Multi-question Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
14.9Exact Match
6
May 29, 2026
8.4Exact Match
6
May 29, 2026
18.5Exact Match
6
May 29, 2026
15.4Exact Match Accuracy
6
May 29, 2026
30.6Exact Match
6
May 29, 2026
17.9Exact Match (EM)
6
May 29, 2026
35.8Exact Match (EM)
6
May 29, 2026
32Exact Match Accuracy (3Q)
6
May 29, 2026