Mathematical Reasoning on Math Math500 Minerva
80.9Math500 ScoreUR2 (GPT Summary)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| UR2 (GPT Summary)Backbone=Qwen-2.5-7B2025.08 | 80.9 | 61 | 71 | |
| Open-Reasoner-ZeroBackbone=Qwen-2.5-7B, Category=CoT-RL Methods2025.08 | 80.7 | 58.8 | 69.8 | |
| UR2 (qw Summary)Backbone=Qwen-2.5-7B2025.08 | 80.2 | 59.4 | 69.8 | |
| Search-o1Backbone=GPT-4o-mini, Category=Advanced RAG Methods2025.08 | 78.6 | 68.3 | 73.5 | |
| Vanilla RLBackbone=Qwen-2.5-7B2025.08 | 78.2 | 59.4 | 68.8 | |
| CoTBackbone=GPT-4o-mini2025.08 | 78 | 65.6 | 71.8 | |
| RATBackbone=GPT-4o-mini, Category=Advanced RAG Methods2025.08 | 77.5 | 64.2 | 70.9 | |
| Standard RAGBackbone=GPT-4o-mini2025.08 | 77.1 | 68.4 | 72.8 | |
| SimpleRL-ZooBackbone=Qwen-2.5-7B, Category=CoT-RL Methods2025.08 | 77.1 | 50.7 | 63.9 | |
| CoTBackbone=Qwen-2.5-7B2025.08 | 76.6 | 59.4 | 68 | |
| General ReasonerBackbone=Qwen-2.5-7B, Category=CoT-RL Methods2025.08 | 76.6 | 62.1 | 68.8 | |
| Standard RAGBackbone=Qwen-2.5-7B2025.08 | 73.8 | 54.6 | 64.2 | |
| UR2 (qw summary)Backbone=Qwen-2.5-3B2025.08 | 68.4 | 47.4 | 57.9 | |
| Vanilla RLBackbone=Qwen-2.5-3B2025.08 | 68 | 43.9 | 56 | |
| UR2 (GPT summary)Backbone=Qwen-2.5-3B2025.08 | 67.8 | 45 | 57.2 | |
| Standard RAGBackbone=Qwen-2.5-3B2025.08 | 65.3 | 40.8 | 53.1 | |
| CoTBackbone=Qwen-2.5-3B2025.08 | 63.6 | 39.9 | 51.8 | |
| Self-AskBackbone=GPT-4o-mini, Category=Advanced RAG Methods2025.08 | 62.9 | 45.2 | 54.1 | |
| SAWG1=true, G2=true, G3=true, G4=true2026.05 | 55.1 | 13.42 | 34.26 | |
| STRG1=true, G2=true, G3=false, G4=true2026.05 | 54.9 | 11.31 | 33.11 | |
| UR2 (GPT summary)Backbone=LLaMA-3.1-8B2025.08 | 54.5 | 39 | 46.8 | |
| DELTG1=true, G2=true, G3=false, G4=false2026.05 | 49.2 | 12.04 | 30.62 | |
| UR2 (qw summary)Backbone=LLaMA-3.1-8B2025.08 | 48.6 | 38.4 | 43.5 | |
| CoTBackbone=LLaMA-3.1-8B2025.08 | 48.4 | 34.4 | 41.4 | |
| CLG1=true, G2=false, G3=true, G4=false2026.05 | 45.7 | 10.48 | 28.09 | |
| RandomG1=false, G2=true, G3=false, G4=true2026.05 | 45.65 | 9.28 | 27.47 | |
| Vanilla RLBackbone=LLaMA-3.1-8B2025.08 | 45.5 | 43.4 | 44.4 | |
| Standard RAGBackbone=LLaMA-3.1-8B2025.08 | 45 | 31.4 | 38.2 |