Truthfulness Evaluation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
75Accuracy
108
Apr 23, 2026
84.7T·I Score
59
Apr 23, 2026
54.7BLEU-Acc
35
Jun 2, 2026
16.9Reliability Score
33
Apr 9, 2026
54.95MC1
30
May 29, 2026
66.63Avg@k
27
May 13, 2026
83.6Health Score
22
Feb 26, 2026
65.75Accuracy
20
May 29, 2026
82.6TruthfulQA Score
12
Apr 6, 2026
1.62NLL
10
Jun 11, 2026
12TruthfulQA Delta
10
May 20, 2026
58.76Normalized Accuracy
10
Mar 18, 2026
0.33Mean Improvement
9
May 18, 2026
54.9TQA Score
9
May 12, 2026
68.69Average Score (@8)
8
Apr 14, 2026
7.9Exclusive Catch Rate (@10%)
3
Apr 29, 2026
100Accuracy under Attack
2
Apr 28, 2026