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

Question Answering on TruthfulQA (Primary Truthfulness Metric)

58.4TruthfulQA

GRPO

25.84834.29942.7551.201Mar 27, 2026Mar 28, 2026Mar 30, 2026Mar 31, 2026Apr 2, 2026Apr 3, 2026Apr 5, 2026
Updated 12d ago

Evaluation Results

MethodLinks
2026.03
58.4
2026.03
57.98
2026.03
56.42
2026.03
56.11
2026.03
55.77
2026.03
54.87
2026.03
54.74
2026.03
54.56
2026.03
54.44
2026.03
54.44
2026.03
54.38
2026.03
54.23
2026.03
54.22
2026.03
54.05
2026.03
54.05
2026.03
52.42
2026.04
30.2
2026.04
29.2
2026.04
29
2026.04
29
2026.04
28.9
2026.04
27.1