schema · evergreen

Saying no: three bands

kNN never says nothing is close; cosine overlaps in the grey zone, so a floor, a note, and a judge.

Nearest-neighbour search has no concept of "nothing is close". Ask a codebase whether you have been to a five-star resort and it returns eight chunks (best score 0.366). Hallucination starts right there.

CRAG-style bands on the top dense score
below 0.45dropped, counted
0.45 – 0.55returned with a weak-match note
0.55 and upnormal

Why not one threshold

Cosine does not separate the grey zone on this model and corpus: real answers have a top-1 dense median of 0.636 and a minimum of 0.526; irrelevant questions have a median of 0.524 and a maximum of 0.587. They overlap. A single cut either drops real answers or lets junk through.

Gateabstain on 13 negativesfalse alarm on 42 positivesextra latency
note below 0.550.8330.0480
floor 0.45 + note 0.55 ✓0.8460.0480
reranker score below 0.051.0000.286+550 ms

The floor removed nothing from the golden set and turned the resort question into an empty result. The two negatives that still slip through are questions with genuinely similar code in the repo — there the call belongs to the reader.

#abstention #crag #calibration

See this note on the whiteboard →