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

Geometric Reasoning on Geo3K

55.2Accuracy

NoisyRollout-7B

38.653642.949347.24551.5407May 27, 2026May 31, 2026Jun 4, 2026Jun 9, 2026Jun 13, 2026Jun 17, 2026Jun 22, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.05
55.2--
2026.05
49.851,438.1852.1
2026.05
49.731,618.4146.1
2026.05
49.564,005.67-33.5
2026.05
49.343,001.24-
2026.05
48.9--
2026.05
48.893,025.72-0.8
2026.05
48.8--
2026.05
46.9--
2026.05
45.8--
2026.05
45.3--
2026.06
44.8--
2026.05
44.4--
2026.05
41.1--
2026.06
41.08--
2026.06
39.29--