Here you will find three handy guides to help you understand the details of S.E.B./C.I.B. Each one increases in technical complexity from the previous one.
What S.E.B. and C.I.B. are, why an outside lab is needed when AI companies grade their own homework, and what we explicitly do not claim.
For anyone, including someone who has never heard of us. · Read online · PDF
The cell as the unit of measurement, blind judging, why a model’s claim is asked rather than inferred from its prose, why most of each battery stays secret — and what an informed sceptic should still hold against us.
Assumes Guide 1. Still no equations. · Read online · PDF
Every formula in S.E.B. and C.I.B., worked through with examples — the trimmed mean, the threat score, the four reliability statistics including the ones where we score badly, the Reliance Gap, Wilson intervals, the cluster bootstrap and d′.
For a reader who wants to check the arithmetic rather than take it on trust. · Read online · PDF
None of these carries a result figure that could go stale in a drawer — the live numbers are on the dashboards and change as we re-test. What they carry is the method, which changes far more slowly and is the part worth arguing with.