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

Multi-Hop Question Answering on HotpotQA (Accuracy)

44.2Accuracy

SkillOrchestra+

12.79220.94629.137.254Feb 23, 2026Feb 25, 2026Feb 27, 2026Mar 1, 2026Mar 3, 2026Mar 5, 2026Mar 7, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
44.2
2026.03
42.1
2026.03
41.6
2026.02
39
2026.03
38.6
2026.03
38.6
2026.03
37
2026.03
36.9
2026.02
35.2
2026.03
34.6
2026.03
34.5
2026.03
32.4
2026.02
27.8
2026.03
27.4
2026.02
26.8
2026.02
24.4
2026.03
23.7
2026.02
23.6
2026.02
23.4
2026.02
23.4
2026.02
22.4
2026.02
21.6
2026.02
21.6
2026.03
20.8
2026.02
20.6
2026.02
19.8
2026.02
19.8
2026.02
16.8
2026.02
15.4
2026.02
14