Mathematical Reasoning on GSM8K (FA, HS)
89.47FACWAC
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CWACModel=Gemma-2-9B2026.04 | 89.47 | 8.38 | |
| LisaModel=Gemma-2-9B2026.04 | 88.25 | 43.48 | |
| SFTModel=Gemma-2-9B2026.04 | 87.6 | 55.42 | |
| ASFTModel=Gemma-2-9B2026.04 | 83.31 | 47.53 | |
| CWACModel=Llama-3-8B2026.04 | 78.13 | 7.19 | |
| BEAModel=Gemma-2-9B2026.04 | 75.96 | 54.21 | |
| SFTModel=Llama-3-8B2026.04 | 75.89 | 66.74 | |
| SPPFTModel=Gemma-2-9B2026.04 | 73.2 | 25.64 | |
| CWACModel=Llama-2-7B2026.04 | 72.56 | 12.54 | |
| SFTModel=Mistral-7B2026.04 | 72.06 | 79.82 | |
| SFTModel=Llama-2-7B2026.04 | 70.81 | 52.17 | |
| CWACModel=Mistral-7B2026.04 | 68.76 | 25.63 | |
| BEAModel=Mistral-7B2026.04 | 64.27 | 53.32 | |
| SafeInstrModel=Gemma-2-9B2026.04 | 62.73 | 38.92 | |
| ASFTModel=Llama-2-7B2026.04 | 57.36 | 20.31 | |
| SPPFTModel=Llama-3-8B2026.04 | 54.67 | 15.27 | |
| BEAModel=Llama-2-7B2026.04 | 45.54 | 38.57 | |
| BEAModel=Llama-3-8B2026.04 | 43.57 | 32.24 | |
| SafeInstrModel=Mistral-7B2026.04 | 43.25 | 32.58 | |
| ASFTModel=Llama-3-8B2026.04 | 40.35 | 23.78 | |
| ASFTModel=Mistral-7B2026.04 | 39.76 | 45.63 | |
| LisaModel=Mistral-7B2026.04 | 38.62 | 56.7 | |
| SPPFTModel=Llama-2-7B2026.04 | 38.32 | 25.58 | |
| SPPFTModel=Mistral-7B2026.04 | 37.83 | 54.27 | |
| LisaModel=Llama-3-8B2026.04 | 35.34 | 28.42 | |
| LisaModel=Llama-2-7B2026.04 | 34.27 | 26.34 | |
| SafeInstrModel=Llama-2-7B2026.04 | 32.29 | 17.68 | |
| SafeInstrModel=Llama-3-8B2026.04 | 29.4 | 20.15 |