Mathematical Reasoning on MATH (Solve Rate %)
58.2Solve Rate (MATH)AGENTCO-OP
Evaluation Results
| Method | Links | |
|---|---|---|
| AGENTCO-OPBackbone=GPT-4o-mini2026.05 | 58.2 | |
| AFlow*Backbone=Mixed2026.05 | 56.2 | |
| AFlowBackbone=GPT-4o-mini2026.05 | 53.1 | |
| MultiPersonaBackbone=GPT-4o-mini2026.05 | 50.8 | |
| CoT SCBackbone=GPT-4o-mini, Prompting Setting=5-shot2026.05 | 50.4 | |
| MedPromptBackbone=GPT-4o-mini2026.05 | 50 | |
| LLM-DebateBackbone=GPT-4o-mini2026.05 | 50 | |
| CoTBackbone=GPT-4o-mini2026.05 | 48.8 | |
| IOBackbone=GPT-4o-mini2026.05 | 48.6 | |
| Self RefineBackbone=GPT-4o-mini2026.05 | 46.1 | |
| ReConcileBackbone=GPT-4o-mini2026.05 | 44.1 | |
| ADASBackbone=GPT-4o-mini2026.05 | 35.4 |