Factual Hallucination on FACTOR Wiki
58.71AccuracyDeLask
Evaluation Results
| Method | Links | |
|---|---|---|
| DeLaskModel=LLaMA3-8B, Decoding=DeLask2026.05 | 58.71 | |
| DeLaskModel=LLaMA2-7B, Decoding=DeLask2026.05 | 57.85 | |
| RegularModel=LLaMA3-8B, Decoding=Regular2026.05 | 57.77 | |
| ITIModel=LLaMA3-8B, Decoding=ITI2026.05 | 57.71 | |
| UADModel=LLaMA3-8B, Decoding=UAD2026.05 | 56.81 | |
| ADModel=LLaMA2-7B, Decoding=AD2026.05 | 56.71 | |
| DOLAModel=LLaMA2-7B, Decoding=DOLA2026.05 | 56.51 | |
| ITIModel=LLaMA2-7B, Decoding=ITI2026.05 | 56.21 | |
| UADModel=LLaMA2-7B, Decoding=UAD2026.05 | 55.89 | |
| RegularModel=LLaMA2-7B, Decoding=Regular2026.05 | 55.77 | |
| DOLAModel=LLaMA3-8B, Decoding=DOLA2026.05 | 54.47 | |
| ADModel=LLaMA3-8B, Decoding=AD2026.05 | 53.72 |