LatentAudit
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Faithfulness Detection | LatentAudit Mistral-7B (evaluation) | AUROC94 | 6 | |
| Faithfulness Detection | LatentAudit Qwen-2.5-7B (evaluation) | AUROC94.5 | 6 | |
| Faithfulness Detection | LatentAudit Llama-3-8B (evaluation set) | AUROC0.948 | 6 |
Showing 3 of 3 rows