Multi-Agent Pathfinding on 5-agent 7x7 (test)
22Valid RateQwen3-4B + GRPO + LLM-as-Environment-Engineer
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Qwen3-4B + GRPO + LLM-as-Environment-Engineer2026.06 | 22 | 12 | |
| Grok-4.22026.06 | 16.67 | 10.67 | |
| Kimi-K2.52026.06 | 16 | 7.33 | |
| Qwen3-4B + GRPOtraining_config=random2026.06 | 16 | 10 | |
| GPT-5.42026.06 | 10 | 4 | |
| Gemini-3.1-Pro2026.06 | 5.33 | 4 | |
| Qwen3-4Bvariant=base2026.06 | 0.67 | 0 |