Question Answering on MMLU Formal Logic (test)
73.67AccuracyMA-PoP
Evaluation Results
| Method | Links | |
|---|---|---|
| MA-PoP2026.05 | 73.67 | |
| LLP2026.05 | 72.01 | |
| Sparse MADT=32026.05 | 71.43 | |
| Sparse MADT=52026.05 | 70.84 | |
| Self-Consistency2026.05 | 70.84 | |
| Decentr. MADT=22026.05 | 70.63 | |
| Decentr. MADT=52026.05 | 69.84 | |
| Decentr. MADT=32026.05 | 69.04 | |
| Sparse MADT=22026.05 | 66.67 | |
| ISP2026.05 | 64.29 | |
| Free MADT=52026.05 | 62.7 | |
| Majority Voting2026.05 | 61.9 | |
| Free MADT=22026.05 | 61.11 | |
| Free MADT=32026.05 | 61.11 | |
| Centr. MADT=32026.05 | 60.32 | |
| Centr. MADT=52026.05 | 60.32 | |
| Falcon-7BNumber of Agents (M)=12026.05 | 58.73 | |
| Centr. MADT=22026.05 | 56.35 | |
| Gemma-9BNumber of Agents (M)=12026.05 | 51.59 | |
| Qwen-7BNumber of Agents (M)=12026.05 | 49.05 | |
| Llama-8BNumber of Agents (M)=12026.05 | 37.94 | |
| Mistral-7BNumber of Agents (M)=12026.05 | 37.3 |