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

Multi-hop Question Answering on HotPotQA (MBE, Recall, Searches)

28.1MBE

Zero-Shot

26.76835.75944.7553.741Jul 10, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.07
28.1--
2025.07
29.1--
2025.07
37.652.61
2025.07
39.3835.94.02
2025.07
39.752.61
2025.07
42.750.11.77
2025.07
45.552.491
2025.07
45.666.93.31
2025.07
46.155.43
2025.07
46.248.21.28
2025.07
49.363.43.23
2025.07
50.464.82.75
2025.07
57.6669.12.22
2025.07
5870.42.89
2025.07
58.264.34
58.570.42.89
61.470.42.89