HH-RLHF on HH-RLHF
10.435Hyper-volumeMOC
Evaluation results
| Method | Links | |
|---|---|---|
| MOCBackbone=Llama3-8B2026.04 | 10.435 | |
| MOCBackbone=Llama2-7B2026.04 | 10.22 | |
| RiCBackbone=Llama2-7B2026.04 | 9.257 | |
| MODPOBackbone=Llama2-7B, Preference=[0.5, 0.5]2026.04 | 3.745 | |
| MetaAlignerBackbone=Llama2-7B2026.04 | 3.41 |