Skip to content

Learning — what a grade actually does

A grade is not a label that sits in a database; it is an instruction the system acts on immediately. Within seconds of a grade landing in ReviewHub, the following happens:

Grade Immediate effects
outlier birth certificate (Lab clone-priority seed) + vocab burst (clone intent) + LLM analysis → notes
catalyst birth certificate (evolve seed) + vocab burst (bridge intent) + notes
on-brand birth certificate (validated-bounds seed) + notes
off-brand / ugly kill phrases extracted → entity notes → injected into future negative prompts + sanitizer
delete excluded from learning corpora

Where learnings live

File-backed memfs (notes/): per-designer, per-colorway, per-garment, and universal.md. Each note carries type (recipe = guidance the LLM reads, kill_phrase = hard ban) and provenance (Shawn vs self-learned). Notes are git-versioned on the volume for audit history.

Where learnings act

  1. Compose system prompt — scoped notes for the manifest's exact entities
  2. Negative prompts — kill-phrase triggers as short noun tokens (filtered so they never negate the look's own entities)
  3. Render-hint variants — banned fabric/detail variants skipped
  4. Lab — seeds, craft blocks, per-seed scoped notes, fuse-guards
  5. Discovery — proven directions + anti-repetition fence
  6. Sleeptime synthesis — periodic clustering of grade evidence into universal craft rules (n≥2 promotion threshold)

Garment-level grades

Clicking a segmented garment in ReviewHub grades THAT garment on THAT model: the learning lands scoped to the garment (J26.md style) and the birth certificate carries the model slot — the Lab knows which outfit won, not just which image.