upload paper.md

paper.md

@@ -28,8 +28,11 @@ smoke (v1.0 paper measurement) reported 0.821 EN / 0.837 RU; the
 4× expanded battery with proper class balance per (lang, genre) cell
 stabilized the numbers upward.
 
-On the 300-sample adversarial paired set, ensemble AUROC reaches
-**0.985** (in-distribution human baseline).
+On the 300-sample adversarial paired set (cross-model paraphrasing
+attack, OOD-augmented baseline), ensemble AUROC reaches **0.998**
+(re-measured 2026-04-29 with current calibration). Earlier v1.0
+paper measurement was 0.985 — the slight increase reflects the
+intervening calibration tuning between Gap-7 and current state.
 
 The contribution of this work is **field-leading reproducibility**, not
 state-of-the-art absolute AUROC. Anyone can clone the repository, run the
@@ -81,7 +84,7 @@ Our headline numbers, reproducible end-to-end on Hetzner CX43-class hardware
 
 - **English ensemble OOD AUROC: 0.864** (176-sample expanded smoke, 2026-04-29)
 - **Russian ensemble OOD AUROC: 0.846** (176-sample expanded smoke, 2026-04-29)
-- **English ensemble adversarial AUROC: 0.984** on 300-sample paraphrase-paired set
+- **English ensemble adversarial AUROC: 0.998** on 300-sample paraphrase-paired OOD-augmented set (re-measured 2026-04-29)
 - **English ensemble p50 latency: 1.2 seconds** (8-core CPU, no GPU)
 
 Earlier v1.0 paper reported 0.802/0.847 on the original 44-text smoke;