I was wondering if Psi4 plans on implementing the “no S^2” approximation when computing second order exchange-induction energies. It seems like the S^2 approximation for exchange-induction breaks down for strongly polarizing systems, and it would be very helpful to have the no S^2 option for exchange-induction.
It is available, please see SAPT: Symmetry-Adapted Perturbation Theory
It is currently a bit awkward to access it as it runs through the sapt(dft) code, the provided setting should give sapt0 results.