SelfAware
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
SelfAware Gemini outputs (test)
52.8AUROC
15
SelfAware GPT outputs (test)
0.528AUROC
15
SelfAware Llama outputs (test)
58.7AUROC
15
SelfAware
0.372Score
10
SelfAware
51.2Accuracy
10
SelfAware
0.587AUROC
9
SELFAWARE
91.4U-Ref
7
SelfAware (out-of-domain)
9.9nAUPC
4
SelfAware
27Accuracy
1