Mathematical Reasoning on ARC-c and GPQA-Diamond
82.3ARC-c (pass@1)RL
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| RLTraining Category=Reinforcement Learning, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 82.3 | 40.4 | 61.4 | |
| DYPOTraining Category=SFT and RL, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 81.8 | 41.4 | 61.6 | |
| SRFTTraining Category=SFT and RL, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 81.6 | 40.4 | 61 | |
| CHORDTraining Category=SFT and RL, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 81.1 | 40.4 | 60.8 | |
| LUFFYTraining Category=SFT and RL, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 80.5 | 39.9 | 60.2 | |
| SuperRLTraining Category=SFT and RL, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 77.8 | 36.9 | 57.4 | |
| SFTTraining Category=Supervised Fine-Tuning, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 75.2 | 24.7 | 50 | |
| ReLIFTTraining Category=SFT and RL, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 74.9 | 40.9 | 57.9 | |
| PRIME-ZeroTraining Category=Reinforcement Learning, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 73.3 | 18.2 | 45.8 | |
| SFT → RLTraining Category=SFT and RL, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 72.4 | 24.2 | 48.3 | |
| Oat-ZeroTraining Category=Reinforcement Learning, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 70.1 | 23.7 | 46.9 | |
| OpenReasoner-ZeroTraining Category=Reinforcement Learning, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 66.2 | 29.8 | 48 | |
| SimpleRL-ZeroTraining Category=Reinforcement Learning, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 30.2 | 23.2 | 26.7 | |
| Qwen2.5-Math-7BTraining Category=Base Model, Backbone=Qwen2.5-Math-7B, Inference Temperature=0.6, Option Shuffling=true2026.04 | 18.2 | 11.1 | 14.6 |