Factual recall
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
77Accuracy
16
Feb 26, 2026
58MMLU Factual Recall Accuracy
14
Feb 26, 2026
81.6Top-1 Accuracy
12
Jun 23, 2026
33.9Hit@1
6
May 20, 2026
0.756AUROC
3
Jul 9, 2026
0.734Tau
3
Jun 11, 2026