ResearchDatasetsLatentAuditLatentAuditFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsFaithfulness DetectionLatentAudit Mistral-7B (evaluation)94AUROC6Faithfulness DetectionLatentAudit Qwen-2.5-7B (evaluation)94.5AUROC6Faithfulness DetectionLatentAudit Llama-3-8B (evaluation set)0.948AUROC6