Model Metric

Noun · AI & Machine Learning

Definitions

  1. A metric used to measure some aspect of model behavior, such as accuracy, latency, cost, calibration, or acceptance rate. Model metrics are most useful when tied clearly to real product outcomes.

    In plain English: A measurement used to judge model performance or behavior.

    Example: "They added a new model metric for citation validity after discovering that high answer ratings still hid factual errors."

Related Terms