Bandit Optimization on fLLM
35.6Cumulative RegretLLMP-Joint
Evaluation Results
| Method | Links | |
|---|---|---|
| LLMP-JointSeeds=52026.04 | 35.6 | |
| LLM-BanditSeeds=52026.04 | 36 | |
| LLMP-BanditSeeds=52026.04 | 40 | |
| LinUCBSeeds=52026.04 | 43.2 | |
| ThompsonSeeds=52026.04 | 54.6 | |
| CGPUCBSeeds=52026.04 | 65.6 | |
| RandomSeeds=52026.04 | 130.4 | |
| ZeroShotSeeds=52026.04 | 180.6 |