Multiple-choice Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
93.84Top-1 Accuracy
9
May 15, 2026
79.85Accuracy
9
May 25, 2026
0.04CMD
9
Feb 26, 2026
63.84Accuracy
9
Feb 26, 2026
77Accuracy
9
Feb 26, 2026
59.21Average Accuracy
9
May 25, 2026
78.6Micro-F1
8
Jun 30, 2026
68.9Exact Match
8
Jun 4, 2026
79.3Accuracy
8
Jun 4, 2026
32.5Group Accuracy
8
Jun 4, 2026
78Accuracy
8
Jun 4, 2026
78.7Normalized Score
8
Jun 4, 2026
79.6GPT Score
8
Jun 4, 2026
85.4GPT Score
8
Jun 4, 2026
91Accuracy (Economics)
8
Jun 2, 2026
65Accuracy
8
May 22, 2026
61.11Accuracy
8
May 15, 2026
45.4Accuracy
8
May 15, 2026
61.13Accuracy
8
May 15, 2026
55.8Accuracy
8
May 15, 2026
0.884Accuracy
8
May 15, 2026
63.8MMLU-Pro Score
8
May 12, 2026
0.88Accuracy
8
May 8, 2026
0.983Calibration Threshold (q-hat)
8
Apr 10, 2026
99.2Calibration Threshold (q-hat)
8
Apr 10, 2026