Search

Few direct matches — filled in with the latest updates.

Tag: #asymptote1 results

New Benchmark Tests Whether AI Can Draw Geometry

New Benchmark Tests Whether AI Can Draw Geometry

Researchers introduce an open-source benchmark separating olympiad geometry solving from accurate diagram construction. Across 954 problems, current foundation models achieved only a 36.14% average diagram compilation success rate, revealing a substantial gap between mathematical reasoning and faithful visual construction.

ArXiv AIResearch22h ago#geometry-reasoning#benchmark#asymptote