Multi-hop Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
81.5F1 Score
17
Apr 10, 2026
71.2F1 Score
17
Apr 10, 2026
59MBE
17
Mar 4, 2026
28.1MBE
17
Mar 4, 2026
71.85EM
17
Feb 27, 2026
58.1Exact Match (EM)
17
Feb 27, 2026
79.39Exact Match (EM)
17
Feb 27, 2026
66.51Exact Match (EM)
17
Feb 27, 2026
56.91EM
17
Feb 27, 2026
31.8Exact Match (EM)
17
Feb 26, 2026
33.2EM Accuracy
17
Feb 26, 2026
22.6Exact Match (%)
16
Jul 8, 2026
36.6Accuracy
16
Jun 17, 2026
31.1Contain-Acc.
16
Jun 12, 2026
67.1Containment Accuracy
16
Jun 12, 2026
51.43EM
16
Jun 1, 2026
66.41EM
16
Jun 1, 2026
50.78Exact Match (EM)
16
Jun 1, 2026
38.34Exact Match (EM)
16
Jun 1, 2026
39.84Exact Match (EM)
16
Jun 1, 2026
57.81EM
16
Jun 1, 2026
44.56EM
16
May 29, 2026
24.12EM
16
May 29, 2026
48.89EM
16
May 29, 2026
50Score
16
Apr 16, 2026