CP on CP
359Generation Performance ScoreGreedy
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Greedythreshold (tau)=0.99, Base LLM=Qwen3-8B2026.05 | 359 | 411.93 | |
| Thompsonthreshold (tau)=0.999, Base LLM=Qwen3-8B2026.05 | 327 | 395.47 | |
| EXP3.Pthreshold (tau)=0.99, Base LLM=Qwen3-8B2026.05 | 207 | 256.12 | |
| Greedythreshold (tau)=0.95, Base LLM=Qwen3-8B2026.05 | 152 | 182.12 | |
| EXP3.Pthreshold (tau)=0.95, Base LLM=Qwen3-8B2026.05 | 139 | 173.57 | |
| Randthreshold (tau)=0.99, Base LLM=Qwen3-8B2026.05 | 138 | 170.88 | |
| UCBthreshold (tau)=0.99, Base LLM=Qwen3-8B2026.05 | 85 | 100.79 | |
| Thompsonthreshold (tau)=0.99, Base LLM=Qwen3-8B2026.05 | 85 | 101.22 | |
| ShinkaEvolvethreshold (tau)=0.99, Base LLM=Qwen3-8B2026.05 | 73 | 90.15 | |
| Randthreshold (tau)=0.95, Base LLM=Qwen3-8B2026.05 | 56 | 66.58 | |
| ShinkaEvolvethreshold (tau)=0.95, Base LLM=Qwen3-8B2026.05 | 45 | 57.9 | |
| UCBthreshold (tau)=0.95, Base LLM=Qwen3-8B2026.05 | 16 | 18.19 | |
| Thompsonthreshold (tau)=0.95, Base LLM=Qwen3-8B2026.05 | 16 | 18.19 |