Truthfulness Detection on True-False dataset
71.47AnimalsPRISM-SAPLMA
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| PRISM-SAPLMALLM=LLaMA2-7B-Chat2024.11 | 71.47 | 89.36 | 82.79 | 67.05 | 75.39 | 78.04 | 77.35 | |
| PRISM-MMLLM=LLaMA2-7B-Chat2024.11 | 70.04 | 90.6 | 79.08 | 63.85 | 70.31 | 77.56 | 75.24 | |
| SAPLMALLM=LLaMA2-7B-Chat2024.11 | 65.39 | 67 | 62.62 | 60.57 | 75.85 | 62.33 | 65.63 | |
| EUBHDLLM=LLaMA2-7B-Chat2024.11 | 56.83 | 61.97 | 59.63 | 60.04 | 55.63 | 55.91 | 58.34 | |
| LN-PPLLM=LLaMA2-7B-Chat2024.11 | 54.96 | 56.53 | 53.57 | 57.27 | 59.48 | 55.27 | 56.18 | |
| MMLLM=LLaMA2-7B-Chat2024.11 | 53 | 56.88 | 56.48 | 55.81 | 61.37 | 50.75 | 55.72 | |
| MINDLLM=LLaMA2-7B-Chat2024.11 | 46.26 | 46.64 | 49.97 | 48.5 | 50.25 | 50.23 | 48.64 |