Multi-hop Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
42.67Recall@2
13
Apr 21, 2026
72.65Recall@2
13
Apr 21, 2026
82.1Accuracy
12
May 27, 2026
89.6EM
11
Feb 26, 2026
93.15Hits@1
10
Apr 15, 2026
82.2Hits@1
10
Apr 15, 2026
48.44Exact Match (EM)
10
Mar 16, 2026
43.12Accuracy
10
Feb 26, 2026
36.67Accuracy
9
Jun 17, 2026
42.3Accuracy
8
Jun 12, 2026
10.4Accuracy
8
May 28, 2026
97.3Accuracy (1 Hop)
8
May 12, 2026
15EM
7
Mar 24, 2026
24Exact Match (EM)
7
Mar 24, 2026
69.9F1 Score
7
Feb 26, 2026
70.3Exact-match Accuracy
6
Jun 30, 2026
0.722Accuracy
5
Jun 17, 2026
64.63Accuracy
5
May 26, 2026
37.1Accuracy (Correct)
5
May 15, 2026
88.4ASR
4
Apr 23, 2026
64.47Relevancy
4
Mar 27, 2026
39.38Relevancy
4
Mar 27, 2026
60.28Relevancy
4
Mar 27, 2026
96.76Relevancy
4
Mar 27, 2026
79API Success Rate
4
Feb 26, 2026