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

Pairwise Accuracy on RubricBench 1.0 (test)

81.8Pairwise Accuracy (Chat)

SVR

41.55252.00162.4572.899Jun 6, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
81.885.582.483.483.882.8
2026.06
80.685.583.686.492.583.1
2026.06
65.36456.853.852.857.6
2026.06
6369.4585277.561
2026.06
55.97165.660.151.260.3
2026.06
55.270.265.660.95060.1
2026.06
54.769.464.45948.859
2026.06
53.166.169.664.956.361.1
2026.06
5143.651.650.228.848.6
2026.06
5069.466.863.136.357.9
2026.06
48.358.161.254.22552
2026.06
44.669.460.858.76055.2
2026.06
43.139.550.447.253.846