Maze Navigation on Maze (test)
0.8Success RateFed-SE
Evaluation results
| Method | Links | |
|---|---|---|
| Fed-SEBase Model=Llama2-7B2025.12 | 0.8 | |
| Fed-SEBase Model=Qwen2.5-7B2025.12 | 0.68 | |
| Fed-SEBase Model=Qwen3-8B2025.12 | 0.64 | |
| LocalBase Model=Qwen2.5-7B2025.12 | 0.48 | |
| CentralizedBase Model=Qwen2.5-7B2025.12 | 0.4 | |
| Pre-TrainedBase Model=Qwen3-1.7B2025.12 | 0.4 | |
| Fed-SEBase Model=Qwen3-1.7B2025.12 | 0.4 | |
| Fed-SEBase Model=Qwen2.5-3B2025.12 | 0.36 | |
| Pre-TrainedBase Model=Qwen2.5-7B2025.12 | 0.32 | |
| CentralizedBase Model=Llama2-7B2025.12 | 0.32 | |
| LocalBase Model=Qwen3-1.7B2025.12 | 0.32 | |
| LocalBase Model=Qwen3-8B2025.12 | 0.32 | |
| CentralizedBase Model=Qwen3-8B2025.12 | 0.32 | |
| FedITBase Model=Qwen2.5-7B2025.12 | 0.28 | |
| FedITBase Model=Llama2-7B2025.12 | 0.28 | |
| LocalBase Model=Qwen2.5-3B2025.12 | 0.28 | |
| Pre-TrainedBase Model=Qwen2.5-3B2025.12 | 0.24 | |
| Pre-TrainedBase Model=Llama2-7B2025.12 | 0.2 | |
| LocalBase Model=Llama2-7B2025.12 | 0.2 | |
| CentralizedBase Model=Qwen2.5-3B2025.12 | 0.2 | |
| FedITBase Model=Qwen2.5-3B2025.12 | 0.2 | |
| FedITBase Model=Qwen3-8B2025.12 | 0.12 | |
| FedITBase Model=Qwen3-1.7B2025.12 | 0.08 | |
| Pre-TrainedBase Model=Qwen3-8B2025.12 | 0.08 | |
| CentralizedBase Model=Qwen3-1.7B2025.12 | 0 |