Executive Proof
Anthrocentrix Thesis Evaluation
True test: "Does decision-time behaviour beat actor history?" — M3 vs M2 on a held-out chronological split.
Anthrocentrix Kill Test
thesis under evaluation
If actor-conditioned prediction does not materially outperform a strong case-only baseline, the Anthrocentrix engine thesis fails.
Seed Anthrocentrix Verdicts (admin)
one-time idempotent load of six committed ExecutionFit verdictsNo Kill Test Result
awaiting a real experimental runThe Results dashboard reflects the most recent kill test on an imported dataset. Run one from Experiments. The dashboard reports both supportive and falsifying outcomes; the engine does not assume the hypothesis will hold.