Faithfulness Detection on LatentAudit Mistral-7B (evaluation)
94AUROC
GPT-4o Judge
Updated 1mo ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2026.04 | 94 | 87 | |
2026.04 | 92.5 | 85.2 | |
2026.04 | 89.5 | 82.5 | |
2026.04 | 87 | 80 | |
2026.04 | 85.8 | 79 | |
2026.04 | 71 | 64.2 |
GPT-4o Judge
| Method | Links | ||
|---|---|---|---|
2026.04 | 94 | 87 | |
2026.04 | 92.5 | 85.2 | |
2026.04 | 89.5 | 82.5 | |
2026.04 | 87 | 80 | |
2026.04 | 85.8 | 79 | |
2026.04 | 71 | 64.2 |