Hybrid Reasoning on MATH_MIX (test)
88AccuracyAgent-GWO
Evaluation results
| Method | Links | |
|---|---|---|
| Agent-GWOBackbone=Gemma-3-12b-it2026.04 | 88 | |
| Agent-GWOBackbone=GPT-4o-mini2026.04 | 87.5 | |
| AoTBackbone=GPT-4o-mini2026.04 | 86.9 | |
| ToTBackbone=GPT-4o-mini2026.04 | 85.6 | |
| AFlowBackbone=GPT-4o-mini2026.04 | 85.3 | |
| GoTBackbone=GPT-4o-mini2026.04 | 84.4 | |
| AoTBackbone=Gemma-3-12b-it2026.04 | 84 | |
| Self-RefineBackbone=GPT-4o-mini2026.04 | 83.8 | |
| Agent-GWOBackbone=Qwen2.5-Coder-7B-Instruct2026.04 | 83.5 | |
| ToTBackbone=Gemma-3-12b-it2026.04 | 83.5 | |
| AFlowBackbone=Gemma-3-12b-it2026.04 | 82.6 | |
| GoTBackbone=Gemma-3-12b-it2026.04 | 82.4 | |
| CoT-SC/n=5Backbone=GPT-4o-mini2026.04 | 81.6 | |
| CoTBackbone=GPT-4o-mini2026.04 | 80.1 | |
| Self-RefineBackbone=Gemma-3-12b-it2026.04 | 79.6 | |
| AoTBackbone=Qwen2.5-Coder-7B-Instruct2026.04 | 78.8 | |
| ToTBackbone=Qwen2.5-Coder-7B-Instruct2026.04 | 77.8 | |
| AFlowBackbone=Qwen2.5-Coder-7B-Instruct2026.04 | 77 | |
| CoT-SC/n=5Backbone=Gemma-3-12b-it2026.04 | 77 | |
| GoTBackbone=Qwen2.5-Coder-7B-Instruct2026.04 | 76.8 | |
| CoTBackbone=Gemma-3-12b-it2026.04 | 75.1 | |
| Self-RefineBackbone=Qwen2.5-Coder-7B-Instruct2026.04 | 74.6 | |
| CoT-SC/n=5Backbone=Qwen2.5-Coder-7B-Instruct2026.04 | 73.9 | |
| CoTBackbone=Qwen2.5-Coder-7B-Instruct2026.04 | 70.4 |