Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
83.32CACC
25
Apr 10, 2026
83.96CACC
25
Apr 10, 2026
72CACC (%)
25
Apr 10, 2026
91.9Accuracy (Sudoku Reasoning)
25
Jun 2, 2026
1.26E0 Score
25
Feb 26, 2026
66ACC1
25
Feb 26, 2026
83.3Exact Match (EM)
24
May 12, 2026
79.3EM
24
May 12, 2026
58.46Pass@1
24
May 12, 2026
73.44Pass@1
24
May 12, 2026
67.66Pass@1
24
May 12, 2026
0.75Pass@1
24
May 12, 2026
7,262Length
24
May 11, 2026
2,993Output Length
24
May 11, 2026
418Solution Length
24
May 11, 2026
39.3Accuracy (superGPQA)
24
Apr 23, 2026
40.6First-Token Accuracy
24
Apr 6, 2026
85.75Accuracy on Date
24
Mar 23, 2026
0.92Accuracy
24
Feb 26, 2026
57.3Accuracy
24
Feb 26, 2026
100Accuracy
24
Feb 26, 2026
85Accuracy
23
Mar 24, 2026
87Score
23
Mar 12, 2026
15.7pass@1
23
May 28, 2026
49.49GPQA-D Score
22
May 13, 2026