ai
Recall
Recall (Evaluation Metric)
Definition
Recall is the fraction of all relevant items that are successfully retrieved or predicted, calculated as true positives divided by all actual positives. High recall means the model misses few relevant items, though potentially at the cost of including irrelevant ones.
In RAG systems, retrieval recall is critical — missing a relevant document means the model cannot use its information.
Can this ship? Are we exposed? Can you prove it?
Every CISO is asking these three questions. We answer them with facts, not guesses, from repository to runtime.