Current Readiness

CogniPrint is a working empirical evidence package supporting a follow-up manuscript. The current readiness decision remains descriptive_only.

311combined readiness rows
0/1valid external reviews recorded
1remaining readiness blocker

Empirical Snapshot

The local empirical snapshot records controlled perturbation campaigns and sample-pair comparisons. It remains corpus-bound and descriptive.

5controlled perturbation campaigns
41local campaign comparison rows
11campaign-004 comparison rows

Benchmark And Holdout Growth

The public benchmark v1.1 and independent holdout layers expand review coverage without turning the project into a general classifier or final validation endpoint.

20released benchmark baselines
120released benchmark variants
50independent holdout rows

Descriptive Validation

The current validation layer includes bootstrap intervals, effect sizes, variance summaries, random-reference comparisons, threshold sensitivity, multiple-comparison correction, and one conventional stylometry baseline. These outputs remain descriptive until external review is recorded.

20benchmark baselines in wave-005 validation
120benchmark variants in wave-005 validation
6multiple-comparison correction tests

Interpretation Guardrail

The evidence visibility layer is descriptive and review-oriented. It does not upgrade the scientific meaning of the current public evidence package into a settled empirical endpoint, a forensic workflow, or a stronger attribution claim.