Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
58.4pass@1
18
May 20, 2026
82.8Pass@1
18
May 20, 2026
62.6pass@1
18
May 20, 2026
70.8Accuracy
18
May 13, 2026
87Accuracy
18
May 13, 2026
57.2NQ Accuracy
18
May 8, 2026
84Precision
18
Apr 28, 2026
65EM Score
18
Apr 28, 2026
4.4CR Score
18
Apr 28, 2026
61.2Accuracy
18
Apr 28, 2026
84Accuracy
18
Apr 21, 2026
53.8HotpotQA Full-Slice Distance (d)
18
Apr 21, 2026
85.42Restate-hard
18
Apr 21, 2026
79.95Accuracy (NQ val)
18
Apr 21, 2026
88.2Accuracy
18
Apr 21, 2026
79.9Accuracy
18
Apr 21, 2026
79Accuracy
18
Apr 21, 2026
90.21Accuracy
18
Apr 21, 2026
62.2Accuracy (Distractor hard)
18
Apr 21, 2026
94.4Accuracy
18
Apr 21, 2026
91.5Exact Match (EM)
18
Jun 19, 2026
0.685vbal
18
Apr 7, 2026
94.1Pass@1
18
May 14, 2026
81.9Accuracy
18
Mar 24, 2026
82.1Accuracy
18
Mar 24, 2026