Code Reasoning on R-Bench-T Code
49.91AccuracyDFT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DFTModel=Qwen3 32B2025.10 | 49.91 | 49.91 | |
| SFTModel=Qwen3 32B2025.10 | 48.54 | 48.54 | |
| RandomModel=Qwen3 32B2025.10 | 47.71 | 47.71 | |
| VCOREModel=Qwen3 32B2025.10 | 45.7 | 45.7 | |
| iw-SFTModel=Qwen3 32B2025.10 | 45.43 | 45.43 | |
| OriginalModel=Qwen3 32B2025.10 | 41.13 | 41.13 | |
| iw-SFTModel=Qwen3 8B2025.10 | 35.47 | 35.47 | |
| DFTModel=Qwen3 8B2025.10 | 35.1 | 35.1 | |
| SFTModel=Qwen3 8B2025.10 | 34.92 | 34.92 | |
| VCOREModel=Qwen3 8B2025.10 | 33.82 | 33.82 | |
| RandomModel=Qwen3 8B2025.10 | 33.46 | 33.46 | |
| VCOREModel=Qwen3 4B2025.10 | 30.25 | 30.25 | |
| DFTModel=Qwen3 4B2025.10 | 30.16 | 30.16 | |
| iw-SFTModel=Qwen3 4B2025.10 | 29.34 | 29.34 | |
| RandomModel=Qwen3 4B2025.10 | 29.34 | 29.34 | |
| SFTModel=Qwen3 4B2025.10 | 28.34 | 28.34 | |
| OriginalModel=Qwen3 8B2025.10 | 24.86 | 24.86 | |
| OriginalModel=Qwen3 4B2025.10 | 23.13 | 23.13 | |
| OriginalModel=LLaMA3.1 8B Instruct2025.10 | 21.76 | 21.76 | |
| VCOREModel=LLaMA3.1 8B Instruct2025.10 | 6.22 | 6.22 | |
| SFTModel=LLaMA3.1 8B Instruct2025.10 | 5.21 | 5.21 | |
| RandomModel=LLaMA3.1 8B Instruct2025.10 | 4.75 | 4.75 | |
| DFTModel=LLaMA3.1 8B Instruct2025.10 | 3.47 | 3.47 | |
| iw-SFTModel=LLaMA3.1 8B Instruct2025.10 | 3.38 | 3.38 |