PairSecurity
The device-independence position and what a pair link refuses
12/12 passed in 2ms
| Test | What it does | Result |
|---|---|---|
test_absent | No device-independent key rate is exposed anywhere, a coincidence-counted click layer's post-selection on detection being the fair-sampling assumption. | ✅ pass |
test_attack | Both plans carry an attack row after security: collective on the violation branch, unstated on the basis-symmetry one. | ✅ pass |
test_background | The receiver background the pair layer derives is the two-gate composition q.Link performs. | ✅ pass |
test_diagnostic | The CHSH value a run reports is labelled a diagnostic, and the plan's device_independent row is False. | ✅ pass |
test_finite | q.FiniteSize on a pair link is refused, the refusal naming coincidences. | ✅ pass |
test_forward | predict() and run() report one modelled violation and one disturbance, and every rate name on the Forecast raises. | ✅ pass |
test_launder | A Forecast handed to q.ViolationBound(source="measured") is refused as circular, while a bare magnitude still prices. | ✅ pass |
test_network | q.PairLink offers run() and explain(), and q.Network refuses it by name. | ✅ pass |
test_pumping | A continuous-wave pump is refused, the pulsed gain carrying no accidental-coincidence term. | ✅ pass |
test_slots | Out-of-domain arguments raise rather than returning a wrong number. | ✅ pass |
test_timing | Detector dead time and jitter are refused rather than approximated. | ✅ pass |
test_tsirelson | A CHSH value past Tsirelson's bound is refused, the refusal naming the detection efficiency a real claim needs. | ✅ pass |