Common-sense reasoning on OpenBookQA (OBQA) (In-Distribution)
88.24AccuracyDALorRA
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| DALorRABackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 88.24 | 3.12 | 0.36 | |
| TFBBackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 88.2 | 4.51 | 0.36 | |
| LABackbone=Llama3.1-8B, Fine-tuning method=LoRA2026.07 | 87.9 | 8.93 | 0.94 | |
| MLEBackbone=Llama3.1-8B, Fine-tuning method=LoRA2026.07 | 87.9 | 9.77 | 0.61 | |
| BLoBBackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 87.66 | 3.35 | 0.38 | |
| ENSBackbone=Llama3.1-8B, Fine-tuning method=LoRA2026.07 | 87.37 | 8.17 | 0.45 | |
| MCDBackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 87.2 | 9.76 | 0.62 | |
| C-LoRABackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 86.93 | 6.5 | 0.4 | |
| MAPBackbone=Llama3.1-8B, Fine-tuning method=LoRA2026.07 | 85.73 | 12.19 | 0.79 |