2Wiki
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
2Wiki
0.019Retrieval Latency (s)
8
2Wiki
42.4Score
8
2Wiki Normal
23.63F1 Score
8
2WIKI (test)
0.92Mean Correct Rate
8
2Wiki
0Tokens (M)
7
2Wiki Extreme
26.99F1 Score
7
2Wiki Noisy
24.17F1 Score
7
2Wiki
67.6FCR
7
2Wiki out-of-domain (test)
37Accuracy
6
2Wiki-5Q
18.5Exact Match
6
2Wiki
57.9F1 (token-overlap)
6
2wiki
36pass@1
6
2Wiki Platinum (test)
84.8Answer Rate
6
2Wiki
52.2Precision
5
2Wiki
52F1 Score
5
2wiki
71,305Triplets
5
2wiki (test)
26.5EM
5
2Wiki
51.83Accuracy
4
2Wiki 100 sampled queries
0.67Latency (s)
4
2Wiki
95.4Recall
3
2Wiki (dev)
46.6Exact Match (EM)
3
2Wiki
41.3EM
3
2Wiki (test)
53.6LASTOP@5
3
2Wiki 1,000 samples (test)
0.388F1 Score
3
2Wiki
39.66Success Rate
2