scikit-learn
diff --git a/‎sklearn/metrics/_classification.py
Copy file name to clipboardExpand all lines: sklearn/metrics/_classification.py
+3-2Lines changed: 3 additions & 2 deletions b/‎sklearn/metrics/_classification.py
Copy file name to clipboardExpand all lines: sklearn/metrics/_classification.py
+3-2Lines changed: 3 additions & 2 deletions
diff --git a/‎sklearn/metrics/tests/test_common.py
Copy file name to clipboardExpand all lines: sklearn/metrics/tests/test_common.py
+31Lines changed: 31 additions & 0 deletions b/‎sklearn/metrics/tests/test_common.py
Copy file name to clipboardExpand all lines: sklearn/metrics/tests/test_common.py
+31Lines changed: 31 additions & 0 deletions
@@ -42,6 +42,7 @@
     _average,
     _union1d,
     get_namespace,
+    get_namespace_and_device,
 )
 from ..utils._param_validation import (
     Hidden,
@@ -217,13 +218,13 @@ def accuracy_score(y_true, y_pred, *, normalize=True, sample_weight=None):
     >>> accuracy_score(np.array([[0, 1], [1, 1]]), np.ones((2, 2)))
     0.5
     """
-
+    xp, _, device = get_namespace_and_device(y_true, y_pred, sample_weight)
     # Compute accuracy for each possible representation
     y_type, y_true, y_pred = _check_targets(y_true, y_pred)
     check_consistent_length(y_true, y_pred, sample_weight)
     if y_type.startswith("multilabel"):
         differing_labels = count_nonzero(y_true - y_pred, axis=1)
-        score = differing_labels == 0
+        score = xp.asarray(differing_labels == 0, device=device)
     else:
         score = y_true == y_pred
 
 
@@ -1826,6 +1826,35 @@ def check_array_api_multiclass_classification_metric(
     )
 
 
+def check_array_api_multilabel_classification_metric(
+    metric, array_namespace, device, dtype_name
+):
+    y_true_np = np.array([[1, 1], [0, 1], [0, 0]], dtype=dtype_name)
+    y_pred_np = np.array([[1, 1], [1, 1], [1, 1]], dtype=dtype_name)
+
+    check_array_api_metric(
+        metric,
+        array_namespace,
+        device,
+        dtype_name,
+        a_np=y_true_np,
+        b_np=y_pred_np,
+        sample_weight=None,
+    )
+
+    sample_weight = np.array([0.0, 0.1, 2.0], dtype=dtype_name)
+
+    check_array_api_metric(
+        metric,
+        array_namespace,
+        device,
+        dtype_name,
+        a_np=y_true_np,
+        b_np=y_pred_np,
+        sample_weight=sample_weight,
+    )
+
+
 def check_array_api_regression_metric(metric, array_namespace, device, dtype_name):
     y_true_np = np.array([[1, 3], [1, 2]], dtype=dtype_name)
     y_pred_np = np.array([[1, 4], [1, 1]], dtype=dtype_name)
@@ -1871,10 +1900,12 @@ def check_array_api_metric_pairwise(metric, array_namespace, device, dtype_name)
     accuracy_score: [
         check_array_api_binary_classification_metric,
         check_array_api_multiclass_classification_metric,
+        check_array_api_multilabel_classification_metric,
     ],
     zero_one_loss: [
         check_array_api_binary_classification_metric,
         check_array_api_multiclass_classification_metric,
+        check_array_api_multilabel_classification_metric,
     ],
     r2_score: [check_array_api_regression_metric],
     cosine_similarity: [check_array_api_metric_pairwise],