Leaderboarddevs-team ledger · judge + FC scores · epistemics counts
Coder
Judge
FC
Bugs
Halluc
Unique
Trunk
Out-tok
Mins
Runs
ScoringJudge = the run judge's adjudicated 0–10 (ground-truth-verified). FC = the independent fact-check agent's score. Bugs / hallucinations / unique strengths are epistemics counts, review-council style — a lie and an honest miss never share a number.