{"result": "verified", "reason": "Public-suite numerical error measured; larger-size run passed sanity gates. Probe error is reported separately; generalization and source honesty are not certified.", "score": 8.881784197001252e-15}