Research

MEI focuses on world models only—neutral, open, reproducible evaluation so users get safer, auditable deployment evidence: replication audits, deployment-risk judgment, and longitudinal tracking.