Value Alignment on HH Balance-8
4.317Conformity ScorePICACO
Evaluation Results
| Method | Links | |
|---|---|---|
| PICACOTarget Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 4.317 | |
| OPROTarget Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 4.303 | |
| CICLTarget Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 4.282 | |
| PICACO2025.07 | 4.257 | |
| URIALTarget Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 4.248 | |
| Q+IF+COTTarget Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 4.204 | |
| Q+IFModel=GPT-4o-mini2025.07 | 4.193 | |
| Q+IFTraining=SFT2025.07 | 4.121 | |
| URIAL+SUMTarget Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 4.086 | |
| Q+IF2025.07 | 4.082 | |
| MODULARTarget Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 4.077 | |
| MP+SYSTEM 2Target Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 3.939 | |
| Q+IFTarget Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 3.934 | |
| QTarget Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 3.884 | |
| MP+SYSTEM 1Target Model=Gemini-1.5-Flash, Judge=GPT-4o2025.07 | 3.877 | |
| Q+IFTraining=SFT+2025.07 | 3.314 | |
| URIALTraining=SFT2025.07 | 2.491 |