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

Code Generation on LiveCodeBench (@1, @32, @128, AUC, Tokens)

75.2Rate @32 Score

DFPO (Orth-Proj)

35.26445.6325666.368Jan 25, 2026Feb 5, 2026Feb 17, 2026Feb 28, 2026Mar 12, 2026Mar 23, 2026Apr 4, 2026
Updated 3d ago

Evaluation Results

MethodLinks
2026.04
75.2----
2026.04
75.1----
2026.04
73.9----
2026.04
71.6----
2026.04
71.6----
2026.04
71----
2026.04
70.9----
2026.04
69.1----
2026.04
64.7----
2026.04
64.5----
2026.01
47.537.249.844.15.8
2026.01
46.836.449.143.45.9
2026.01
46.536.548.843.26
2026.01
45.535.147.8424.4
2026.01
44.535.546.841.56
2026.01
42.132.444.138.83.6
2026.01
36.826.838.933.51.6
2026.01
-36.850.244.44.6