False-acceptance frequency of external decision mechanisms

Determine how frequently an external decision mechanism incorrectly accepts a candidate as ready, including under requirement-based sensor-coordinate and polarity binding tasks.

Background

The paper emphasizes that moving release authority outside a language-model generator does not by itself establish the reliability of the resulting acceptance layer. An external gate may still produce incorrect acceptances, so its false-release frequency must be assessed independently of the model’s confidence or self-assessment. The paper reports a benchmark-specific measurement, but the broader question identified in the literature remains unresolved.

References

However, two questions are left unresolved: whether a generator's output can be checked reliably by the generator itself, and how often an incorrect acceptance is produced by the external decision mechanism.