All Publications
METR
OrganizationHigh(4)
Model Evaluation and Threat Research
Credibility Rating
4/5
High(4)High quality. Established institution or organization with editorial oversight and accountability.
20
Resources
52
Citing pages
1
Tracked domains
Tracked Domains
metr.org
Resources (20)
20 resources
Citing Pages (52)
AI Accident Risk CruxesAjeya CotraAlignment Research CenterCapability ElicitationAI Capability Threshold ModelAutonomous CodingAI Governance Coordination TechnologiesCorporate AI Safety ResponsesDangerous Capability EvaluationsAI Safety Defense in Depth ModelEmergent CapabilitiesEpistemic Virtue EvalsEpoch AIEval Saturation & The Evals GapAI EvaluationsEvals-Based Deployment GatesAI EvaluationHeavy Scaffolding / Agentic SystemsInternational AI Safety Summit SeriesAI Safety Intervention Effectiveness MatrixAI Safety Intervention PortfolioIntervention Timing WindowsAI Lab Safety CultureLarge Language ModelsLong-Horizon Autonomous TasksMesa-Optimization Risk AnalysisMETRThird-Party Model AuditingPersuasion and Social ManipulationAI ProliferationAI Development Racing DynamicsRed TeamingResponsible Scaling Policies (RSPs)Reward HackingAI Risk Activation Timeline ModelAI Risk Cascade Pathways ModelResponsible Scaling PoliciesAI Safety CasesAI Safety Culture Equilibrium ModelSandboxing / ContainmentScalable Eval ApproachesScheming & Deception DetectionSelf-Improvement and Recursive EnhancementSeoul Declaration on AI SafetySurvival and Flourishing FundSituational AwarenessSycophancyTechnical AI Safety ResearchTool-Use RestrictionsAI Risk Warning Signs ModelAI Whistleblower ProtectionsWhy Alignment Might Be Easy
Publication ID:
metr