Found 1 document(s)
Page 1 of 1
-
Stance classifiers make errors that can be costly to identify under limited human-review budgets. Prioritizing potentially incorrect predictions remains challenging, particularly across heterogeneous base models. We intr...