NIST TrojAI oct2024 FULL-FT (gemma-2-2b)
Matched reference — read against a recipe-matched benign population from the same corpus — the strong regime.
NIST TrojAI oct2024, gemma-2-2b FULL fine-tune substrate (delta = FT - base).
The reading
VerdictPARTIAL + currently INVERTING
At n=6 (3 clean / 3 poison) the matched-reference read is fully inverted (AUC 0.0; clean read-mean > poison read-mean). Tiny-n OR a real full-FT substrate inversion — undetermined; needs the full 16c/16p. DO NOT claim full-FT coverage.
The aggregate
The rank-separability score and the shipped operating point, side by side. The verdicts are a projection at a false-positive budget, not a raw score — coarse reads only, no detector numbers.
No AUC is reported for this cohort — see the caveat above. We don’t manufacture a rank-separability number where the design doesn’t support one.