Machine Translation on Korean-Italian
18.02BLEUPIVOTE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| PIVOTEMerging Module=GPT-4o, Candidates=Direct & English Pivot2025.02 | 18.02 | 43.46 | 86.19 | |
| PIVOTEMerging Module=GPT-4o, Candidates=top32025.02 | 17.77 | 43.38 | 85.46 | |
| PIVOTEMerging Module=GPT-4, Candidates=Direct & English Pivot2025.02 | 17.1 | 43.29 | 85.92 | |
| PIVOTEMerging Module=GPT-4, Candidates=top32025.02 | 16.66 | 42.85 | 86.82 | |
| NLLBtype=Standalone NMT System2025.02 | 16.27 | 41.14 | 84.6 | |
| PIVOTEMerging Module=Llama-3, Candidates=top32025.02 | 15.6 | 39.86 | 84.1 | |
| GPT-4otype=Standalone NMT System2025.02 | 15.11 | 42.59 | 85.93 | |
| MBRtype=Prior Ensemble Method2025.02 | 14.1 | 42.24 | 86.7 | |
| GPT-4type=Standalone NMT System2025.02 | 14.07 | 42.22 | 86.8 | |
| PIVOTEMerging Module=Llama-3, Candidates=Direct & English Pivot2025.02 | 13.85 | 37.36 | 69.96 | |
| Llama-3type=Standalone NMT System2025.02 | 11.79 | 34.82 | 77.37 | |
| Baizetype=Standalone NMT System2025.02 | 10.62 | 31.87 | 73.62 | |
| Vicunatype=Standalone NMT System2025.02 | 10.11 | 31.15 | 70.29 | |
| LLM-Blendertype=Prior Ensemble Method2025.02 | 8.77 | 28.74 | 82.8 | |
| EVAtype=Prior Ensemble Method2025.02 | 2.53 | 15.26 | 39 |