Yahoo
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Yahoo
57.13Micro F1
64
Yahoo (test)
326.7NLL
48
Yahoo
70.3Accuracy
42
Yahoo (test)
77.1Accuracy
36
Yahoo
72.5Micro F1 Score
33
Yahoo Arts 1
0.751Macro-AUC
32
Yahoo (test)
30Trigger Inversion Success Rate
26
Yahoo high poison rate
29Success Rate
26
Yahoo
10Tell me seriously
24
Yahoo
3.53Expected Calibration Error (ECE)
24
yahoo Business1
81.15Macro-AUC
24
yahoo Business1
0.127Ranking Loss
24
yahoo Business 1
36.82Macro-F1
24
Yahoo Arts1
28.31Macro-F1
24
Yahoo 10 cls.
0.001MSE
22
Yahoo
30Trigger Inversion Success Rate
22
Yahoo
93.5ASR
22
Yahoo StandardCL (test)
70.34Accuracy
20
YAHOO
86.9PA-F1
18
Yahoo
-325.7Prior LL
18
Yahoo (test)
88.5F1 Score
17
Yahoo mb - 1 token high poison rate (test)
10Inversion Success Rate
14
Yahoo Tell me seriously - 3 tokens high poison rate (test)
1Inversion Success Rate
14
Yahoo B (test)
87.4Accuracy
14
Yahoo (F) (test)
73.7Accuracy
14