Medical Task on MMLU Med
83.8PerformanceNTK-Selector
Evaluation Results
| Method | Links | |
|---|---|---|
| NTK-SelectorModel=QWEN3-8B2025.11 | 83.8 | |
| BaseModel=QWEN3-8B2025.11 | 83.4 | |
| DSIRModel=QWEN3-8B2025.11 | 82.7 | |
| RandomModel=QWEN3-8B2025.11 | 82.6 | |
| LESSModel=QWEN3-8B2025.11 | 82.6 | |
| TSDSModel=QWEN3-8B2025.11 | 82.3 | |
| GradientModel=QWEN3-8B2025.11 | 82.2 | |
| Domain-OnlyModel=QWEN3-8B2025.11 | 81.9 | |
| EmbeddingModel=QWEN3-8B2025.11 | 81.5 | |
| GradientModel=LLAMA3-8B-INSTRUCT2025.11 | 74 | |
| NTK-SelectorModel=LLAMA3-8B-INSTRUCT2025.11 | 73.8 | |
| LESSModel=LLAMA3-8B-INSTRUCT2025.11 | 73.1 | |
| RandomModel=LLAMA3-8B-INSTRUCT2025.11 | 73 | |
| EmbeddingModel=LLAMA3-8B-INSTRUCT2025.11 | 72.7 | |
| DSIRModel=LLAMA3-8B-INSTRUCT2025.11 | 72.7 | |
| BaseModel=LLAMA3-8B-INSTRUCT2025.11 | 72.3 | |
| Domain-OnlyModel=LLAMA3-8B-INSTRUCT2025.11 | 71.8 | |
| TSDSModel=LLAMA3-8B-INSTRUCT2025.11 | 71.4 |