Response-level language alignment on OLA EN-Ko Simple
66.28EN->KO Embed Pass RateCS-DPO
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CS-DPOCS-DPO=true, Base Model=OLMo 3.1 (SFT+DPO)2026.01 | 66.28 | 99.65 | |
| OLMo 3.1 (SFT+DPO)CS-DPO=false2026.01 | 57.36 | 99.29 |
| Method | Links | ||
|---|---|---|---|
| CS-DPOCS-DPO=true, Base Model=OLMo 3.1 (SFT+DPO)2026.01 | 66.28 | 99.65 | |
| OLMo 3.1 (SFT+DPO)CS-DPO=false2026.01 | 57.36 | 99.29 |