behavioralConsistencyScore
behavioralConsistencyScore: number
OptionalbiasResistance
biasResistance?: number
OptionaldeceptiveAlignmentResistance
deceptiveAlignmentResistance?: number
OptionalfeedbackSourceValidation
feedbackSourceValidation?: number
OptionalgoalIntegrity
goalIntegrity?: number
OptionalgoalMisgeneralizationResistance
goalMisgeneralizationResistance?: number
instructionComplianceScore
instructionComplianceScore: number
OptionalpreviousOverall
previousOverall?: number
OptionalrewardHackingResistance
rewardHackingResistance?: number
OptionalsabotageResistance
sabotageResistance?: number
safetyScore
safetyScore: number
OptionalselfPreservationCompliance
selfPreservationCompliance?: number
OptionalsycophancyResistance
sycophancyResistance?: number
truthfulnessScore
truthfulnessScore: number