HotpotQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
HotpotQA ANCE
49.6MAP@10
3
HotpotQA
40.2Context Relevance Score
3
HotpotQA
20.88QPS
3
HotpotQA 200 Samples
0.621Planner
2
HotpotQA
35.8Session Coverage
2
HotPotQA
37.8Text EM
2
HotpotQA FullWiki n=250 shared-corpus stress run
218Answer Count
2
HotpotQA FullWiki
72.1Accuracy
2
HotpotQA
65.5Accuracy
2
HotpotQA full-wiki first 200 examples (val)
0Mean Jaccard
2
HotpotQA
38.32Success Rate
2
HotpotQA
83Sel-AUC
2
HotpotQA 104
1.9Harm Rate (HotpotQA 104)
2
HotpotQA 80q
8Blind Assessment Score
2
HotpotQA Offline (test)
61Pass@1
2
HotpotQA (Online Collection Split)
61.4Pass@1
2
HotpotQA 500 samples (evaluation)
71.6Accuracy
2
HotpotQA RAG v3
89.51Macro-F1
2
HotpotQA n=500 (downstream)
23EM
2
HotpotQA corrected external
100Relevance Accuracy
2
HotpotQA relation-heavy
100RelAcc
2
HotPotQA
1PTR Accuracy
2
HotpotQA
89.2Accuracy
2
HotpotQA
16.2PPE
2
HotpotQA KET-RAG
79.8Accuracy
2