Multi-hop Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
82.1EM
559
May 20, 2026
80.79F1
334
Jul 7, 2026
77.4F1 Score
294
Apr 23, 2026
73.9EM
247
Jun 30, 2026
74.9Exact Match
215
May 8, 2026
46EM
209
Jun 29, 2026
76.4F1 Score
175
Jul 3, 2026
50.2Exact Match (EM)
167
Jun 12, 2026
55.68F1
151
Jul 7, 2026
56Exact Match
128
Apr 10, 2026
48.35F1
125
Jun 18, 2026
67.5EM
110
Jun 4, 2026
85.56F1 Score
85
Jun 5, 2026
74.9F1
79
Feb 26, 2026
86.07F1
77
Apr 14, 2026
80LLM Judge Score
72
Jun 5, 2026
74.2Accuracy
72
Feb 26, 2026
85.81Answer F1
72
Jun 30, 2026
74.3F1
71
May 12, 2026
47.1Exact Match (EM)
66
May 19, 2026
57.02HotpotQA Score
65
Jun 11, 2026
64.6HotpotQA Score
64
Jun 11, 2026
75.9Answer F1
64
Feb 26, 2026
75.2Accuracy
62
Mar 11, 2026
25.3Exact Match (EM)
58
Apr 10, 2026