# Teen and child classifier recheck

Requested dataset scope: **276 identities × 10 expressions = 2,760 images**, covering babies, preschoolers, both school-age groups and both teen groups, including 18-year-olds in the older-teen category. **Six previously published child/teen model-gallery examples** are checked separately; their scores and any replacement commit are in the gallery reports. The published snapshot has checked **2,760 current dataset images**; **1,135 pass the operational criteria** and **1,625 require further processing**. This is an ongoing snapshot unless the linked completion report explicitly says complete.

The exact [policy](metadata/minor-safety/policy.json), [detector provenance](metadata/minor-safety/detector-provenance.json), [per-image scores](metadata/minor-safety/scores.jsonl) and [current validation](metadata/minor-safety/current-validation.json) are checksum-specific. They identify the actual PNG bytes, not only filenames. Already-public TAR contents are matched to their remote SHA256 and byte size before scoring the identical local images.

Two official LAION image models are used: the [CLIP ViT-L/14 binary classifier](https://github.com/LAION-AI/CLIP-based-NSFW-Detector) and the [LAION-SAFETY EfficientNetV2 B2 classifier](https://github.com/LAION-AI/LAION-SAFETY). The latter's drawing/neutral classes are SFW; its hentai/porn/sexy softmax scores are summed for the NSFW score. Each model sees the original portrait and a foreground crop fitted into a neutral square. CLIP uses L2-normalized projected OpenAI ViT-L/14 image embeddings; EfficientNet uses RGB0–1 and its documented260-pixel image input.

The deliberately strict operational threshold is **0.10**: an image is queued for replacement if either model/view reaches it. Scores at least0.50 are recorded separately as positive. These thresholds are project criteria, not documented guarantees or a calibration for this illustrated domain. The two models can disagree; a flagged score does not establish that an image depicts sexual content. Model scores can produce false positives and false negatives.

New teen/child TARs are held until the exact image checksums pass. During repair the default/base Datasets configurations whitelist already-uploaded, checksum-approved original TARs; entire flagged TARs are temporarily omitted. The [stream-filter snapshot](metadata/minor-safety/public-stream-filter.json) records that reduced subset. Previously uploaded files may still be visible in the repository while their replacements are pending. The complete500-person plan remains the target, not a claim about this temporarily filtered stream. Full wildcard streaming is restored after the recheck and all replacement uploads complete. Replacements use fully covered, age-appropriate opaque clothing and are checked again. If a neutral reference changes, all ten emotions are regenerated against the new reference. When the original neutral already passes, its exact pixels and other passing expressions are retained; only flagged expressions are regenerated against that same neutral. Every released group still has ten passing expressions. Published replacement audits distinguish changed and retained PNG hashes, parameters and the corresponding TAR changes. Individual image assets remain inside TARs.