Personalized Response Generation on Review Interpolated Users
84.6WinrateFSPO + RAT
Evaluation results
| Method | Links | |
|---|---|---|
| FSPO + RATNumber of shots=8, Optimization strategy=Few-Shot Preference Optimization + User Description Rationalization2025.02 | 84.6 | |
| FSPONumber of shots=8, Optimization strategy=Few-Shot Preference Optimization2025.02 | 73.6 | |
| FSPONumber of shots=4, Optimization strategy=Few-Shot Preference Optimization2025.02 | 71.3 | |
| Pref-FTNumber of shots=8, Optimization strategy=Preference Fine-Tuning2025.02 | 70.7 | |
| Pref-FTNumber of shots=4, Optimization strategy=Preference Fine-Tuning2025.02 | 66.1 | |
| PromptedNumber of shots=4, Optimization strategy=Prompted2025.02 | 61.9 | |
| PromptedNumber of shots=8, Optimization strategy=Prompted2025.02 | 59.1 | |
| Llama 3.2 3B InstructNumber of shots=0, Optimization strategy=Instruction-tuned2025.02 | 50 |