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

Geometry3K

Benchmarks

Task NameDataset NameSOTA ResultTrend
Geometry Problem SolvingGeometry3K (test)
Choice Accuracy95.2
76
Geometric ReasoningGeometry3K
Accuracy@169.1
42
Geometry problem solvingGeometry3K
Accuracy88.8
41
Mathematical ReasoningGeometry3k
Accuracy79.9
26
Geometric ReasoningGeometry3K (test)
Accuracy40.93
22
Geometry ReasoningGeometry3K (val)
Accuracy44.93
21
Geometry Problem SolvingGeometry3K 1.0 (test)
Overall Score90.9
12
Multi-modal reasoningGeometry3K
Pass@137.9
11
Geometric ReasoningGeometry3K
Avg@8 Accuracy44.11
10
Geometry Problem SolvingIMP-Geometry3K
Accuracy76
10
Geometric ReasoningGeometry3K Geo-Sub-Aux (test)
Accuracy52.4
8
Geometric ReasoningGeometry3K Geo-Sub (test)
Accuracy50.9
8
Geometry reasoningGeometry3k (test)
Test Score48.11
8
Geometry reasoningGeometry3K
Geometry Score66.7
7
Geometric ReasoningGeometry3K mid-school level
Accuracy81.03
6
Visual ReasoningGeometry3K (test)
Pass@1 Accuracy56.18
5
Multi-step theorem predictionGeometry3K
Accuracy95.16
3
MathGeometry3K
Accuracy39
2
Showing 18 of 18 rows