1 paper
Jingzhou Jiang, Yixuan Tang, Yi Yang +1
When task-specific labels are not available, it becomes difficult to select an embedding model for a specific target corpus. Existing labelless measures based on kernel estimators…