Metadata Predictability Is Not Evidence Dependence: An Intervention-Based Audit for Weak-Label Benchmarks

by [unverified] · free · Last verified 2026-06-21T03:07:59.154Z

We study a protocol-level test for weak-label benchmarks: whether benchmark outputs change when the provided evidence is intervened on. Metadata-only shortcut checks answer a different question, namely whether outputs are predictable from metadata priors. We therefore combine a metadata statistic...