Factual Grounding Evaluation on TRUE (100-sample subset)
0.86ROC-AUCDeepchecks Grounded in Context
Evaluation results
| Method | Links | |
|---|---|---|
| Deepchecks Grounded in ContextEvaluator=Proprietary SLM2026.05 | 0.86 | |
| RAGAS FaithfulnessEvaluator LLM=GPT-4o2026.05 | 0.65 | |
| Langsmith Answer FaithfulnessEvaluator LLM=GPT-4o2026.05 | 0.63 |