HotpotQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
HotpotQA 10K context
84.8Accuracy
19
HotpotQA (val)
57.3F1 Score
19
HotpotQA full wiki (dev)
86.5F1 Score
19
HotpotQA full wiki (dev)
88.17PEM
19
HotpotQA 128K
25Accuracy
18
HotpotQA (held-out)
81.2F1 Score
18
HotpotQA
0.07Attack Success Rate (ASR)
18
HotpotQA
91.47F1 Score
18
HotpotQA
4.4CR Score
18
HotPotQA
84Accuracy
18
HotpotQA (full-slice)
53.8HotpotQA Full-Slice Distance (d)
18
HotpotQA
83.2F1 Score
18
HotPotQA
1Unlearn Efficacy
18
HotPotQA
4.7Average Steps per Episode
18
HotpotQA Passage
60.03Score
18
HotpotQA Fixed Chunk 2048
60.03QA Score
18
HotpotQA
61.13EM
18
HotpotQA (test)
68EM
18
HotpotQA
38.2EM
18
HotpotQA ID
56.8Accuracy
18
HotpotQA
52.4Cover EM
18
HotpotQA (test val)
47.1Exact Match (EM)
18
HotpotQA 100K context
79.4Accuracy
18
HotpotQA 30K context
78.9Accuracy
18
HotpotQA (test)
0.822F1
18