Science Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
86.9Accuracy
10
Feb 26, 2026
70.1Accuracy
10
Feb 26, 2026
76.37Accuracy@8
9
Jul 3, 2026
72.44Score
9
Jul 3, 2026
50.96Accuracy (ARC zero-shot)
9
Jun 1, 2026
79.3Mean@16
9
May 28, 2026
65.4Mean@16
9
May 28, 2026
81.6Mean@16
9
May 28, 2026
80.6Mean@16
9
May 28, 2026
63.1Accuracy
9
May 15, 2026
69.5EM
9
Mar 27, 2026
180.31OTPS
9
Feb 26, 2026
61.5Score
9
Feb 26, 2026
50.54Accuracy
9
Feb 26, 2026
70.14Accuracy
9
Feb 26, 2026
4.5Steps
8
Jun 16, 2026
85Accuracy
8
Jun 15, 2026
57.32Pass@1
8
Jun 9, 2026
3.41Speedup Ratio
8
Jun 2, 2026
58.1Accuracy
8
May 13, 2026
83.9Accuracy
8
May 13, 2026
35.59Score
8
Mar 31, 2026
67.4Accuracy (Avg@8)
8
Jun 5, 2026
90.2Accuracy
8
Feb 26, 2026
34.75Accuracy
7
Jun 26, 2026