Skip to content

Phase 2 no-QPU cross-checks

Date: 2026-05-05

These checks spend no additional IBM QPU time. They use only promoted Phase 2 raw counts and derived summaries.

Readout mitigation boundary

Script: scripts/analyse_phase2_readout_mitigation.py

Output: data/phase2_readout_mitigation/phase2_readout_mitigation_summary_2026-05-05.json

The existing raw counts include selected readout-only calibration states, not a complete computational-basis calibration set. Therefore the repository cannot honestly claim a full 2^n x 2^n confusion-matrix inversion without new QPU calibration circuits.

Implemented correction:

  • exact-state parity-confusion inversion,
  • only for initial states with matching readout-only calibration rows,
  • no full-count redistribution across all basis states.

Main result after correction:

  • Phase 2 A+G n=4 replication: Fisher chi2=73.894393, p=4.160825e-08, significant depths 5/10.
  • Popcount original contrast: Fisher p=9.267543e-18.
  • Popcount within-even swap: Fisher p=4.466548e-20.
  • Popcount within-odd swap: Fisher p=3.590153e-15.
  • Popcount excitation-inversion arm: Fisher p=5.929924e-18.

Conclusion: readout-only parity correction weakens the Phase 2 A+G combined statistic but does not remove the promoted sign/statistical conclusion. The popcount-control downgrading of the causal claim also survives correction.

Offline GUESS-readiness calibration

Script: scripts/analyse_phase2_guess_calibration.py

Output: data/phase2_guess_calibration/phase2_guess_calibration_summary_2026-05-05.json

This is not GUESS zero-noise extrapolation because the promoted Phase 2 datasets do not include explicit folded-noise scale factors. The analysis fits log parity-survival versus circuit depth as a cumulative-noise proxy.

Fit quality:

  • Phase 2 A+G even: alpha=0.034190, R2=0.9989.
  • Phase 2 A+G odd: alpha=0.034662, R2=0.9997.
  • Phase 2 A+G sector mean: alpha=0.034427, R2=0.9995.
  • Popcount state fits: R2=0.9582 to 0.9921.

Conclusion: parity leakage is smooth enough versus depth to justify preparing a future GUESS-style folded-noise protocol. It is not yet a demonstrated error mitigation result.

Hardware follow-up preparation

Prepared protocol: docs/campaigns/next_validation_protocols_2026-05-05.md

Recommended order:

  1. Keep the current paper as a no-new-QPU preprint after incorporating the readout and GUESS-readiness boundaries.
  2. Use the state/layout-randomization protocol only if reviewers or collaborators ask for mechanism separation beyond the popcount control.
  3. Use the minimal multi-device protocol only if backend-transfer evidence is required before journal submission.