List your company

Spatial decision-making on Maze size 4

94.38Success Rate (%)Gemini-3-Flash
60.413669.231878.0586.8682Jun 1, 2026

Evaluation results

MethodLinks
Gemini-3-FlashSetting=Naive2026.0694.38
SpecFlowSetting=Flow(image) + AR(Text)2026.0694.12
DiffThinkerSetting=Flow(image) Only2026.0693.1
MVoTSetting=Autoregressive2026.0692.03
GPT-5.1Setting=Naive2026.0691.76
Qwen3-VL-32BSetting=SFT2026.0686.21
Qwen3-VL-32BSetting=GRPO2026.0683.02
Qwen3-VL-8BSetting=SFT2026.0678.92
Qwen3-VL-8BSetting=GRPO2026.0675.31
Qwen3-VL-32BSetting=Naive2026.0672.18
Qwen3-VL-8BSetting=Naive2026.0661.72