Reading Comprehension
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
87.39Accuracy
8
Mar 16, 2026
94.22Accuracy
8
Mar 16, 2026
0.673Accuracy
8
Mar 6, 2026
52.9Accuracy
8
Feb 26, 2026
93.01EM
8
Feb 26, 2026
54.11BLEU-1
8
Feb 26, 2026
36.55BLEU-1
8
Feb 26, 2026
33,350.02Average Context Length (tokens)
8
Feb 26, 2026
81.7Accuracy
7
May 12, 2026
35.33Accuracy (Normalized)
7
Apr 23, 2026
75.1Accuracy
7
Feb 26, 2026
2.3Speedup
7
Feb 26, 2026
59.7EM
7
Feb 26, 2026
90.1Maximum Accuracy
7
Feb 26, 2026
31Exact Match
6
Jul 7, 2026
64.73Client 1 Accuracy
6
May 28, 2026
59.01Accuracy (DROP Single Client)
6
May 28, 2026
79.87Exact Match (EM)
6
May 18, 2026
81.17F1 Score
6
May 12, 2026
82.86Score
6
Apr 21, 2026
74.1Accuracy
6
Mar 20, 2026
90.2Accuracy
6
Feb 26, 2026
76.4P@1
6
Feb 26, 2026
64.89Accuracy
6
Feb 26, 2026
71.78Accuracy
6
Feb 26, 2026