Multi-step Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
92.72Accuracy
35
Mar 24, 2026
66.99Accuracy
20
Jul 8, 2026
80.09Task Performance
14
May 19, 2026
94Accuracy
11
Jun 30, 2026
64.63Accuracy
11
May 12, 2026
76.1Mean Accuracy
10
Feb 26, 2026
32.2Pass@1
9
Feb 26, 2026
75.2Accuracy
9
Jun 9, 2026
71.02Accuracy
8
Jun 4, 2026
95Accuracy
7
Feb 26, 2026
71.5Accuracy
5
Jun 2, 2026
31Level 1 Accuracy
4
May 12, 2026
57.36Murder Mystery Score
4
May 12, 2026
29.5Macro Accuracy
4
Apr 24, 2026
31.6Accuracy (macro)
4
Apr 24, 2026
34.3Macro Accuracy
4
Apr 24, 2026
74Accuracy
4
Feb 26, 2026
90Accuracy
3
Jun 2, 2026
100Accuracy
3
Jun 2, 2026
100Accuracy
3
Jun 2, 2026
61Accuracy
3
Jun 2, 2026
58Accuracy
3
Jun 2, 2026
68Accuracy
3
Jun 2, 2026
61.6Average Score
3
Feb 26, 2026
95Success Rate
2
Jun 29, 2026