Regulator feedback
RegLab predicts a verdict; the regulator/ARM returns the actual one. Reconciling the two closes the loop. A report RegLab expected to pass that the regulator rejected is a rule RegLab is missing (surprise-reject → a proposed coverage gap); one RegLab expected to fail that the regulator accepted is a false positive to tune. Every divergence is a proposed, human-reviewed finding — never an auto-change to the ruleset (invariant RG-1).
a submitted population + the ARM response (with planted surprise-rejects/accepts)
Or bring your own data — paste your reports + the ARM response file
reports (what you submitted)
ARM/NCA response file · CSV or auth.031 XML