List your company

Geometry Problem Solving on PyGeoX Medium

57AccuracyProprietary-A
38.2843.144852.86Jun 8, 2026

Evaluation results

MethodLinks
Proprietary-AInference=default, Evaluation Protocol=zero-shot2026.0657
Proprietary-BInference=reasoning, Evaluation Protocol=zero-shot2026.0654
Qwen3-8B-RLInference=RL-trained, Evaluation Protocol=zero-shot2026.0650
Proprietary-CInference=reasoning, Evaluation Protocol=zero-shot2026.0649
DeepSeek-V3.2Inference=default, Evaluation Protocol=zero-shot2026.0639