Answers are grounded in this paper and cite their sources.
The distance-3 guarantee of the Shor [[9,1,3]] code promises correction of any single-qubit error, but the code's degeneracy makes a much larger, structured set of multi-error patterns correctable. We enumerate that set exactly. Restricting bit-flip (X) support to at most one qubit per 3-qubit block and phase (Z) support to a single block defines a favourable family of 1,408 Pauli patterns; we prove the family is uniquely decodable by a complete syndrome-collision check (any two members sharing a syndrome are stabilizer-equivalent), and we prove the family is maximal in the phase direction: a Z pattern spread across two blocks shares its syndrome with a single-block pattern from which it differs by a logical operator. An exhaustive census of all 324 weight-2 and 2,268 weight-3 patterns finds the resulting decoder corrects 55.6% and 35.4% of all patterns respectively, against 13.9% and 8.1% for the standard single-error lookup decoder, with in-family patterns corrected at 100%. In a pre-registered paired Monte Carlo benchmark (depolarizing noise, noisy syndrome readout, 20,000 trials per point), the favourable decoder reduces total logical failure by 25–27% relative across physical error rates p = 0.005–0.05, while exhibiting a higher silent logical failure rate — it never erases where the single-error decoder gives up. Under strongly Z-biased noise the total-failure advantage erodes to a statistical tie, which we map explicitly. The Pauli-frame results are validated against full statevector simulation (0.200% vs 0.210% failure at p = 0.01). All constructions are verified by an open, self-testing reference implementation.
No takes yet. Share an insight, caveat, or question.
Derrick Irons (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: