We report phase-resolved measurements revealingamplitude–phase coupling (APC) in a strongly drivensuperconducting qubit. Using 1000 shots per point acrossfive drive amplitudes on a Rigetti Ankaa-3 transmonqubit (Qubit 58; T1 = 36.7 μs, T2 = 22.2 μs), we comparea rectangular reference pulse and a self-intersectingFigure-8 trajectory under matched root-mean-squaredrive amplitude. From harmonic fits to the phasedependentsurvival probability, we extract the optimalcontrol phase ϕ∗(A) and show that it shifts by more than3 rad across the explored amplitude range for both pulsefamilies. To quantify phase sensitivity, we define thephase contrast C(A) = maxϕ ˆ F −minϕ ˆ F and the contrastratio R(A) = CF8/CRect. At low amplitude (A ≤ 0.6),the Figure-8 trajectory reduces phase contrast relativeto the rectangular reference (R = 0.48–0.57), whereas athigh amplitude (A ≥ 0.8) it increases it (R = 1.22–1.45),revealing a crossover in trajectory-dependent phase sensitivity.An independent replication acquired in a separatecalibration session at 100 shots per point preserves thesame low- and high-amplitude sign structure. Additionalpulse-train continuity tests show that internal timinggaps reduce phase contrast without measurably shiftingthe optimal phase, and that a fixed 4 ns gap showsno statistically significant dependence on its placementwithin the train. These observations identify APC as anexperimentally accessible feature of the measured controllandscape and show that fixed-phase calibration canbecome systematically biased across drive amplitudes inthe strongly driven regime.

Submitted to Quantum Science and Technology (IOP Publishing) in 2026. A provisional patent application covering aspects of this work was filed in November 2025.