List your company

Geometric Reasoning on Geomverse-109

28.9Accuracy@1VisuoThink
4.66810.95917.2523.541Apr 12, 2025

Evaluation results

MethodLinks
VisuoThinkBackbone=GPT-4o2025.0428.9
VisuoThinkBackbone=Claude-3.5-sonnet2025.0427.8
VisuoThink w/o rollout searchBackbone=Claude-3.5-sonnet2025.0426.7
VisuoThinkBackbone=Qwen2-VL-72B-Instruct2025.0425.6
VisuoThink w/o rollout searchBackbone=GPT-4o2025.0424.4
VisuoThink w/o rollout searchBackbone=Qwen2-VL-72B-Instruct2025.0419
VisualSketchpad + Equation SolverBackbone=Claude-3.5-sonnet2025.0417.8
VisualSketchpadBackbone=Claude-3.5-sonnet2025.0416.7
CoTBackbone=Claude-3.5-sonnet2025.0414.4
VisualSketchpad + Equation SolverBackbone=GPT-4o2025.0413.3
CoTBackbone=GPT-4o2025.0411.1
VisualSketchpad + Equation SolverBackbone=Qwen2-VL-72B-Instruct2025.0411.1
VisualSketchpadBackbone=GPT-4o2025.048.9
VisualSketchpadBackbone=Qwen2-VL-72B-Instruct2025.046.7
CoTBackbone=Qwen2-VL-72B-Instruct2025.045.6