Section 09 · Evidence

What the benchmarks say — and what they do not.

REMORA is not a raw accuracy maximizer. Its commercial center of gravity is governed tool execution, where unsafe-execution rate is the metric that matters.

09.1Full-coverage QA

Across the committed QA corpora, REMORA does not dominate single-model or naive-ensemble baselines. This is expected: the system trades coverage for selective reliability.

CategoryN = 302N = 544 (cached)
Single model57.08.8
Naive ensemble61.412.3
REMORA (governed)54.29.1
Full-coverage QA accuracy. REMORA is not positioned as a raw accuracy maximizer.
09.3Tool-call safety

With the full policy gate engaged, unsafe-execution rate collapses by roughly two orders of magnitude relative to a single-model baseline on the v2 benchmark.

ConfigurationUnsafe executionDecision accuracyMean utility
Single-model heuristic20.0%0.20-0.25
Majority-vote heuristic10.0%0.300.00
REMORA full gate0.0%0.900.62
Tool-call safety v2. Full policy gate dominates benchmark-scoped safety.
09.4Calibration & negative results