Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Question Answering on TruthfulQA (Domain-level split, o=3)

92.8Accuracy

W/O Decontamination

24.26442.05759.8577.643Jan 27, 2026
Updated 4d ago

Evaluation Results

MethodLinks
2026.01
92.80.476
2026.01
92.30.471
2026.01
89.90.327
2026.01
89.90.447
2026.01
88.50.313
2026.01
86.10.592
2026.01
86.10.592
2026.01
84.60.274
2026.01
79.30.524
2026.01
740.471
2026.01
72.60.274
2026.01
64.10.069
2026.01
63.60.367
2026.01
62.50.173
2026.01
61.40.042
2026.01
57.2-
2026.01
56.50.296
2026.01
55.30.101
2026.01
45.2-
2026.01
39.40.178
2026.01
26.9-