Aggregate Language and Logic Tasks on HumanEval++, MATH, MMLU-Redux
94.61Average AccuracyHieraMAS
Evaluation results
| Method | Links | |
|---|---|---|
| HieraMASMulti=true, Topo=true, Role=true, Node=true2026.02 | 94.61 | |
| Full-GraphMulti=true, Topo=false, Role=false, Node=false2026.02 | 92.78 | |
| AFlowMulti=true, Topo=true, Role=false, Node=false2026.02 | 92.69 | |
| Self-Consistency+CoTMulti=true, Topo=false, Role=false, Node=false2026.02 | 91.6 | |
| LLM-DebateMulti=true, Topo=false, Role=false, Node=false2026.02 | 91.45 | |
| MASRouterMulti=true, Topo=true, Role=true, Node=false2026.02 | 90.89 | |
| Random-GraphMulti=true, Topo=false, Role=false, Node=false2026.02 | 90.85 | |
| GDesignerMulti=true, Topo=true, Role=false, Node=false2026.02 | 90.68 | |
| CoTMulti=false, Topo=false, Role=false, Node=false2026.02 | 89.81 | |
| Self-ConsistencyMulti=true, Topo=false, Role=false, Node=false2026.02 | 89.77 | |
| BaseMulti=false, Topo=false, Role=false, Node=false2026.02 | 83.14 |