diff --git a/PythonModels/scripts/run_test_mql_full_state_comparison.py b/PythonModels/scripts/run_test_mql_full_state_comparison.py
index 1ac96b4..5a7930d 100644
--- a/PythonModels/scripts/run_test_mql_full_state_comparison.py
+++ b/PythonModels/scripts/run_test_mql_full_state_comparison.py
@@ -53,6 +53,20 @@ class TestMqlFullStateComparisonRun:
     def signal_count(self) -> int:
         return len(self.output.data_paths)
 
+    def metrics_by_max_abs_error(self):
+        return tuple(
+            sorted(
+                self.comparison.metrics,
+                key=lambda metric: metric.max_abs_error,
+                reverse=True,
+            )
+        )
+
+    @property
+    def largest_abs_error_metric(self):
+        metrics = self.metrics_by_max_abs_error()
+        return metrics[0] if metrics else None
+
 
 @dataclass(frozen=True)
 class TestMqlFullStateComparisonPathConfig:
@@ -175,7 +189,14 @@ def format_test_mql_full_state_comparison_summary(
         f"Max absolute error: {run.comparison.max_abs_error}",
         f"Max relative error: {run.comparison.max_rel_error}",
     ]
-    for metric in run.comparison.metrics:
+    largest_metric = run.largest_abs_error_metric
+    if largest_metric is not None:
+        lines.append(
+            "Largest absolute error: "
+            f"{largest_metric.data_path}={largest_metric.max_abs_error}"
+        )
+    lines.append("Metrics by max absolute error:")
+    for metric in run.metrics_by_max_abs_error():
         lines.append(
             f"  - {metric.data_path}: max_abs_error={metric.max_abs_error}, "
             f"final_abs_error={metric.final_abs_error}"
diff --git a/tests/test_run_test_mql_full_state_comparison.py b/tests/test_run_test_mql_full_state_comparison.py
index a770b7a..fcb39f9 100644
--- a/tests/test_run_test_mql_full_state_comparison.py
+++ b/tests/test_run_test_mql_full_state_comparison.py
@@ -46,9 +46,15 @@ class RunTestMqlFullStateComparisonScriptTests(unittest.TestCase):
         self.assertEqual(run.sample_count, 1)
         self.assertEqual(run.signal_count, 2)
         self.assertAlmostEqual(run.comparison.max_abs_error, 0.0, delta=1.0e-8)
+        self.assertEqual(
+            run.metrics_by_max_abs_error()[0].data_path,
+            "press@pn_c1_8",
+        )
         self.assertTrue(csv_summary_exists)
         self.assertIn("Mode: Python 132 full-state closure comparison", summary)
         self.assertIn("Compared signals: 2", summary)
+        self.assertIn("Largest absolute error: press@pn_c1_8=", summary)
+        self.assertIn("Metrics by max absolute error:", summary)
         self.assertIn("python.press@pn_c1_8", csv_header)
         self.assertIn("amesim.vol1@pn_brp2_8", csv_header)
 
@@ -95,6 +101,8 @@ class RunTestMqlFullStateComparisonScriptTests(unittest.TestCase):
         self.assertIn("Model: test_mql", summary)
         self.assertIn("Output schema signals: 858", summary)
         self.assertIn("Compared signals: 1", summary)
+        self.assertIn("Largest absolute error: press@pn_c1_8=", summary)
+        self.assertIn("Metrics by max absolute error:", summary)
         self.assertTrue(summary.endswith("\n"))