ginc-seeds-reporting

Summary

GINC does not report a single number from one run; it repeats each experiment five times with different random starts and shows how much the results wobble, so readers can tell whether a finding is stable or just a lucky draw. This means any downstream claim that leans on GINC numbers inherits that variability, and comparisons between GINC and other methods should account for the spread rather than treating the mean as a fixed fact.