Combinatorial Reasoning
100Graph AccuracyLlama 70B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Llama 70BAlgorithm=BoN2025.03 | 100 | 96.7 | 80 | |
| Llama 1BAlgorithm=SEM-CTRL2025.03 | 100 | 100 | 100 | |
| Llama 8BAlgorithm=SEM-CTRL2025.03 | 100 | 100 | 100 | |
| Llama 70BAlgorithm=SEM-CTRL2025.03 | 100 | 100 | 100 | |
| o4-miniAlgorithm=API2025.03 | 96.1 | 100 | 100 | |
| o1-previewAlgorithm=API2025.03 | 92 | 100 | 100 | |
| DeepSeek-R1Algorithm=API2025.03 | 85 | 100 | 100 | |
| Llama 8BAlgorithm=BoN2025.03 | 52.4 | 70 | 80 | |
| Llama 70BAlgorithm=Base2025.03 | 37.5 | 90 | 30 | |
| Llama 8BAlgorithm=Base2025.03 | 25 | 40 | 10 | |
| Llama 1BAlgorithm=BoN2025.03 | 24.3 | 0 | 0 | |
| Llama 1BAlgorithm=Base2025.03 | 10.4 | 0 | 0 |