Logical Reasoning on Reasoning Gym Color Cube
32Accuracy (*)TokUR (EU)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| TokUR (EU)Base Model=Qwen-2.5-3B-Instruct2025.05 | 32 | 60.08 | 33.71 | |
| TokUR (TU)Base Model=Qwen-2.5-3B-Instruct2025.05 | 31.33 | 58.33 | 31.14 | |
| Self-CertaintyBase Model=Qwen-2.5-3B-Instruct2025.05 | 28 | 50.62 | 30.8 | |
| TokUR (AU)Base Model=Qwen-2.5-3B-Instruct2025.05 | 28 | 56.57 | 30.02 | |
| CoT (Lower-Bound)Base Model=Qwen-2.5-3B-Instruct2025.05 | 26 | — | — | |
| LLBase Model=Qwen-2.5-3B-Instruct2025.05 | 24 | 46.36 | 26.01 | |
| DeepConfBase Model=Qwen-2.5-3B-Instruct2025.05 | 24 | 49.01 | 24.41 | |
| PEBase Model=Qwen-2.5-3B-Instruct2025.05 | 22 | 46.73 | 28.75 |