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

Truthfulness on TruthfulQA (Metric: TruthfulQA)

59.2TruthfulQA

DPO + MaPPO

36.94442.72248.554.278Mar 17, 2026Mar 25, 2026Apr 3, 2026Apr 12, 2026Apr 20, 2026Apr 29, 2026May 8, 2026
Updated 21d ago

Evaluation Results

MethodLinks
2026.05
59.2
2026.05
57.1
2026.05
56.4
2026.05
56.2
2026.05
55.9
2026.04
42.9
2026.04
42.8
2026.04
42.5
2026.04
42.3
2026.03
42
2026.04
41.5
2026.03
41.1
2026.04
40.5
2026.03
40.4
2026.04
40.1
2026.04
40
2026.03
39.8
2026.03
39.7
2026.04
39.7
2026.04
39.7
2026.03
39.4
2026.03
39
2026.04
38.6
2026.04
38.6
2026.04
38.4
2026.04
38.2
2026.04
38.2
2026.03
38.1
2026.04
38.1
2026.04
38
2026.04
37.8
2026.04
37.8