Could postselection on a different quantum state yield a different repulsive magnitude, and would that consistency strengthen the artifact-rejection case?

Your question probes a subtle and important point about the experimental signature of the repulsive gravity effect.

Yes, in principle postselecting on *different* quantum states of the source mass should yield different repulsive magnitudes [AG-2026.02-534, AG-2026.02-268]. Both papers emphasize that the repulsion arises from the *quantum interference* between distinct spatial superposition states of the source, conditioned on a specific measurement outcome. The magnitude of this interference—and hence the strength of the effective repulsive force—depends on the details of the initial superposition and the choice of final postselected state [AG-2026.02-268]. Different postselection choices would couple to different components of the superposition, producing different weak values and thus different repulsive signals.

Why this matters for artifact rejection: if you observe *consistent, state-dependent repulsion* across multiple postselection choices, that would be much harder to dismiss as a classical measurement artifact or systematic error. A genuine artifact would likely show up as noise or non-reproducible signals independent of the quantum state you postselect on. But a true quantum effect—mediated by negative weak values and spacetime superposition—should show *structured variation* that tracks the quantum state you've prepared. Demonstrating this structure would be a powerful way to rule out simple calibration errors or detector biases, since those wouldn't "know" which quantum state you'd chosen [AG-2026.02-268].

The provided preprints don't directly address whether experiments have yet mapped out this state-dependent repulsion landscape, so I can't tell you whether that consistency check has been performed.