Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Multi-hop Question Answering on HotpotQA (official evaluation)

33.2EM Accuracy

M3PO

8.2414.7221.227.68Dec 1, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
33.2
2025.12
31.6
2025.12
30.8
2025.12
30.4
2025.12
29.9
2025.12
28.7
2025.12
27.3
2025.12
25.6
2025.12
25.5
2025.12
22.8
2025.12
20.2
2025.12
18.7
2025.12
18.6
2025.12
18.3
2025.12
13.3
2025.12
12.9
2025.12
9.2