qector-ler-methodologylisted

# QECTOR LER Methodology Source of authority: v1.0.0 reference manual, chapters 15, 19, 22. ## Definitions - **Logical error rate**: the probability that the predicted observable flips differ from the sampled observable flips on the same Stim circuit. Scoring with `correction != error` is explicitly rejected (manual 15). - **Wilson 95% score interval**: a binomial confidence interval with correct coverage at small `k` and extreme `p`, unlike the Wald approximation. Always `z = 1.959963985`. ## Wilson formula (manual 15.2) For `k` logical errors in `n` trials: centre = ( p + z^2/(2n) ) / ( 1 + z^2/n ) margin = z * sqrt( p(1-p)/n + z^2/(4 n^2) ) / ( 1 + z^2/n ) CI_95 = ( max(0, centre - margin), min(1, centre + margin) ) with `p = k/n` and `z = 1.959963985`. Worked example (manual 15.2, appendix E.2): `k=10, n=1000` -> CI is approximately `(0.00544, 0.01831)`. The `qector-research.wilson_ci` tool returns exactly this; `qector-research.wilson_table` returns a batch of intervals for a series of `k` values at fixed `n`. ## Required-artifact metadata (manual 22.3, table 22.1) Every evidence artifact must include: | Field | Required detail | | --------------- | -------------------------------------------------------------------------------- | | code_family | `repetition`, `ring`, `rotated_surface`, `toric`, `heavy_hex`, `color_code`, `bicycle`, `bivariate_bicycle`, cust