Classification on MTS-Specialty 200 random samples (test)
14Macro F1 ScoreQwen3-30B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen3-30BModel Category=New-gen general (Mnew)2026.01 | 14 | 35 | |
| CAPTComponents=Mnew + Mold-clin + Mold2026.01 | 13.5 | 32 | |
| Proxy TuningComponents=Mold-L + Mold-clin + Mold2026.01 | 9.5 | 27.5 | |
| LLaMA-2-70B-chatModel Category=Old-gen large general (Mold-L)2026.01 | 9.4 | 28 | |
| MeLLaMA-13B-chatModel Category=Old-gen clinical (Mold-clin)2026.01 | 8 | 21 | |
| UniTEComponents=Mnew + Mold-clin2026.01 | 6 | 10 |