Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
71.1Mean@16
3
May 28, 2026
81.5Mean@16
3
May 28, 2026
70.5Accuracy
3
May 28, 2026
55Accuracy
3
May 28, 2026
68Accuracy
3
May 28, 2026
12Response Content Detail Score
3
May 27, 2026
60.1ROUGE-L Score
3
May 26, 2026
11.63F1 Score
3
May 22, 2026
25.4OBQA
3
May 22, 2026
7.8Risk
3
May 21, 2026
0Pathwise Violations (PathV)
3
May 21, 2026
0Pathwise Violations (PathV)
3
May 21, 2026
0Pathwise Violations
3
May 21, 2026
6.9Risk
3
May 21, 2026
46.6Exact Match (EM)
3
May 19, 2026
0.065Capability Delta (Δ)
3
May 15, 2026
92.8Precision
3
May 13, 2026
55.7Accuracy (0-4k Context)
3
May 12, 2026
67.22wikimqa
3
May 12, 2026
87.63Performance
3
May 12, 2026
83.1Accuracy
3
May 12, 2026
83.03Precision
3
May 11, 2026
0.091F1 Score
3
May 11, 2026
27.4F1
3
May 11, 2026
0.858AUROC
3
May 8, 2026