Common-sense reasoning on ARC-Challenge (In-Distribution)
82.55AccuracyENS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ENSBackbone=Llama3.1-8B, Fine-tuning method=LoRA2026.07 | 82.55 | 13.45 | 0.8 | |
| DALorRABackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 81.74 | 5.6 | 0.56 | |
| C-LoRABackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 81.7 | 10.75 | 0.68 | |
| MCDBackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 81.64 | 13.69 | 0.99 | |
| MLEBackbone=Llama3.1-8B, Fine-tuning method=LoRA2026.07 | 81.08 | 16.35 | 1.2 | |
| MAPBackbone=Llama3.1-8B, Fine-tuning method=LoRA2026.07 | 81.08 | 15.77 | 1.12 | |
| BLoBBackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 80.86 | 5.87 | 0.59 | |
| TFBBackbone=Llama3.1-8B, Fine-tuning method=LoRA, Monte Carlo samples (M)=102026.07 | 80.18 | 6.19 | 0.62 | |
| LABackbone=Llama3.1-8B, Fine-tuning method=LoRA2026.07 | 79.95 | 9.26 | 1.18 |