CaseHOLD
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
CaseHOLD (test)
89.22Test Accuracy
22
CaseHold
83.13Accuracy (CaseHold)
16
CaseHOLD (test)
78.5Mean macro F1
12
CaseHOLD
100AR (%)
9
CaseHOLD
41.5Agreement (%)
8
CaseHold
96Cumulative Score (CS)
8
CaseHOLD (eval)
7.8Risk
3
CaseHOLD (CHOLD)
74.1Accuracy/F1 Score
2