Hallucination prediction on Explain original
80.91AccuracyConf + Probe (SCAO)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Conf + Probe (SCAO)Model=8B2025.09 | 80.91 | 15.26 | |
| Conf + Probe (SCAO)Model=LLaMA3-8B-Instruct2025.09 | 80.91 | 15.26 | |
| Conf + ProbeModel=8B2025.09 | 79.12 | 13.47 | |
| Conf + ProbeModel=LLaMA3-8B-Instruct2025.09 | 79.12 | 13.47 | |
| Probe_dnnModel=8B2025.09 | 78.92 | 13.27 | |
| ProbednnModel=LLaMA3-8B-Instruct2025.09 | 78.92 | 13.27 | |
| Conf + Probe (SCAO)Model=70B2025.09 | 78.75 | 16.79 | |
| Conf + Probe (SCAO)Model=LLaMA3-70B-Instruct2025.09 | 78.75 | 16.79 | |
| Probe_dnnModel=70B2025.09 | 76.28 | 14.32 | |
| ProbednnModel=LLaMA3-70B-Instruct2025.09 | 76.28 | 14.32 | |
| Conf + ProbeModel=70B2025.09 | 76 | 14.04 | |
| Conf + ProbeModel=LLaMA3-70B-Instruct2025.09 | 76 | 14.04 | |
| Conf (SCAO)Model=70B2025.09 | 60.99 | — | |
| Conf (SCAO)Model=LLaMA3-70B-Instruct2025.09 | 60.99 | — | |
| Conf (SCAO)Model=8B2025.09 | 59.71 | — | |
| Conf (SCAO)Model=LLaMA3-8B-Instruct2025.09 | 59.71 | — | |
| ConfModel=8B2025.09 | 57.66 | — | |
| ConfModel=LLaMA3-8B-Instruct2025.09 | 57.66 | — | |
| ConfModel=70B2025.09 | 55.71 | — | |
| ConfModel=LLaMA3-70B-Instruct2025.09 | 55.71 | — |