HotpotQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
HotpotQA
173Total Triples Count
10
HotpotQA
51.6EM
10
HotpotQA (test)
70.3EM
10
HotpotQA
88.4Expected Utility
10
HotpotQA
73LLM Judge Score
10
HotpotQA (evaluation)
70.4Accuracy
10
HotpotQA
40.2Exact Match (EM)
10
HotpotQA RAG v3
90.75Macro F1
10
HotpotQA
55.321Accuracy
10
HotpotQA
56.3Accuracy (Acc)
10
HotpotQA
30.81Exact Match (EM)
10
HotpotQA
32EM
10
HOTPOTQA
61EM
10
HotpotQA
99Precision
10
HotpotQA
53.2F1 Score
10
HotpotQA + type
75.51AUROC
10
HotpotQA (original)
83.39AUROC
10
HotpotQA 49 (test)
61.99F1 Score
10
HotpotQA-WC (test)
0.86AuROC
10
HotpotQA 1,000 queries (test)
60.7EM
10
HotpotQA
60Exact Match
10
HotpotQA multi-hop (test)
44Avg@16
10
HotpotQA
27.87F1 Score
10
HotpotQA
36.73EM
10
HotpotQA
64.83F1 Score
10