Math on Math (pass@8)
61.1pass@8IGDS
Evaluation Results
| Method | Links | |
|---|---|---|
| IGDSModel=Qwen3-8B-Base, Data Selection Strategy=IGDS2026.04 | 61.1 | |
| FullModel=Qwen3-8B-Base, Data Selection Strategy=Full2026.04 | 60.8 | |
| IFDModel=Qwen3-8B-Base, Data Selection Strategy=IFD2026.04 | 60.5 | |
| ZIPModel=Qwen3-8B-Base, Data Selection Strategy=ZIP2026.04 | 60.3 | |
| LossModel=Qwen3-8B-Base, Data Selection Strategy=Loss2026.04 | 59.4 | |
| RandomModel=Qwen3-8B-Base, Data Selection Strategy=Random2026.04 | 58.7 | |
| OriginalModel=Qwen3-8B-Base, Data Selection Strategy=Original2026.04 | 55.7 | |
| IGDSModel=Llama-3.1-8B, Data Selection Strategy=IGDS2026.04 | 45.8 | |
| FullModel=Llama-3.1-8B, Data Selection Strategy=Full2026.04 | 45 | |
| RandomModel=Llama-3.1-8B, Data Selection Strategy=Random2026.04 | 40.8 | |
| IGDSModel=Gemma-2-2B, Data Selection Strategy=IGDS2026.04 | 37.8 | |
| LossModel=Llama-3.1-8B, Data Selection Strategy=Loss2026.04 | 37.2 | |
| IFDModel=Llama-3.1-8B, Data Selection Strategy=IFD2026.04 | 36.8 | |
| ZIPModel=Llama-3.1-8B, Data Selection Strategy=ZIP2026.04 | 35.4 | |
| FullModel=Gemma-2-2B, Data Selection Strategy=Full2026.04 | 32.2 | |
| IFDModel=Gemma-2-2B, Data Selection Strategy=IFD2026.04 | 30 | |
| ZIPModel=Gemma-2-2B, Data Selection Strategy=ZIP2026.04 | 29.4 | |
| RandomModel=Gemma-2-2B, Data Selection Strategy=Random2026.04 | 29.2 | |
| OriginalModel=Llama-3.1-8B, Data Selection Strategy=Original2026.04 | 26.6 | |
| LossModel=Gemma-2-2B, Data Selection Strategy=Loss2026.04 | 24.6 | |
| OriginalModel=Gemma-2-2B, Data Selection Strategy=Original2026.04 | 11.2 |