Evidence Retrieval
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
22.5Hit@2
56
Jun 30, 2026
68Hit@2
52
Jun 30, 2026
21.6Recall@10
26
May 8, 2026
63.5Recall@10
26
May 8, 2026
63.8Recall@10
26
May 8, 2026
59.5Recall@10
26
May 8, 2026
78.1Recall@10
26
May 8, 2026
57.55F1 Score
14
Feb 26, 2026
54.19MAP
13
Feb 26, 2026
57.34F1 Score
13
Feb 26, 2026
8.4Hit@2
10
Jun 30, 2026
2.7Hit@2
10
Jun 30, 2026
87.02Recall@20
10
May 28, 2026
88.48Recall@20
10
May 28, 2026
93.8Recall
10
Mar 4, 2026
87.7Recall
10
Mar 4, 2026
31.9Recall@5
9
May 8, 2026
40.7Recall@5
9
May 8, 2026
46Hit@5
8
Jun 30, 2026
38.4Hit@5
8
Jun 30, 2026
55Accuracy
7
Jul 9, 2026
82.16F1 Score
7
Jun 9, 2026
5.9Recall@1
7
Feb 26, 2026
24.4Recall@1
7
Feb 26, 2026
56Recall@1
7
Feb 26, 2026