Glossary · Term

AUROC

← all terms

Definition

Plain language

A score from zero to one for how well a classifier separates the right answers from the wrong ones.

As stated in the literature

Area Under the Receiver Operating Characteristic curve, a threshold-free measure of binary classification quality; a value of 1.0 indicates perfect separation.

Also called: AUC

Why it matters: It summarizes classifier quality in a single number that doesn't depend on a particular operating point, making models easy to compare.

For example, a spam filter with AUROC 0.95 ranks almost all spam messages above almost all real ones, regardless of where you draw the threshold.

Heard on the show

“Train a linear classifier on the model's internal activations, with oracle labels, and it reads per-answer fatality at an AUROC of 0.”
Episode 244 — The Open-Weight Defense That Feeds Attackers Confident, Falsified Answers

Mentioned in 5 episodes

  1. 244
    The Open-Weight Defense That Feeds Attackers Confident, Falsified Answers
  2. 173
    The Free Step-Level Grader Hiding in Every RL Training Run
  3. 171
    The Safety Decision a Model Makes Before It Thinks a Word
  4. 153
    Catching a Lie From the Inside, When the Words Look Completely Honest
  5. 037
    Why Hallucination Detectors Miss Stale Facts: A Geometric Story About What Models Know But Don't Say