ProCreations/repro-a-coin-flip-for-safety-llm-judges-fail-to-reliably-measure-adversarial-robustness

A Coin Flip for Safety — released-data reproduction

Six claim-matched executions over the authors' full released audit, ReliableBench, and JudgeStressTest. Run python reproduce.py and python validate_evidence.py for deterministic cold replay and fail-closed validation.