SST
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
SST
1ROUGE-1
32
SST2
95.27Macro-F1
30
SST5
45.92Macro-F1
30
SST-2
96ACC
30
SST (val)
68.76Top-1 Acc
30
SST binary
91.7Accuracy
29
SST-2
79.49False Acceptance Rate (FA)
28
SST2
0.563AUC π-Soft (NS)
27
SST2
95.4FA Score
27
SST-2
94Accuracy
26
SST-5
52.6Accuracy
25
SST-2
91.1Accuracy
25
SST
0.507MSE (sigma=0.5)
24
SST-2
52.5Positive Rate
24
SST-2
-0.675Delta AF
24
SST-2
100ROUGE-1
24
SST
88.19Accuracy
24
SST5
76.65Spearman's rho (x100)
23
SST-2
54.6Attack Success Rate (ASR)
22
SST
24.2PPL Change (%)
20
SST-2 (OOD)
82.4Accuracy
18
SST2
89.9Accuracy
17
SST2
2.1Attack Success Rate (ASR)
17
SST2
88.52FPR
17
SST2 (test)
0.928AUROC
17