SAI LabsBook a demo

Safety IndexEvaluationsVerbalised evaluation awareness

Verbalised evaluation awareness

Evaluation integrity · Evaluation awarenessEvaluation-integrity profileBuilt, not yet run

How often the model says out loud that it thinks it is being tested, judged over the answers it already gave on other benchmarks in this index.

Metric
% answers verbalising test awareness. Capability, not scored.
Dataset
–
Grader
–
Status
Built, not yet run
Replicability
public

Results

Built and tested; the first full runs are pending.