Science Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
99.8Accuracy
916
Jun 26, 2026
98.2Accuracy
357
Jul 9, 2026
96Accuracy
354
May 12, 2026
97.4Accuracy
338
Jul 1, 2026
98.3Average Accuracy
273
Jun 1, 2026
96.3Accuracy
268
Jun 4, 2026
97.53Accuracy
240
May 29, 2026
92.8Accuracy
186
Apr 24, 2026
98Accuracy
162
May 26, 2026
96.6Accuracy
149
Jul 1, 2026
97.7Normalized Accuracy
137
Mar 17, 2026
94.3Accuracy (SciQ)
134
Jun 18, 2026
93Accuracy
111
Jun 12, 2026
86.4Accuracy
108
Jun 24, 2026
94.6Accuracy
89
Jun 24, 2026
87.6pass@1 Accuracy
85
Feb 26, 2026
91.9Accuracy
84
Jun 11, 2026
98.8ARC Accuracy
82
Jun 4, 2026
98.4Accuracy
74
Apr 7, 2026
95.6Score
70
May 14, 2026
63.51Accuracy
69
Apr 21, 2026
55.8Accuracy
69
May 27, 2026
88.6IMG Score
64
Mar 10, 2026
40.85Exact Match Accuracy
60
Apr 14, 2026
94.3Accuracy
60
Jul 3, 2026