Code Generation on MBPP (Acc (%), Avg Cost)
49.23Accuracy (%)MACA (Ours)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MACA (Ours)Model=Llama-3.1-8B, Structure Adaptation=✓, Orchestration Adaptation=✓2026.05 | 49.23 | 2,412.3 | |
| AgentVerseModel=Llama-3.1-8B, Structure Adaptation=✗, Orchestration Adaptation=✗2026.05 | 48.94 | 5,017.5 | |
| MacNetModel=Llama-3.1-8B, Structure Adaptation=✗, Orchestration Adaptation=✗2026.05 | 47.83 | 6,794.8 | |
| Self-RefineModel=Llama-3.1-8B, Structure Adaptation=✗, Orchestration Adaptation=✗2026.05 | 47.53 | 498.7 | |
| PuppeteerModel=Llama-3.1-8B, Structure Adaptation=✗, Orchestration Adaptation=✓2026.05 | 47.17 | 3,696.5 | |
| SC (CoT×5)Model=Llama-3.1-8B, Structure Adaptation=✗, Orchestration Adaptation=✗2026.05 | 46.36 | 1,206.9 | |
| VanillaModel=Llama-3.1-8B, Structure Adaptation=✗, Orchestration Adaptation=✗2026.05 | 45.97 | 138.4 | |
| CoTModel=Llama-3.1-8B, Structure Adaptation=✗, Orchestration Adaptation=✗2026.05 | 44.62 | 146.1 | |
| ComplexCoTModel=Llama-3.1-8B, Structure Adaptation=✗, Orchestration Adaptation=✗2026.05 | 44.18 | 247.4 | |
| MaASModel=Llama-3.1-8B, Structure Adaptation=✓, Orchestration Adaptation=✗2026.05 | 42.05 | 3,215.7 | |
| AgentPruneModel=Llama-3.1-8B, Structure Adaptation=✓, Orchestration Adaptation=✗2026.05 | 31.44 | 2,787.8 | |
| DyLANModel=Llama-3.1-8B, Structure Adaptation=✗, Orchestration Adaptation=✗2026.05 | 12.5 | 11,647.8 |