a measure of how well the model can recognize and localize concepts

    All notes