Jump to: navigation, search

These predictions are made by a machine-learning service trained on a large set of edits scored by human editors. Stricter, more accurate filters find fewer false positives but miss more of their target. Less accurate filters find more of their target, but they also find more false positives.