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

Single-Hop Question Answering on PopQA out-of-domain

52.4Accuracy

GEPO

-0.84812.97626.840.624Oct 30, 2025Dec 5, 2025Jan 10, 2026Feb 15, 2026Mar 23, 2026Apr 28, 2026Jun 4, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2025.10
52.4
2026.02
51.5
2025.10
51.5
2025.10
47.2
2025.10
46.1
2026.02
45.9
2025.10
44.8
2026.06
44.8
2026.02
44.6
2026.06
44.1
2025.10
42.4
2026.06
41.5
2026.02
39.7
2025.10
39.7
2025.10
37.8
2026.06
37.8
2025.10
19.9
19.9
2026.02
17.8
2026.02
17.1
2025.10
17.1
2026.02
11.4
2026.02
3.8
2026.02
1.2