Agentic AI Reference Architecture

Cognition · Software component

Local Surrogate Explainer

Software componentCognitionCognition & Memoryarc:LocalSurrogateExplainer

A decision explainer that perturbs a single input and fits a local approximation of the model to estimate each feature's contribution to that prediction.

Responsibility. Produces per-prediction feature contributions for black-box models.

Also known as: LIME, SHAP, Local explanation

Variant of Decision Explainer abstract

When to choose. Choose when the decision model is opaque and a model-agnostic local approximation of top feature contributions is needed.

invokesis target of alternativeTois target of alternativeTospecializesis target of alternativeToInference Server: invokesInference ServerCounterfactual Explainer: is target of alternativeToCounterfactual ExplainerDecision Factor Explainer: is target of alternativeToDecision Factor ExplainerDecision Explainer: specializesDecision ExplainerExample-Based Explainer: is target of alternativeToExample-Based Explainer
Direct neighbourhood (hover for relationship types)

Relationships

invokes dependency

alternative to variability

Classification

Patterns
Local model-agnostic explanation
Technologies
LIMESHAP
Quality attributes
Explainability (NIST AI RMF: explainable and interpretable)

Sources

  1. Ref10.02: "Explainability and Interpretability in Agent Systems," unpublished reference note (02-Explainability-Interpretability.md), Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam supplementary materials, 2026. unpublished note