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

Geometric Reasoning on GeoSym-Bench (test)

18.79Accuracy (%)

Qwen3-VL-8B + GeoSym

11.260413.215215.1717.1248May 10, 2026
Updated 15d ago

Evaluation Results

MethodLinks
2026.05
18.79
15.66
14.68
2026.05
11.55