Agent Maturity Compass API - v1.2.0
    Preparing search index...

    Interface InterpretabilityEvent

    Interpretability Scoring — Can you understand why the agent decided?

    Inspired by Google DeepMind's Gemma Scope: interpretability tools that trace decisions through model internals. AMC can't access model weights, but CAN score the observable interpretability surface:

    • Decision explanations: Does the agent explain its reasoning?
    • Chain-of-thought faithfulness: Do explanations match actions?
    • Confidence calibration: Are confidence signals accurate?
    • Attribution quality: Can decisions be traced to inputs?
    • Refusal transparency: When the agent refuses, does it explain why?
    interface InterpretabilityEvent {
        actionId: string;
        actionMatchesExplanation: boolean;
        actualOutcome?: "partial" | "success" | "failure";
        attributionSources: number;
        explanationLength: number;
        hasAttribution: boolean;
        hasExplanation: boolean;
        isRefusal: boolean;
        refusalHasReason: boolean;
        statedConfidence?: number;
        timestamp: number;
    }
    Index
    actionId: string
    actionMatchesExplanation: boolean
    actualOutcome?: "partial" | "success" | "failure"
    attributionSources: number
    explanationLength: number
    hasAttribution: boolean
    hasExplanation: boolean
    isRefusal: boolean
    refusalHasReason: boolean
    statedConfidence?: number
    timestamp: number