FUMA + CKA / REEF

Description

Forensic detection of model merges and lineage via centered kernel alignment between candidate parent models and a target.

Technical signature

in: (candidate_parents[], target) → out: { lineage_graph }

Adversarial robustness

Defeated by post-merge SFT large enough to overwhelm shared representations.

Daubert eligibility

Method is peer-reviewed, has published error rates, and is reproducible from the chain-of-custody hash. Eligible for Tier-A admissibility packaging. Ultimate admissibility is the court's determination.