HotpotQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
HOTPOTQA
26.1EM
7
HotpotQA 1% v1.1 (train)
63.1F1
7
HotpotQA
76.4F1 Score
7
HotpotQA
30,205.1Average Token Count
7
HotpotQA
72.94Accuracy
7
HotpotQA
80.4FCR
7
HotpotQA (HQA) (val)
1.6TFLOPs (Read)
7
HotpotQA (dev)
97.52EM
7
HotpotQA (50% coverage)
79.37F1
7
HotpotQA (20% coverage)
86.69F1
7
HotpotQA (dev)
88.89Passage EM
7
ADVHOTPOT (test)
0.688AUC
7
HotpotQA Challenge Set 1.0 (test)
50.15F1 Score
7
HotpotQA
30.04Answer in Context
7
HotpotQA
94.5Accuracy (No Attack)
7
HotpotQA
78.07LLM Judge Score
6
HotpotQA
45EM
6
HotpotQA
73.2F1
6
HotpotQA
97.65TPR
6
HotpotQA (val)
75Initial Token-level F1
6
HotpotQA
63.6Truthfulness
6
HotpotQA in-domain (test)
37.6Accuracy
6
HotpotQA (cross)
0.81Correlation Coefficient (rho)
6
HotpotQA (DPO)
0.9Rho
6
HotpotQA
58.2Exact Match (EM)
6