Preference Elicitor

Software componentModel AdaptationModelsVariation point (abstract)arc:PreferenceElicitor

An abstract component that derives utility or reward function parameters representing a principal's preferences from observed behaviour instead of explicit engineering.

Responsibility. Infers utility-function parameters from observed choices or demonstrations.

Also known as: Utility function learner, Preference learning

Direct neighbourhood (hover for relationship types)

Variants

Variant	When to choose
Inverse Reward Learner	Choose when expert demonstrations (e.g., thousands of hours of human driving) are available and explicit preference specification is impractical.
Revealed Preference Learner	Choose when users' selections among presented options and feedback are continuously observable, enabling personalization without explicit preference configuration.

Relationships

writes dependency

Utility Function Specification Ch5.10

Design guidance

SHOULD resolve the ambiguity of multiple utility functions explaining the same behaviour through simplicity assumptions, priors over plausible utilities, or active queries.

Classification

Risks mitigated: Misalignment between engineered utility functions and actual preferences

Sources

Ch5.10: T. Nguyen, "Utility-Based Decision Making Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.10. ISBN: 9798244538229.