Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Spatial Decision-Making on FrozenLake size 8
Loading...
81.74
Success Rate
Gemini-3-Flash
35.4704
47.4827
59.495
71.5073
Jun 1, 2026
Success Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Success Rate
Gemini-3-Flash
Setting=Naive
2026.06
81.74
SpecFlow
Setting=Flow(image) +...
2026.06
81.73
DiffThinker
Setting=Flow(image) Only
2026.06
79.83
GPT-5.1
Setting=Naive
2026.06
79.68
MVoT
Setting=Autoregressive
2026.06
78.94
Qwen3-VL-32B
Setting=SFT
2026.06
62.91
Qwen3-VL-32B
Setting=GRPO
2026.06
59.47
Qwen3-VL-8B
Setting=SFT
2026.06
53.72
Qwen3-VL-8B
Setting=GRPO
2026.06
50.21
Qwen3-VL-32B
Setting=Naive
2026.06
48.12
Qwen3-VL-8B
Setting=Naive
2026.06
37.25
Feedback
Search any
task
Search any
task