Machine Translation on Neko (test)
83.42XCOMETNeoAMT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| NeoAMTBackbone=Qwen3-8B2026.01 | 83.42 | 5.1 | |
| Qwen3-14BModel Category=General Purpose LLMs2026.01 | 83.36 | 5.6 | |
| Qwen3-32BModel Category=General Purpose LLMs2026.01 | 83.33 | 5.2 | |
| X-ALMAModel Category=Translation-Specific LLMs2026.01 | 82.88 | 5.4 | |
| Hunyuan-MT-7BModel Category=Translation-Specific LLMs2026.01 | 82.36 | 5.2 | |
| Qwen3-8B GRPOBackbone=Qwen3-8B, Strategy=GRPO2026.01 | 82.3 | 5.5 | |
| GPT-OSS-20BModel Category=General Purpose LLMs2026.01 | 81.39 | 5.8 | |
| Qwen3-8BModel Category=General Purpose LLMs2026.01 | 81.12 | 5.7 | |
| Aya-expanse-8BModel Category=General Purpose LLMs2026.01 | 81.09 | 5.8 | |
| mExTrans-7BModel Category=Translation-Specific LLMs2026.01 | 80.81 | 5.7 | |
| Qwen3-8B SFTBackbone=Qwen3-8B, Strategy=SFT2026.01 | 80.6 | 5.6 | |
| MT-R1-Zero-7B-MixModel Category=Translation-Specific LLMs2026.01 | 79.86 | 6 | |
| Aya-23-8BModel Category=General Purpose LLMs2026.01 | 79.27 | 6.3 | |
| ALMA-R-7BModel Category=Translation-Specific LLMs2026.01 | 79.11 | 6.1 | |
| Llama-3.1-8B-InstructModel Category=General Purpose LLMs2026.01 | 78.81 | 6.6 | |
| TowerInstruct-7B-v0.2Model Category=Translation-Specific LLMs2026.01 | 78.03 | 6.6 | |
| Qwen3-8B RAGBackbone=Qwen3-8B, Strategy=RAG2026.01 | 75.42 | 6.7 |