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

Truthful Question Answering on TruthfulQA (MC1 & MC2)

81.94MC1 Score

MARI

24.542439.443754.34569.2463May 27, 2026
Updated 5d ago

Evaluation Results

MethodLinks
2026.05
81.9470.59
2026.05
67.9371.92
2026.05
66.6764.96
2026.05
65.7468.84
2026.05
64.3561.13
2026.05
61.8167.39
2026.05
59.855.6
2026.05
55.648.9
2026.05
52.3362.8
2026.05
50.5852.51
2026.05
50.4650.2
2026.05
49.8359.69
2026.05
47.9266.23
2026.05
46.8468.33
2026.05
46.7864.43
2026.05
44.7760.05
2026.05
43.9859.81
2026.05
42.263.65
2026.05
40.4462.25
2026.05
39.8160.71
2026.05
36.1953.12
2026.05
36.1953.12
2026.05
35.4553.95
2026.05
35.4553.95
2026.05
33.846.9
2026.05
33.551.9
2026.05
32.8452.33
2026.05
32.7650.75
2026.05
32.0349.51
2026.05
31.349.38
2026.05
31.1252.32
2026.05
30.8149.8
2026.05
29.6446.95
2026.05
28.745.03
2026.05
28.4449.25
2026.05
26.7548.31