Question Answering on ARC Challenge (Mean Improvement and Statistical Significance)
0.066Mean ImprovementPASf
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| PASfModel=DeepSeek-R1-Distill-8B2025.09 | 0.066 | 0.046 | 0 | |
| iPASaModel=DeepSeek-R1-Distill-8B2025.09 | 0.051 | 0.037 | 0 | |
| iPASwoModel=DeepSeek-R1-Distill-8B2025.09 | 0.039 | 0.022 | 0 | |
| iPASaModel=Llama-3.1-8B-Instruct2025.09 | 0 | -0.01 | 0.5 | |
| PASfModel=Llama-3.1-8B-Instruct2025.09 | 0 | -0.01 | 0.84 | |
| iPASwoModel=Nous-Hermes-2-Mistral-7B-DPO2025.09 | -0.003 | -0.012 | 0.77 | |
| PASfModel=Nous-Hermes-2-Mistral-7B-DPO2025.09 | -0.004 | -0.01 | 0.9 | |
| iPASaModel=Nous-Hermes-2-Mistral-7B-DPO2025.09 | -0.005 | -0.011 | 0.96 | |
| iPASwoModel=Llama-3.1-8B-Instruct2025.09 | -0.01 | -0.02 | 0.95 |