Machine Translation on WMT22 Ru-Sah news
58.2COMETChatGPT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ChatGPTMode=Dual-Reflection2024.10 | 58.2 | 74.2 | 2.6 | |
| IBUTMode=Iterative Bilingual Understanding (IBUT)2024.10 | 54.5 | 76.9 | 4.9 | |
| ChatGPTMode=Reranking2024.10 | 53.8 | 75.9 | 2.6 | |
| ChatGPTMode=Iterative Refinement2024.10 | 53.8 | 76.5 | 2.5 | |
| ChatGPTMode=Multi-Agent Discussion (MAD)2024.10 | 53.5 | 76.4 | 2.6 | |
| ChatGPTMode=Translation Error Analysis and Reduction (TEaR)2024.10 | 53.4 | 75.3 | 2.5 | |
| ChatGPTMode=Multi-Agent Prompt Selection (MAPS)2024.10 | 53.3 | 76.1 | 2.2 | |
| ChatGPTMode=5-shot prompting2024.10 | 53.1 | 75.4 | 2.3 | |
| ChatGPTMode=Zero-shot2024.10 | 52.8 | 73.2 | 1.9 |