zerokelvin The residual population in the excited state after a pi-pulse on my qubit didn’t match the Rabi model at all, even after calibrating the drive amplitude. Tried detuning the pulse by a few MHz and the leftover |0> fraction actually increased.
QCJunkie Anharmonicity in the qubit levels might be shifting things so your drive is leaking instead of just flipping |0> to |1>.
QMachineX If the drive strength is shifting the levels via AC Stark, the effective resonance moves with amplitude, so a static detune scan won’t catch it. Measure the shift directly at your pi-pulse power first.