diff --git a/tensorflow_privacy/privacy/membership_inference_attack/data_structures.py b/tensorflow_privacy/privacy/membership_inference_attack/data_structures.py
index f915f55..d710671 100644
--- a/tensorflow_privacy/privacy/membership_inference_attack/data_structures.py
+++ b/tensorflow_privacy/privacy/membership_inference_attack/data_structures.py
@@ -566,7 +566,7 @@ class SingleRiskScoreResult:
     """
     meaningful_threshold_list, precision_list, recall_list = self.attack_with_varied_thresholds(threshold_list)
     summary = []
-    summary.append('\nPrivacy risk score analysis over slice: \"%s\"' %
+    summary.append('Privacy risk score analysis over slice: \"%s\"' %
                    str(self.slice_spec))
     for i in range(len(meaningful_threshold_list)):
       summary.append('  with %.5f as the threshold on privacy risk score, the precision-recall pair is (%.5f, %.5f)' %
@@ -589,6 +589,7 @@ class RiskScoreResults:
       single_summary = single_result.collect_results()
       for line in single_summary:
         summary.append(line)
+      summary.append('\n')
     return '\n'.join(summary)