List your company

Plane Geometry Problem Solving on Formalgeo (test)

0.857AccuracyMLLM Interpreter
0.561640.638320.7150.79168Jan 29, 2026

Evaluation results

MethodLinks
MLLM InterpreterTraining Data Size=5.5k2026.010.857
Gemini2.5-ProTraining Data Size=-2026.010.818
Gemini2.5-FlashTraining Data Size=-2026.010.805
DFE-GPSTraining Data Size=238k2026.010.753
GLM4.1-VTraining Data Size=-2026.010.734
Claude-Sonnet-4Training Data Size=-2026.010.691
Claude-Opus-4.1Training Data Size=-2026.010.691
GeoUniTraining Data Size=235k2026.010.598
GPT-4oTraining Data Size=-2026.010.58
Qwen2.5-VL 32BTraining Data Size=-2026.010.573