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

Conversational Alignment on Alpaca-Evals

58.12Alpaca-Evals Score

ODRPO Gini-Med (GRPO)

53.034454.354755.67556.9953May 12, 2026
Updated 2mo ago

Evaluation Results

MethodLinks
2026.05
58.12
2026.05
57.21
2026.05
57.11
2026.05
56.91
2026.05
56.91
2026.05
56.61
2026.05
56.41
2026.05
56.34
2026.05
55.97
2026.05
55.33
2026.05
55.29
2026.05
53.23