Semantic Tagging on USAS Chinese (test)
47.9Top-1 AccWSD Bi-Encoder Model
Evaluation results
| Method | Links | ||
|---|---|---|---|
| WSD Bi-Encoder ModelModel Type=Neural, Language Scope=Multilingual, Model Scale=Base2026.01 | 47.9 | 70.4 | |
| WSD Bi-Encoder ModelModel Type=Neural, Language Scope=Multilingual, Model Scale=Small2026.01 | 42.2 | 66.3 | |
| Hybrid Rule/Neural ModelModel Type=Hybrid, Language Scope=Multilingual, Model Scale=Small2026.01 | 39.8 | 55.6 | |
| Hybrid Rule/Neural ModelModel Type=Hybrid, Language Scope=Multilingual, Model Scale=Base2026.01 | 39.8 | 56.3 | |
| PyMUSASModel Type=Rule-based2026.01 | 32.6 | 43.6 |