Combinatorial Search on N-Queens N=8
22.2AccuracySoftCoT
Evaluation Results
| Method | Links | |
|---|---|---|
| SoftCoTBase model=Qwen2.5, Transfer cost=full retrain2026.05 | 22.2 | |
| HyperGuideBase model=Qwen2.5, Transfer cost=small MLP2026.05 | 22.2 | |
| HyperGuideBase model=GPT-OSS, Transfer cost=small MLP2026.05 | 19.8 | |
| SoftCoTBase model=GPT-OSS, Transfer cost=full retrain2026.05 | 18.5 | |
| PT-SFTBase model=Mistral, Transfer cost=full retrain2026.05 | 16.1 | |
| HyperGuideBase model=Mistral, Transfer cost=small MLP2026.05 | 16.1 | |
| Tree of ThoughtsBase model=GPT-OSS, Transfer cost=N/A2026.05 | 13.6 | |
| Tree of ThoughtsBase model=Mistral, Transfer cost=N/A2026.05 | 13.6 | |
| Self-ConsistencyBase model=Qwen2.5, Transfer cost=N/A2026.05 | 11.1 | |
| Few-shotBase model=Mistral, Transfer cost=N/A2026.05 | 11.1 | |
| Self-ConsistencyBase model=Mistral, Transfer cost=N/A2026.05 | 11.1 | |
| Few-shotBase model=Qwen2.5, Transfer cost=N/A2026.05 | 9.9 | |
| PT-SFTBase model=Qwen2.5, Transfer cost=full retrain2026.05 | 9.9 | |
| PT-SFTBase model=GPT-OSS, Transfer cost=full retrain2026.05 | 9.9 | |
| SoftCoTBase model=Mistral, Transfer cost=full retrain2026.05 | 9.9 | |
| Few-shotBase model=GPT-OSS, Transfer cost=N/A2026.05 | 6.1 | |
| OVMBase model=Qwen2.5, Transfer cost=full retrain2026.05 | 4.9 | |
| Self-ConsistencyBase model=GPT-OSS, Transfer cost=N/A2026.05 | 4.9 | |
| OVMBase model=GPT-OSS, Transfer cost=full retrain2026.05 | 4.9 | |
| Tree of ThoughtsBase model=Qwen2.5, Transfer cost=N/A2026.05 | 3.7 | |
| OVMBase model=Mistral, Transfer cost=full retrain2026.05 | 3.7 |