Reward Modeling on HHH-Alignment Standard
91.8AccuracyArmoRM
Evaluation Results
| Method | Links | |
|---|---|---|
| ArmoRM2026.02 | 91.8 | |
| ICRMN (Number of in-context demonstrations)=322026.02 | 88.9 | |
| GRM2026.02 | 88.2 | |
| BTRM2026.02 | 86.4 | |
| ICRMN (Number of in-context demonstrations)=162026.02 | 86.4 | |
| ICRMN (Number of in-context demonstrations)=82026.02 | 82.8 | |
| ICRMN (Number of in-context demonstrations)=42026.02 | 79.9 | |
| ICRMN (Number of in-context demonstrations)=22026.02 | 68.2 | |
| ICRMN (Number of in-context demonstrations)=12026.02 | 64.6 |