Multi-hop Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
54.4Exact Match (EM)
55
May 20, 2026
100CoT Match Rate
54
Mar 13, 2026
58.9F1
54
Apr 21, 2026
46.3EM
53
Jun 9, 2026
22.6Exact Match (EM)
51
May 20, 2026
78.4EM
51
Mar 10, 2026
46.49Average QA Score
50
Apr 28, 2026
15.35EM
50
Apr 28, 2026
55Exact Match (EM)
50
Apr 28, 2026
40EM
50
Apr 28, 2026
74.2Exact Match (EM)
50
Jun 12, 2026
36.8EM
50
May 28, 2026
75.97F1
48
Feb 26, 2026
86Accuracy
46
Jun 18, 2026
48.4String Accuracy
44
Jun 2, 2026
70.3String Accuracy
44
Jun 2, 2026
47.6Accuracy
44
May 14, 2026
43.6Accuracy (2Wiki)
44
May 14, 2026
69.3Exact Match (EM)
44
May 12, 2026
46.1F1
38
Mar 17, 2026
81.5Answer F1
38
Mar 4, 2026
89.342Wiki Accuracy
37
Apr 24, 2026
68Pass@1
36
Apr 21, 2026
77.4Pass@1
36
Apr 21, 2026
18.8Pass@1
36
Apr 21, 2026