Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-Agent Pathfinding on 5-agent 7x7 (test)
Loading...
22
Valid Rate
Qwen3-4B + GRPO + LLM-as-Environment-Engineer
-0.1832
5.5759
11.335
17.0941
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
22
12
Grok-4.2
2026.06
16.67
10.67
Kimi-K2.5
2026.06
16
7.33
Qwen3-4B + GRPO
training_config=random
2026.06
16
10
GPT-5.4
2026.06
10
4
Gemini-3.1-Pro
2026.06
5.33
4
Qwen3-4B
variant=base
2026.06
0.67
0
Feedback
Search any
task
Search any
task