Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
90.17SQuAD v1.1 F1
21
Mar 10, 2026
98.78Accuracy (All)
21
Feb 26, 2026
54.3Average F1
21
Feb 26, 2026
83.3F1
21
Feb 26, 2026
63.8EM
21
Feb 26, 2026
85.1Accuracy
20
Jun 30, 2026
0.27Normalized Generalization Error Approximation
20
Jun 15, 2026
1.51Generalization Error (G(W))
20
Jun 15, 2026
4.416Score
20
Jun 11, 2026
67.1Concept Accuracy
20
Jun 4, 2026
84.4Concept Accuracy
20
Jun 4, 2026
78.3TruthfulQA Mean Score
20
Jun 1, 2026
0.23Loss
20
Jun 1, 2026
0.22Loss
20
Jun 1, 2026
63.2Accuracy
20
May 28, 2026
86.9E2E Wall Time (s)
20
May 25, 2026
74Answer-letter Accuracy
20
May 22, 2026
41.84EM
20
May 19, 2026
86.23Accuracy
20
May 15, 2026
86Accuracy
20
May 13, 2026
67.3Accuracy
20
May 13, 2026
64EM
20
May 8, 2026
53.6OBQA Accuracy
20
Apr 23, 2026
67.26Normalized Accuracy
20
May 12, 2026
82.9Avg Score (All Datasets)
20
Apr 10, 2026