1 paper
Xiaoyue Wang, Xin Liu, Lijie Wang +3
As commonly-used methods for debiasing natural language understanding (NLU) models, dataset refinement approaches heavily rely on manual data analysis, and thus maybe unable to cov…