Machine Translation Ranking on GenMT22 (MQM) Zh→En
67.75AccuracyGemini-2.5-Pro
Evaluation results
| Method | Links | |
|---|---|---|
| Gemini-2.5-ProParadigm=GQM, Reasoning Strategy=with Reasoning2026.02 | 67.75 | |
| GRRM (RLVR)Paradigm=GQM, Reasoning Strategy=with Reasoning, Training Phase=RLVR2026.02 | 66 | |
| DeepSeek-R1-0528Paradigm=GQM, Reasoning Strategy=with Reasoning2026.02 | 65.79 | |
| BT-RMParadigm=SQM, Reasoning Strategy=without Reasoning2026.02 | 64.92 | |
| GRRM (SFT)Paradigm=GQM, Reasoning Strategy=with Reasoning, Training Phase=SFT2026.02 | 64.29 | |
| Gemini-2.5-ProParadigm=SQM, Reasoning Strategy=with Reasoning2026.02 | 61.99 | |
| CometKiwi-XXLParadigm=SQM, Reasoning Strategy=without Reasoning2026.02 | 61.6 | |
| SQM-GenRM (RLVR)Paradigm=SQM, Reasoning Strategy=with Reasoning, Training Phase=RLVR2026.02 | 60.37 | |
| DeepSeek-R1-0528Paradigm=SQM, Reasoning Strategy=with Reasoning2026.02 | 58.09 | |
| SQM-GenRM (SFT)Paradigm=SQM, Reasoning Strategy=with Reasoning, Training Phase=SFT2026.02 | 57.68 | |
| RandomParadigm=N/A2026.02 | 38.74 |