TruthfulQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
TruthfulQA AR
87.8F1-Score
18
TruthfulQA EN
88.34F1-Score
18
TruthfulQA
57.89Factuality
18
TruthfulQA
68AUROC
18
TruthfulQA
68.2Accuracy
18
TruthfulQA
62.6Truthfulness Score
18
TruthfulQA
0.685vbal
18
TruthfulQA MCQ
39.33MCQ Accuracy
18
TruthfulQA Generalization
14.77ROUGE-L
18
TruthfulQA f_con o=5
100Acc (Exact)
18
TruthfulQA
83TruthRate
17
TruthfulQA
41.98Truthfulness Score
16
TruthfulQA Open-ended
99.6True Score
16
TruthfulQA latest (test)
84.57Accuracy
16
TruthfulQA
0.023ECE
16
TruthfulQA MC2
47.12Truthfulness Score (MC2)
15
TruthfulQA MC
0.073Mean Per-Step Regret
15
TruthfulQA
0.138Mean Per-Step Regret
15
TruthfulQA
74.24Neutral Accuracy
15
TruthfulQA (full dataset)
86.8Accuracy
14
TruthfulQA
79.4Accuracy
14
TruthfulQA 4 options (test)
80.1Accuracy (ACC)
14
TruthfulQA
35.3PRR (AlignScore)
14
TruthfulQA
0.871AUROC
14
TruthfulQA
25Sycophancy Rate
14