Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-Agent Pathfinding on 5-agent 6x6 (evaluation set)
Loading...
26
Valid Rate
Qwen3-4B + GRPO + LLM-as-Environment-Engineer
1.7368
8.0359
14.335
20.6341
Jun 16, 2026
Valid Rate
Optimal Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Valid Rate
Optimal Rate
Qwen3-4B + GRPO + LLM-as-Environment-Engineer
2026.06
26
17.33
Qwen3-4B + GRPO
training_config=random
2026.06
21.33
14
Kimi-K2.5
2026.06
18
12
Grok-4.2
2026.06
13.33
9.33
GPT-5.4
2026.06
10
6
Gemini-3.1-Pro
2026.06
6.67
4.67
Qwen3-4B
variant=base
2026.06
2.67
2.67
Feedback
Search any
task
Search any
task