{ "Exact-Match": 0.038, "Token-F1": 0.4954, "BLEU-4": 0.2101, "BERTScore-F1": 0.5793, "Safety-Score": 0.8401, "Halluc-Rate": 0.378, "Overconf-Rate": 0.0, "Abstention-Rate": 0.004, "Mean-Confidence": 0.7397, "ECE": 0.7016, "N": 500 }