Likelihood Splits
method
ID: likelihoodsplits
Entity Details
Basic Information
Entity ID
likelihoodsplits
Entity Type
Method
English Label
Likelihood Splits
Russian Label
Разделы вероятности
Kazakh Label
Мүмкіндік бөліністері
Frequency
1
Description
English Description
A method to create challenging benchmarks by splitting datasets based on the likelihood assigned by a pre-trained language model.
Russian Description
-
Kazakh Description
-