Instruction Following on AlpacaEval (805 prompts)
7.22OLMo-2 Reward ScoreAAD
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| AADBackbone=OLMo-2-13B2025.09 | 7.22 | 98 | — | |
| EFTBackbone=OLMo-2-13B2025.09 | 6.97 | 96 | 55.8 | |
| Bo2Backbone=OLMo-2-13B2025.09 | 6.94 | 96 | 65.2 | |
| Greedy DPOBackbone=OLMo-2-13B2025.09 | 6.87 | 95 | 75.2 | |
| Greedy SFTBackbone=OLMo-2-13B2025.09 | 5.02 | — | 93.9 | |
| AADBackbone=OLMo-2-7B2025.09 | 3.84 | 98 | — | |
| EFTBackbone=OLMo-2-7B2025.09 | 3.5 | 96 | 58.1 | |
| Bo2Backbone=OLMo-2-7B2025.09 | 3.44 | 97 | 64.6 | |
| Greedy DPOBackbone=OLMo-2-7B2025.09 | 3.27 | 94 | 71.5 | |
| Greedy SFTBackbone=OLMo-2-7B2025.09 | 1.51 | — | 91.6 |