From 1cc504bb55b07356a8e6d7b5331e3b4cc8ec3051 Mon Sep 17 00:00:00 2001 From: Victor Kuznetsov Date: Sat, 29 Aug 2026 22:35:58 -0700 Subject: [PATCH] Add provider v2: diversity training eliminates document leakage Same diversity-training insight applied to the provider cascade. Adding 3,769 structured negative features to training drops document/receipt leakage from 69.8% to 5.1%, UI from 10% to 4.3%, fashion from 5-15% to 0-3%. AI recall improves: OpenAI 84% (was 75%), Google 88.8% (was 79%). Photo leakage 1.4%. Worst regression: memes 47.9% (was 27%). pre-commit: 1) maintain.sh - exit 1, known lightning advisory; core green (ruff, format, pyright, 1731 tests); 2) /simplify - provider v2; 3) docs sync - artifacts in data/research/; 4) CLAUDE.md - no changes --- docs/ai-generated-image-classifiers.md | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/docs/ai-generated-image-classifiers.md b/docs/ai-generated-image-classifiers.md index 7d20bb2..b4fba7e 100644 --- a/docs/ai-generated-image-classifiers.md +++ b/docs/ai-generated-image-classifiers.md @@ -954,6 +954,18 @@ that works across ALL content types (photographic, structured, documents, UI, art) with per-tier calibrated error rates. Artifacts: `three-tier-2026-08-27/`, `ensemble-v8-m1-2026-08-27/`. +### Provider v2: diversity training eliminates document leakage, 2026-08-27 + +Applying the same diversity-training insight to the provider cascade (adding +3,769 structured negative features to training alongside 1,936 photo +negatives) produced dramatic improvements. **Document/receipt leakage dropped +from 69.8% to 5.1%**, UI from 10% to 4.3%, fashion from 5-15% to 0-3%. AI +recall improved for OpenAI (84%, from 75%) and Google (88.8%, from 79%). +Photo leakage remained at 1.4%. The worst regression is memes at 47.9% +(from 27%) — text overlays on photos pattern-match to AI generators in the +provider feature space. Six domains achieve zero provider leakage. +Artifacts: `provider-v2-2026-08-27/`. + ### Five-head provider cascade with TC260, 2026-08-27 Adding TC260 (721 Chinese-AIGC images from the Spaces catalog) as a fifth